توقع التغيرات في غلات الزراعة تحت سيناريوهات تغير المناخ وآثارها على الأمن الغذائي العالمي Predicting changes in agricultural yields under climate change scenarios and their implications for global food security

المجلة: Scientific Reports، المجلد: 15، العدد: 1
DOI: https://doi.org/10.1038/s41598-025-87047-y
PMID: https://pubmed.ncbi.nlm.nih.gov/39843615
تاريخ النشر: 2025-01-22

تقارير علمية

مفتوح

توقع التغيرات في غلات الزراعة تحت سيناريوهات تغير المناخ وآثارها على الأمن الغذائي العالمي

كريستين لي □ , جيمس كاماك , أندرو روبنسون & توم كومباس

لتغير المناخ آثار مباشرة على الإنتاجية الزراعية الحالية والمستقبلية. يمكن استخدام نماذج التحليل الإحصائي لتوليد توقعات استجابة غلة المحاصيل للعوامل المناخية من خلال تجميع البيانات من التجارب المنضبطة. ومع ذلك، فإن التحديات المنهجية في إجراء هذه التحليلات الميتا، جنبًا إلى جنب مع عدم اليقين المدمج من مصادر مختلفة، تجعل من الصعب التحقق من صحة نتائج النماذج. نقدم تحديثات للتقديرات المنشورة لاستجابة غلة المحاصيل لدرجات الحرارة المتوقعة، وهطول الأمطار، وأنماط ثاني أكسيد الكربون، ونظهر أن نماذج التأثيرات المختلطة تؤدي بشكل أفضل من نماذج OLS المجمعة من حيث خطأ الجذر التربيعي المتوسط (RMSE) والانحراف المفسر، على الرغم من الاستخدام الشائع لنماذج OLS المجمعة في التحليلات الميتا السابقة. بناءً على تحليلنا، قد يؤدي استخدام نماذج OLS المجمعة إلى تقدير خسائر الغلة بشكل أقل. نستخدم أيضًا نهج البلوك-بوتستراب لتحديد عدم اليقين عبر أبعاد متعددة، بما في ذلك اختيارات النماذج، وتوقعات المناخ من مشروع المقارنة بين النماذج المتصلة السادس (CMIP6)، وسيناريوهات الانبعاثات من المسارات الاجتماعية والاقتصادية المشتركة (SSP). تظهر تقديراتنا استجابات غلة متوقعة تبلغ – (الذرة)، – (الأرز)، – (الصويا)، و 14% (القمح) من 2015 إلى 2080-2100 تحت سيناريو الأعمال كالمعتاد من SSP5-8.5، والتي تقل إلى , و على التوالي تحت سيناريو الانبعاثات المنخفضة من SSP1-2.6. بدون التخفيف والتكيف، قد تصبح الدول في جنوب آسيا، وأفريقيا جنوب الصحراء الكبرى، وأمريكا الشمالية، وأوقيانوسيا في خطر عدم القدرة على تلبية الطلب الوطني من السعرات الحرارية بحلول نهاية القرن تحت سيناريو الانبعاثات الأكثر شدة.
من المتوقع أن يؤثر تغير المناخ بشكل مباشر على الإنتاج الزراعي من خلال تقليل كل من غلة المحاصيل والجودة عبر تغيير أنماط درجات الحرارة والمياه والغازات والمواد الغذائية. علاوة على ذلك، يمكن أن يكون لتغير المناخ أيضًا آثار غير مباشرة على الغلات من خلال تغيير التأثيرات الناتجة عن الآفات والأمراض والأعشاب الضارة. يمكن تقدير نماذج الاستجابة لغلة المحاصيل لتغير المناخ باستخدام بيانات من التجارب المنضبطة. يمكن استخدام هذه النماذج لتوليد توقعات لتأثيرات المناخ في بيئات جديدة مثل الفترات الزمنية المستقبلية أو الجغرافيا التي تفتقر نسبيًا إلى الدراسات الأولية. . التوقعات من مثل هذه النماذج ذات صلة كبيرة بسياسة المناخ – تم تضمين وظائف استجابة غلة المناخ في عدد من نماذج التقييم المتكاملة (IAMs) وتقييمات الهيئة الحكومية الدولية المعنية بتغير المناخ (IPCC) لتأثيرات تغير المناخ على إنتاج الغذاء. . ومع ذلك، من الصعب التحقق من الفائدة العامة لوظائف استجابة الغلة المطبقة على ‘إعدادات خارج العينة’. ترجع هذه الصعوبة جزئيًا إلى التحديات المنهجية في تقدير العلاقات العامة ونقل هذه العلاقات عبر الإعدادات، وجزئيًا لأن مصادر عدم اليقين المختلفة طوال عملية النمذجة يمكن أن تتجمع وتنتشر بطرق متداخلة لتوقعات النمذجة. تقدير العلاقات العامة باستخدام البيانات المجمعة من الدراسات – تقنية التحليل الميتا – مليء بالتحديات المنهجية التي يمكن أن تؤدي إلى تمثيل مفرط أو ناقص لبعض الأنماط في البيانات. وجدت تحليل ميتا سابق لتقديرات تأثير المناخ أدلة قوية على تحيز التكرار من الارتباط بين الدراسات، فضلاً عن الارتباط داخل الدراسة من تضمين تقديرات متعددة. . في حالة استجابات الغلة لتغير المناخ، يمكن أن تؤدي مثل هذه التحديات إلى استجابات غلة أعلى أو أقل مما هو مبرر بالنظر إلى الهيكل الحقيقي للبيانات. بمجرد تقدير العلاقات العامة، فإن كيفية نقل هذه التقديرات إلى إعدادات جديدة وتفسيرها في سياقها تخضع أيضًا للعديد من عدم اليقين. على مستوى أساسي، هناك درجة كبيرة من التباين عبر استراتيجيات أخذ عينات بيانات نماذج تأثير المناخ، والاختيارات لتجاهل أو تقدير القيم المفقودة، والاختيارات أو الافتراضات المتعلقة بمواصفات النموذج ومعايرته مع بيانات الإدخال. من المفيد
أن يوضح المودلون هذه الغموض، وهو ما نقوم به في هذه الدراسة. للذهاب أبعد من ذلك، غالبًا ما يُشار إلى مسار التنمية الاقتصادية العالمية وتأثير انبعاثات البشر على أنظمة الأرض والمناخ على أنه غير قابل للاختزال، حيث قد لا تقلل المزيد من المعلومات بالضرورة من هذا عدم اليقين. . لا يمكن تقليص هذه ‘اللا يمكن تقليصها’ أو ‘العميقة’ من عدم اليقين، لكننا نؤكد أن نماذج التأثير يجب أن تهدف إلى تضمين مدخلات من مجموعة واسعة من نماذج المناخ وسيناريوهات الانبعاثات، مما سيساعد المستخدمين وصانعي السياسات على فهم نطاق التأثيرات المحتملة بشكل أفضل. الهدف هو توضيح مصادر عدم اليقين، مما يسهل الثقة واستخدام توقعات استجابة الغلة على مستويات صنع السياسات.
لدراستنا ثلاثة أهداف: (1) تحسين تقدير استجابات الغلة للعوامل المناخية مثل درجة الحرارة، وهطول الأمطار وثاني أكسيد الكربون، باستخدام بيانات من قاعدة بيانات استجابة غلة المحاصيل المعتمدة (بيانات CGIAR)، (2) تحليل نطاق عدم اليقين في التوقعات الناتجة إلى مصادرها المنفصلة، و (3) تقديم تقديرات ذات صلة بالسياسة تتعلق بالأمن الغذائي العالمي. كان هدفنا هو تقديم تقدير شامل ولكنه محدث لاستجابات غلة المناخ العالمية مع مراعاة الآثار على الأمن الغذائي وعدم اليقين المرتبط تحت مجموعة أكبر من سيناريوهات الانبعاثات والفترات الزمنية. أولاً، افترضنا أن ملاءمة النماذج المختلطة لبيانات CGIAR ستساعد في حساب الارتباط داخل الدراسة بين تقديرات متعددة في مجموعة البيانات، مما يؤدي إلى أداء أفضل للنموذج الإحصائي مقارنة بنماذج OLS المجمعة المستخدمة في التحليلات الميتا السابقة لنفس مجموعة البيانات. . ثم يمكننا مقارنة التوقعات العامة لاستجابات غلة المحاصيل لتغير المناخ التي تم الحصول عليها من خلال اتباع هذا النهج، بالتوقعات المستمدة من هذه التحليلات الميتا القديمة. ثانيًا، استخدمنا تقنية أخذ عينات البوتستراب المتكررة على مجموعة البيانات عبر أبعاد متعددة من عدم اليقين، لنرى كيف تنتشر مصادر عدم اليقين المختلفة طوال عملية النمذجة إلى استجابات الغلة العالمية المقدرة. سمح لنا ذلك بتحديد حصة التباين التي تعزى إلى أخذ عينات البيانات، والقيم المفقودة، ومواصفات النموذج، وبيانات إدخال نموذج المناخ من مشروع المقارنة بين النماذج المتصلة السادس (CMIP6) من مجموعة نماذج الدورة العالمية (GCMs). أخيرًا، قدرنا تأثيرات استجابات غلة المحاصيل المتوقعة باستخدام ليس فقط درجات الحرارة المتوقعة في المستقبل ولكن أيضًا هطول الأمطار المتوقع في المستقبل، وهو ما لم يتم القيام به في التحليلات الميتا السابقة التي تم ملاءمتها على نفس مجموعة البيانات. قمنا بتطبيق مخرجات نموذج الاستجابة المفضل لحساب فجوات السعرات الحرارية على مستوى الدولة وحالة الأمن الغذائي المحلي لثلاثة سيناريوهات انبعاثات ذات شدة متفاوتة.

النتائج

وظائف استجابة الغلة المودلة

بدأنا بتحديث قاعدة بيانات CGIAR من خلال فحص وتقدير البيانات المفقودة. تجمع هذه القاعدة 74 دراسة تتعلق بالذرة، والأرز، والصويا، والقمح (تشمل من السعرات الحرارية المستهلكة عالميًا في ) تحتوي على أكثر من 8800 تقدير نقطي للتغيرات في غلة المحاصيل عبر درجات حرارة وهطول أمطار متغيرة، و وعوامل أخرى. لم نضف دراسات إلى قاعدة البيانات؛ لذلك تظل مجموعة غير شاملة من المعلومات حول استجابات الغلة للعوامل المناخية. تحتوي قاعدة بيانات CGIAR على هيكل متداخل أو هرمي حيث يمكن أن تأتي تقديرات متعددة من دراسة واحدة؛ يجب اعتبار هذه التقديرات مرتبطة أو غير مستقلة. وبالمثل، يتم اشتقاق العديد من التقديرات من نفس المنطقة أو الدولة، ونفترض أيضًا أن هذه مرتبطة.
نقوم بتطبيق نماذج إحصائية لكل محصول على حدة (انظر قسم “الطرق” للحصول على المواصفات الكاملة). نماذجنا الخمسة المرشحة هي: (1) نموذج المربعات الصغرى العادية (OLS) المجمعة، (2) نموذج مختلط خطي عام (GLMM) مع تقاطعات عشوائية، (3) GLMM مع تقاطعات عشوائية ومنحدرات عشوائية، (4) نموذج مختلط إضافي عام (GAMM) مع تقاطعات عشوائية و(5) GAMM مع تقاطعات عشوائية ومنحدرات عشوائية، كما هو موضح في الشكل 1. النموذج 1 هو نموذج OLS مجمع يعامل التقديرات النقطية من نفس الدراسة والدولة كما لو كانت مستقلة، وهو مشابه لنماذج OLS المستخدمة في التحليل التلوي المنشور سابقًا على نفس مجموعة بيانات CGIAR، باستثناء أن دوال الاستجابة لدينا تُقدّر بشكل منفصل لكل من المحاصيل الأربعة، ولا تشمل مصطلح تفاعل بين التغير في درجة الحرارة التربيعية ودرجة الحرارة الأساسية (انظر قسم “الطرق”). تشمل النماذج 2 و4 تقاطعات عشوائية للدراسات والدول في مجموعة البيانات، حيث تأخذ تقاطعات الدراسة العشوائية في الاعتبار الارتباط بين تقديرات متعددة من نفس الدراسة، وتأخذ تقاطعات الدولة العشوائية في الاعتبار الارتباط بين الدراسات التي تشمل تقديرات لنفس الدولة. يساعد تضمين تقاطعات الدولة العشوائية في التقاط العوامل غير المناخية المرتبطة بالموقع، مثل قابلية الإصابة بالآفات والأمراض، خصوبة التربة، وسلوك الزراعة، كما يسمح أيضًا بتوقع التقديرات الملائمة في دول جديدة غير مدرجة في بيانات CGIAR. تشمل النماذج 3 و5 تقاطعات عشوائية للدراسات والدول ولكن أيضًا منحدرات عشوائية للدول، مما يسمح بتغير دوال درجة الحرارة، والهطول وثاني أكسيد الكربون لكل دولة (أي تقدير دالة منفصلة لكل دولة). تسمح التأثيرات العشوائية بتقسيم التباين المتبقي بين المستويات الهرمية للدول والدراسات في البيانات. تعتبر GAMMs نوعًا غير بارامتري من النموذج الخطي العام الذي يسمح للمتغير الاستجابي (هنا تغيير العائد) بالاعتماد على دوال سلسة لمصطلحات النموذج. تم اختيار مصطلحات النموذج بناءً على الأدبيات السابقة وملاءمتها لنمو المحاصيل (انظر قسم “الطرق”). تم معايرة دوال الاستجابة لتغيرات في متوسط درجة الحرارة، والهطول وظروف ثاني أكسيد الكربون، وبالتالي، لا تمثل التأثيرات الناتجة عن درجات الحرارة المتطرفة والمتغيرة وأحداث الأمطار بما في ذلك الفيضانات والجفاف، ولا تلتقط تقلبات التربة والعناصر الغذائية على المدى الطويل.
وجدنا أن دوال الاستجابة لدرجة الحرارة والعائد تميل إلى الانحدار لأسفل للذرة، الأرز والقمح، مما يشير إلى أن تأثيرات درجة الحرارة بمفردها تميل إلى أن يكون لها تأثير مخفض على العائدات (الشكل 1أ). لا يُرى هذا الاستجابة في فول الصويا. تصف دوال الاستجابة للهطول علاقة إيجابية بين التغيرات في الأمطار والعائدات لجميع المحاصيل باستثناء الأرز، وقد يكون ذلك بسبب أن الأرز يُروى في الغالب (الشكل 1ب). لأغراض التصوير، تُظهر هذه الرسوم البيانية دوال الاستجابة عند مستويات خط الأساس المتوسطة من درجة الحرارة والهطول، لكنها لا تُظهر التفاعلات بين التغيرات في درجة الحرارة والهطول، أو تركيبات من مستويات خط الأساس المختلفة مع التغيرات. ما يمكن أيضًا رؤيته من نقاط البيانات المرسومة هو تقديرات متعددة من نفس الدراسة، والتي قد تمثل تجارب فريدة تختبر الاستجابات لتركيبات من
الشكل 1. دوال الاستجابة المتوسطة الهامشية ( ) لتغير درجة الحرارة (أ) وتغير الهطول (ب)، مشروطة على درجة حرارة خط الأساس المتوسطة، هطول خط الأساس المتوسطة، وغياب التغيرات في الهطول (أ) أو التغيرات في درجة الحرارة (ب)، التغيرات في وتكيف المزارع. النقاط تمثل تقديرات الدراسة المستمدة من بيانات CGIAR. تُظهر أشرطة الخطأ فترة الثقة 95% التي تجمع التباين داخل وبين 5 تقديرات ملائمة معززة. الأسطورة: GLM RI، نموذج مختلط خطي عام (GLMM) مع تقاطعات عشوائية؛ GLM RS، GLMM مع تقاطعات ومنحدرات عشوائية؛ GAM RI، نموذج مختلط إضافي عام (GAMM) مع تقاطعات عشوائية؛ GAM RS، GAMM مع تقاطعات ومنحدرات عشوائية؛ LM، نموذج المربعات الصغرى العادية (OLS) المجمعة. الشكل التوضيحي 3 يُظهر بعض المقارنات خارج العينة التي لم تُدرج في قاعدة بيانات CGIAR. .
محصول نموذج عينة نماذج المناخ العالمية البيانات المفقودة مجموعة
ذرة 10.9 8.67 4.52 8.58 32.7
أرز 19.3 12.0 9.46 12.4 53.1
فول 51.1 48.2 18.6 47.5 165.0
قمح 15.3 8.57 6.08 8.64 38.6
الجدول 1. مصادر عدم اليقين الناتجة عن اختيار النموذج، العينة، نماذج المناخ العالمية ومجموعات البيانات المعززة (البيانات المفقودة)، كما تم قياسها بواسطة متوسط الانحراف المعياري لتغير العائدات العالمية الموزونة (%). عدم اليقين المجمّع لا يضيف إلى حيث أن عدم اليقين ليس مستقلًا عبر المصادر.
درجة الحرارة، والهطول وثاني أكسيد الكربون. تثير هذه الدراسات متعددة التقديرات الحاجة إلى نمذجة التأثيرات المختلطة لأخذ في الاعتبار انحياز الارتباط داخل الدراسة.

مصادر عدم اليقين

قمنا بتحديد عدم اليقين بشكل منهجي من خلال أخذ 100 عينة من البيانات باستخدام طريقة البلوك-بووتستراب ومقارنة متوسط الانحراف المعياري الناتج للاستجابات العالمية المقدرة عبر خمسة أبعاد. تشمل هذه الأبعاد: (1) 100 عينة من البلوك-بووتستراب، مع حجب حسب الدراسة حيث يتم اختيار مجموعة مختلفة من الدراسات في كل عينة (أي عدم اليقين بشأن استراتيجية العينة لتجميع مجموعة بيانات CGIAR)، (2) خمس تعويضات مختلفة لمجموعة بيانات CGIAR لملء القيم المفقودة، باستخدام معادلات التعويض المتعددة المتسلسلة (MICE، انظر قسم “الطرق”) (أي عدم اليقين بشأن معالجة البيانات المفقودة)، (3) خمسة نماذج إحصائية (أي عدم اليقين بشأن المواصفة الاقتصادية)، (4) 23 مجموعة من بيانات المدخلات لدرجة الحرارة، والهطول وثاني أكسيد الكربون من مجموعة CMIP6 من 23 نموذج مناخ عالمي، لكل من ثلاثة سيناريوهات انبعاثات (أي عدم اليقين بشأن تأثير انبعاثات الإنسان على نظام المناخ) (انظر قسم “الطرق”).
وجدنا أن عدم اليقين الناتج عن اختيار النموذج يساوي بين 10 و من العائدات الزراعية العالمية للذرة، الأرز والقمح وأكثر من لفول الصويا (الجدول 1). جميع أنواع عدم اليقين وعدم اليقين المجمّع هي الأعلى في فول الصويا، تليها الأرز، القمح والذرة. يعكس هذا تقريبًا أحجام العينات المختلفة لتقديرات CGIAR غير المعوضة المتعلقة بكل محصول (الذرة: 3426؛ الأرز: 2776؛ فول: 788؛ القمح: 1972). تعكس نتائج عدم اليقين العالية حول فول الصويا حجم العينة المنخفض والتباين الكبير في هذه التقديرات. التباينات عبر نماذج المناخ العالمية هي الأدنى مقارنة بأنواع عدم اليقين الأخرى عبر جميع المحاصيل. الشكل 2 يُظهر 575 توزيعًا لنماذج تقديرات التأثير من 2081-2100 (تجميع التوزيعات عبر نماذج المناخ العالمية لتحسين الوضوح).
الشكل 2. توزيعات التغير المتوقع في العائدات العالمية الموزونة في 2081-2100 من 575 مجموعة نماذج لكل محصول (5 مواصفات نموذجية مجموعات البيانات المعززة نماذج المناخ العالمية)، مجمعة حسب نموذج المناخ العالمي لتمثيل فقط الاختلافات عبر اختيار النموذج ومجموعة البيانات المعززة. الأسطورة: GLM_RI، نموذج مختلط خطي عام (GLMM) مع تقاطعات عشوائية؛ GLM_RS، GLMM مع تقاطعات ومنحدرات عشوائية؛ GAM_RI، نموذج مختلط إضافي عام (GAMM) مع تقاطعات عشوائية؛ GAM_RS، GAMM مع تقاطعات ومنحدرات عشوائية؛ LM، نموذج OLS المجمّع.
يولد نموذج OLS المجمّع توزيعًا أضيق للاستجابات المقدرة مقارنةً بكل من GLMMs للذرة، فول الصويا والقمح، ومقارنةً بنماذج GAM للأرز. وهذا يشير إلى أن عدم أخذ الارتباطات المتعلقة بالدراسة والدولة في البيانات في الاعتبار قد يؤدي إلى تقديرات منخفضة لعدم اليقين في التنبؤ.

أداء النموذج

قمنا بتقييم أداء النموذج التنبؤي من خلال إجراء اختبار التحقق المتقاطع k-fold مع عشرة طيات. وجدنا أن GAMM وGLMM مع تقاطعات عشوائية ومنحدرات عشوائية أدت أفضل أداء كما تم قياسه بواسطة الجذر التربيعي لمتوسط الخطأ (RMSE)، وأدى نموذج OLS المجمّع أسوأ أداء. الشكل 3 يُظهر ذلك جنبًا إلى جنب مع الانحراف المفسر لكل من المواصفات الخمس للنموذج (كل منها مجمعة عبر خمس تقديرات ملائمة معززة). كان أداء نموذج OLS المجمّع هو الأسوأ، مع أعلى متوسط RMSE للنموذج وأقل انحراف مفسر عبر جميع المواصفات الخمس للنموذج.
أداء نماذج GLMM و GAMM مع تقاطعات وانحدارات عشوائية أفضل مقارنة بجميع نماذج GLMM و GAMM مع تقاطعات عشوائية فقط. وهذا ليس مفاجئًا نظرًا للاختلاف الهيكلي في الدراسات عبر قاعدة بيانات CGIAR؛ مما يسمح بمتغيرات المناخ (درجة الحرارة، هطول الأمطار و التغيير) للتنوع بشكل عشوائي لكل دراسة يساعد في استخراج العلاقات ‘الحقيقية’ من كل دراسة. ومع ذلك، فإن الطبيعة المرنة لنماذج GAMM غير المعلمية (حيث لا نقيد عدد العقد في الأشكال) وشكل دوال الاستجابة الخاصة بها تشير إلى أن GAMM مع تقاطعات وانحدارات عشوائية قد يتجاوز ملاءمة البيانات. من بين النماذج الخمسة، نفضل نموذج GLMM مع تقاطعات وانحدارات عشوائية لأنه يميل إلى تجاوز ملاءمة البيانات بشكل أقل مقارنة بنموذج GAMM على الرغم من أن أدائه التنبؤي جيد بشكل مشابه.

تغيرات العائد المتوقع على المستوى العالمي

استخدمنا التقديرات الملائمة من النموذج المفضل (نموذج الانحدار الخطي العام مع تقاطعات وانحدارات عشوائية) لتقدير التغيرات المستقبلية المتوقعة في العائد على شبكة ( عند خط الاستواء) وقمنا بتحويل هذه التأثيرات إلى مقاييس سعرات حرارية. استخدمنا هذه المقاييس لتحليل مخاطر انعدام الأمن الغذائي على مستوى الدول من 2021 إلى 2100. نعتبر ثلاثة سيناريوهات انبعاثات باستخدام تركيبات المسارات الاجتماعية والاقتصادية المشتركة (SSPs) في CMIP6: SSP1-2.6 و SSP2-4.5 و SSP5-8.5. تماشيًا مع التقرير السادس للهيئة الحكومية الدولية المعنية بتغير المناخ (IPCC) SSP1-2.6 يوضح سيناريو للاستدامة مع نمو منخفض نسبيًا في السكان والاستهلاك وتركيز غازات الدفيئة (GHG)، مع أفضل تقدير لـ ارتفاع متوسط درجة حرارة سطح الأرض العالمية بحلول 2081-2100 مقارنةً بالفترة من 1850-1900. سيناريو SSP2-4.5 هو سيناريو سكاني معتدل، مع تحديات متوسطة للتخفيف والتكيف. الاحترار بحلول 2081-2100. سيناريو SSP5-8.5 هو سيناريو للنمو الاقتصادي السريع جنبًا إلى جنب مع استمرار العولمة والتنمية المعتمدة على الوقود الأحفوري، مع تركيزات عالية من غازات الدفيئة، مع الاحترار بحلول 2081-2100.
نظهر توقعات عالمية مجمعة من CMIP6 للنموذج المفضل لـ SSP5-8.5 في الشكل 4 ولـ SSP12.6 و SSP2-4.5 في الشكلين التكميلين 5 و 6. أكبر تخفيضات في العائدات تُلاحظ في الذرة، مع خسائر قدرها تحت SSP1-2.6 و تحت سيناريو SSP5-8.5 في الفترة من 2081 إلى 2100 (الجدول 2، 3)، وهو ما يتماشى مع دراسات أخرى غير مدرجة في مجموعة بيانات CGIAR. من بين كبار منتجي الذرة في العالم، من المتوقع أن تعاني الولايات المتحدة والصين من أكبر انخفاضات في الغلة. و على التوالي (الجدول 3). من المتوقع أن تنخفض غلات الأرز في
الشكل 3. متوسط خطأ الجذر التربيعي للنموذج (RMSE) من التحقق المتقاطع بعشر طيات (الأعلى) والانحراف المفسر بواسطة كل نموذج (الأسفل) عبر 5 تقديرات لمجموعات البيانات المدخلة. الدوائر السوداء تظهر متوسط RMSE والانحراف المفسر الخاص بالنموذج، المحسوب عبر مجموعات البيانات المدخلة. الأسطورة: GLM RI، نموذج خطي مختلط عام (GLMM) مع تقاطعات عشوائية؛ GLM RS، GLMM مع تقاطعات ومنحدرات عشوائية؛ GAM RI، نموذج مختلط إضافي عام (GAMM) مع تقاطعات عشوائية؛ GAM RS، GAMM مع تقاطعات ومنحدرات عشوائية؛ LM، نموذج OLS مجمع.
الشرق الأوسط، أمريكا الجنوبية، وجنوب/جنوب شرق آسيا، تتفاقم على مدار القرن مع خسارة متوسطة عالمية مرجحة في العائد قدرها تحت SSP1-2.6 و تحت سيناريو SSP5-8.5 بحلول 2081-2100. من المتوقع أن تنخفض غلات الأرز في الهند وبنغلاديش بحوالي بينما هذه الخسائر أقل في الصين ) وإندونيسيا ( ) (الجدول 3). تواجه عوائد فول الصويا عدم يقين كبير مع توقعات تظهر مناطق من التغير السلبي والإيجابي الصافي عبر مناطق البرازيل ( ) والولايات المتحدة ( )، بينما تظهر العوائد العالمية المتوسطة زيادة صافية من تحت SSP1-2.6 ولكن بانخفاض صافي من تحت SSP5-8.5. يظهر القمح انخفاضًا نسبيًا صغيرًا في العائد العالمي تحت SSP1-2.6 بحلول 2081-2100، ولكن من المتوقع حدوث خسائر كبيرة على مستوى العالم ( ) ، في الهند ( ) والولايات المتحدة ( ) تحت SSP5-8.5 (الجدول 3).

الفجوة العالمية المتوقعة في سعرات الطعام

قمنا بتحويل تقديرات التغير في العائد العالمي المتوقع إلى إمدادات السعرات الحرارية الوطنية وقارناها بتقديرات الطلب الوطني على السعرات الحرارية باستخدام بيانات من منظمة الأغذية والزراعة (الفاو) وأدبيات أخرى. يمثل فجوة السعرات الحرارية الطلب الوطني المستقبلي على السعرات الحرارية الذي لا يمكن تلبيته من خلال الإنتاج والواردات. تظهر نتائجنا انخفاض إمدادات السعرات الحرارية في العديد من البلدان في جنوب وجنوب شرق آسيا، وأفريقيا جنوب الصحراء الكبرى، مع توقعات بخسائر أكثر حدة تحت سيناريو SSP5-8.5 مقارنة بالسعرات الحرارية الأساسية الموردة في عام 2015 (الشكل 5أ، ب، الشكل التكميلية 7أ). بالإضافة إلى هذه البلدان، قد تصبح أمريكا الشمالية وأوقيانوسيا معرضتين لخطر عدم القدرة على تلبية الطلب الوطني على الغذاء تحت سيناريو SSP5-8.5 (الشكل 5ج، د). اتبعنا نهجًا بسيطًا عمدًا لتوضيح الآثار المحتملة لاستجابات العائد المقدرة على الأمن الغذائي العالمي. يعامل هذا النهج العديد من العوامل الديناميكية كعوامل ثابتة، مثل المساحات المحصودة، ومتطلبات الطاقة اليومية المتوسطة، ونسبة إنتاج كل بلد الموجهة إلى كل شريك تصدير، ونسبة المحاصيل غير الأساسية في نظام غذاء البلد. وبالتالي، افترضنا أنماطًا ثابتة لتوزيع المحاصيل، وتكرار الحصاد، وإجمالي مساحة الأراضي الزراعية القائمة، وتركيبات السكان في الدول، وأنماط التصدير وأنماط النظام الغذائي. (انظر قسم “الطرق”).
قمنا أيضًا بإعادة إنتاج خرائط فجوة السعرات الحرارية باستخدام نموذج OLS المجمّع (الشكل التوضيحي 7b، d). يُظهر هذا أن العديد من البلدان تعاني من تخفيضات أقل حدة في السعرات الحرارية تحدث في وقت لاحق من القرن مقارنةً بنموذج GLMM المفضل مع تقاطعات وانحدارات عشوائية. قد يكون ذلك بسبب الدراسات متعددة التقديرات التي تحتوي على عائدات أقل حدة.
الشكل 4. التغيرات المتوقعة في العائد المستمدة من نموذج الانحدار الخطي المختلط العام (GLMM) مع تقاطعات وانحدارات عشوائية مجمعة عبر 5 نماذج تعويض و23 مجموعة بيانات توقعات GCM لنموذج SSP5-8.5، حيث تشير التظليل إلى التوزيع المكاني لإنتاج المحاصيل. .
تؤثر الاستجابات على نتائج OLS المجمعة، أو لأن الاستجابات السلبية للعائدات واضحة ضمن الدراسات الفردية، ولكنها خافتة عند دمج البيانات من تلك الدراسات (أي مفارقة سيمبسون).

نقاش

نقدم مراجعة موجزة لنتائجنا في سياق الأدبيات العلمية الواسعة والعميقة التي تغطي تغير الفينولوجيا لمحاصيل الحبوب استجابةً لارتفاع درجات الحرارة وظروف الجفاف. . تحدد نتائجنا خسائر أكبر في العائدات للذرة والأرز في العديد من المناطق مقارنةً بالتقديرات السابقة باستخدام نفس مجموعة بيانات CGIAR، بحوالي نقاط مئوية مقارنة بـ نتائج القمح قابلة للمقارنة مع المرجع. مع فرق نقطة مئوية في أي اتجاه لمعظم المناطق. نتائج من المرجع. تظهر خسائر العائدات بشكل أكبر قليلاً في وسط آسيا، شمال إفريقيا، أمريكا الجنوبية وغرب أستراليا. نعتقد أن هذه الاختلافات ناتجة عن المعالجات المنهجية للبيانات واحتساب تغيرات هطول الأمطار في تقديرنا لخسائر العائدات العالمية الموزعة على الشبكة.
محصول سيناريو 2021-2040 2041-2060 2061-2080 ٢٠٨١-٢١٠٠
ذرة SSP1-2.6 – 1.6 [- 9.0, 5.7] – 3.3 [- 10.5, 4.0] -4.0 [-11.2, 3.3] -3.8 [- 11.1, 3.6]
SSP2-4.5 -1.6 [-9.0, 5.7] – 4.6 [- 11.9, 2.7] -7.2 [-14.8, 0.4] – 13.4 [- 17.0, – 1.2]
SSP5-8.5 – 2.0 [- 9.3, 5.3] -6.8 [-14.3, 0.8] -12.9 [-22.3, -4.5] -22.2 [-34.3, -10.1]
أرز SSP1-2.6 0.5 [- 9.1, 10.0] – 1.2 [- 10.8, 8.4] – 2.2 [- 12.0, 7.5] – 2.7 [- 12.5, 7.1]
SSP2-4.5 1.0 [- 8.5, 10.5] – 0.9 [- 10.8, 9.0] – 2.7 [- 13.5, 8.0] – 4.3 [- 15.9, 7.2]
SSP5-8.5 1.2 [- 8.4, 10.7] – 1.8 [- 12.5, 8.9] – 5.8 [- 19.2, 7.7] -9.0 [-26.8, 8.8]
صويا SSP1-2.6 2.9 [-21.3، 27.0] 3.3 [-20.5, 27.1] 2.4 [- 21.5, 26.3] 1.4 [- 22.7, 25.5]
SSP2-4.5 3.5 [-20.5, 27.5] 5.3 [- 19.3, 30.0] 4.4 [- 22.4, 31.2] 2.6 [- 26.1, 31.4]
SSP5-8.5 4.5 [- 19.7, 28.6] 5.5 [- 21.4, 32.3] -0.6 [-36.0, 34.8] – 15.4 [- 69.2, 38.5]
قمح SSP1-2.6 1.3 [- 11.0, 13.6] -0.4 [-12.3, 11.5] – 1.3 [- 13.2, 10.6] – 1.5 [- 13.5, 10.5]
SSP2-4.5 1.7 [-10.6، 13.9] -0.7 [-12.6, 11.1] – 3.1 [- 15.1, 9.0] -4.9 [-17.3, 7.5]
SSP5-8.5 1.6 [-10.6, 13.7] – 2.4 [- 14.4, 9.7] – 7.8 [- 21.5, 6.0] – 14.1 [- 32.6, 4.3]
الجدول 2. توقعات التغير في متوسط العائد الموزون عالميًا (%) من نموذج الانحدار الخطي العام المختلط (GLMM) لـ SSP1-2.6 (الأعلى)، SSP2-4.5 (الوسط) و SSP5-8.5 (الأسفل)، التقديرات المركزية بالخط العريض [فترة الثقة 95%]. الوزن بناءً على قيم إنتاج المحاصيل الموزعة على الشبكة. التقديرات تتجاهل التغيرات في المناطق غير المزروعة حاليًا.
محصول أكبر المنتجين SSP1-2.6 SS2-4.5 SSP5-8.5
ذرة الولايات المتحدة -6.5 [-13.3, 0.3] – 12.6 [- 20.5, – 4.6] -26.0 [-39.3، -12.7]
الصين -4.7 [-11.7, 2.2] – 10.1 [- 17.8, – 2.4] -24.7 [-37.2, -12.1]
البرازيل 1.0 [- 7.5, 9.5] – 3.0 [- 10.7, 4.7] – 13.1 [- 22.6, – 3.7]
الأرجنتين – 0.9 [- 8.3, 6.5] – 4.3 [- 11.6, 3.1] – 12.1 [- 20.8, – 3.5]
أرز الصين -5.6 [- 13.0, 1.8] -6.1 [-15.1, 3.0] – 9.8 [- 24.6, 5.0]
الهند -7.2 [-14.1، -0.3] – 10.2 [- 18.6, – 1.7] – 21.3 [- 35.0, – 7.6]
إندونيسيا 7.4 [- 8.2, 23.0] 3.7 [- 11.7, 19.1] -6.5 [-22.5, 9.6]
بنغلاديش – 10.3 [- 19.8, – 0.7] – 11.9 [- 22.6, – 1.2] – 21.6 [- 36.4, – 6.9]
فول الصويا البرازيل 0.4 [- 28.4, 29.1] 5.4 [-20.0، 30.9] -5.6 [-45.9, 34.8]
الولايات المتحدة – 3.5 [- 22.7, 15.8] – 6.0 [- 32.6, 20.7] – 30.9 [- 89.1, 27.3]
الصين 1.4 [-20.7، 23.5] 2.1 [- 25.6, 29.7] – 17.2 [- 73.4, 39.0]
الأرجنتين 2.9 [- 17.5, 23.3] 7.7 [- 14.0, 29.4] 2.9 [- 28.1, 33.8]
قمح الصين -4.0 [-12.9, 4.8] – 5.7 [- 15.3, 3.9] – 12.7 [- 29.2, 3.9]
الهند -4.6 [-16.3, 7.1] – 8.5 [- 20.8, 3.8] – 21.7 [- 39.9, – 3.6]
روسيا – 1.9 [- 14.3, 10.5] -4.9 [-17.6, 7.7] – 13.3 [- 33.1, 6.4]
الولايات المتحدة -4.3 [-14.5, 6.0] -7.2 [-18.1, 3.8] – 14.4 [- 32.5, 3.6]
كندا -2.1 [- 13.0, 8.9] -3.4 [-15.5, 8.8] -8.1 [-31.0, 14.8]
الجدول 3. توقعات التغير في متوسط العائد الموزون بالإنتاج للفترة 2081-2100 من نموذج الانحدار الخطي المختلط العام (GLMM) مع تقاطعات وانحدارات عشوائية لأكبر المنتجين (%) لسيناريوهات SSP1-2.6 وSSP2-4.5 وSSP5-8.5، التقديرات المركزية بالخط العريض [فترة الثقة 95%]. الوزن بقيم إنتاج المحاصيل الموزعة على الشبكة يتجاهل التغيرات في المناطق غير المزروعة حاليًا. انظر البيانات التكميلية لجميع البلدان.
مقارنةً بالتحليلات الميتا السابقة لنفس مجموعة بيانات CGIAR إن تضميننا لتغير هطول الأمطار وتفاعله مع تغير درجات الحرارة ومستويات هطول الأمطار الأساسية أدى إلى خسائر أكبر في العائدات العالمية للذرة والأرز. وهذا يوضح أهمية تضمين تفاعلات درجات الحرارة وهطول الأمطار في نمذجة تأثيرات المناخ على الزراعة. غالبًا ما تتزامن درجات الحرارة العالية مع، أو تكون ناتجة عن، انخفاض هطول الأمطار، ويمكن أن يخلق تفاعلها تأثيرات لا يمكن التقاطها من خلال النظر إلى درجات الحرارة أو هطول الأمطار بشكل منفصل. وقد وجدت المراجعات الأخيرة لتفاعلات الحرارة المجمعة والجفاف تأثيرات أكبر على عائدات المحاصيل الحبوب مقارنة بالضغوط الفردية وحدها. بالإضافة إلى التفاعل مع توفر المغذيات تسبب ارتفاع درجات الحرارة المتوسطة في تبخر المزيد من الماء من سطح التربة (خسائر التبخر والنتح) وبالتالي زيادة متطلبات المياه للمحاصيل. الهطول أكثر تباينًا مكانيًا من درجة الحرارة، وهناك اختلاف كبير بين نماذج المناخ العالمية بشأن اتجاه تغير الهطول في العديد من المناطق. . هذا يبرز أهمية تضمين نطاق التباين في توقعات هطول الأمطار المستقبلية كما فعلنا هنا من خلال استخدام بيانات الإدخال من 23 نموذجًا من نماذج المناخ العالمية.
تتوافق الخسائر النسبية الأكثر حدة في العائدات للذرة مقارنة بالقمح مع تحليلات مقارنة أخرى في الأدبيات. كما وجدوا أن الذرة أكثر حساسية للجفاف مقارنة بالقمح، جزئيًا بسبب طول فترة الفتحة والتلقيح (ASI) مما يؤدي إلى عدم التزامن بين الفتحة والتلقيح، مما يؤثر على التلقيح والتكاثر. وقد افترضوا أن الاختلافات في حساسية الجفاف يمكن أن تُعزى إلى
الشكل 5. الآثار المترتبة على الأمن الغذائي المقدرة من نموذج الانحدار الخطي العام المختلط (GLMM) مع تقاطعات وانحدارات عشوائية. الأعلى: انخفاض في إجمالي إمدادات الغذاء بالسعرات الحرارية بالنسبة لخط الأساس لعام 2015 (%) لـ SSP24.5 (أ) و SSP5-8.5 (ب). الأسفل: تغيير في خطر عدم تلبية الطلب الوطني على السعرات الحرارية من الإنتاج والواردات لـ SSP2-4.5 (ج) و SSP5-8.5 (د). الخرائط لـ SSP1-2.6 موجودة في الشكل التكميلية 7.
أصول الذرة في مناطق أقل محدودية بالمياه مقارنة بالقمح الذي نشأ في المناطق الجافة. وبالتالي، اكتسب القمح صفات التكيف بما في ذلك تأخير الشيخوخة، والتي يمكن أن تكون مفيدة في الجفاف المتأخر في الموسم، بالإضافة إلى أنظمة جذرية أعمق للوصول إلى مياه التربة تحت السطحية. . من ناحية أخرى، يحتوي القمح على حدود درجات حرارة دنيا أقل ( ) من الذرة ( ) أو فول الصويا ( 46,47، لذا نتوقع انخفاضات أو خسائر حيث يتم تجاوز هذه العتبات حيث يتم زراعتها حاليًا (انظر قسم “الطرق”). ومع ذلك، قد تصبح العديد من المناطق التي تكون حاليًا باردة جدًا للقمح قابلة للزراعة حتى تصل إلى النقطة المثلى لدرجات الحرارة لملء الحبوب. حددت دراسة تحليلية سابقة لاستجابة الذرة والقمح للجفاف أن العوائد تكون أكثر حساسية للجفاف إذا كانت المياه محدودة خلال المرحلة التناسلية مقارنةً بمرحلة النمو الخضري للنبات. من المتوقع أن تؤثر درجات حرارة النهار والليل التي تتحرك خارج نطاق درجات الحرارة المثلى لتحقيق أقصى إنتاجية للحبوب بشكل مباشر على إنتاجية الحبوب (الجدول التكميلي 1). . ومع ذلك، لا يمكن التقاط الفروق التفصيلية في تغير درجة الحرارة وهطول الأمطار طوال موسم النمو من خلال دوال الاستجابة لدينا هنا. من بين جميع المحاصيل الأساسية، فإن الذرة لديها أكبر مساحة مزروعة حالياً. إذا تم زراعة الذرة في مناطق تكون فيها درجات حرارة موسم النمو أقل ولكن قريبة من درجات الحرارة المثلى للتطوير والإنتاج، فقد تؤدي الزيادة في درجات الحرارة إلى دفع هذه المواقع خارج النطاق المثالي لدرجات حرارة النمو.
خسائر محصول الأرز أكبر مما تم تقديره في التحليلات الميتا السابقة باستخدام نفس مجموعة بيانات CGIAR، ومن المتوقع هنا أن تنخفض الغلات في أفريقيا وأمريكا الجنوبية وأستراليا حتى نهاية القرن استجابةً لارتفاع درجات الحرارة. وهذا يتماشى مع الفهم العلمي لتأثير إجهاد درجات الحرارة العالية على حبوب لقاح الأرز، والخصوبة، وغلة الحبوب. مرجع تقديرات انخفاض العائد عند درجة واحدة من الاحترار العالمي للصين روسيا )، فرنسا ( )، كانساس في الولايات المتحدة ( ، الهند ( )، مقارنة بتوقعاتنا في 2041-2060 و 2061-2080 التي تظهر انخفاض غلة القمح بـ في الصين، في الهند، في الولايات المتحدة، في روسيا ومرتفعًا بـ في فرنسا. مثل نرى أيضًا زيادات في إنتاج الأرز في مناطق الارتفاع العالي والعرض (مثل هضبة التبت، غرينلاند، سيبيريا، الأنديز). نتائج فول الصويا لدينا تختلف بشكل حاد عن تلك المقدرة سابقًا باستخدام نفس مجموعة بيانات CGIAR، حيث مشروع الفشل التام للمحاصيل في أفريقيا والبرازيل وجنوب أستراليا. وقد ناقشت دراسات أخرى درجة كبيرة من عدم اليقين المحيط باستجابة محصول الصويا لتغير المناخ، والتي قد تكون ناتجة عن ارتفاع درجة حرارته الحرجة.
تحمل الإجهاد الحراري حول الإزهار (مقارنة بالذرة والقمح) ولكن أيضًا تحملها لدرجات الحرارة المنخفضة. يمكن رؤية ذلك في دراسات مماثلة تتنبأ بوجود مناطق عديدة ذات غلات منخفضة كما هو الحال مع المناطق ذات الغلات المرتفعة، لا سيما في الولايات المتحدة والبرازيل. وجدنا أيضًا نطاقًا واسعًا من عدم اليقين حول توقعات فول الصويا عبر جميع أبعاد اختيار النموذج، والعينات، ونماذج المناخ العالمية، والبيانات المفقودة.
تظهر خرائطنا للتغيرات المتوقعة في غلة المحاصيل العالمية الموزعة على الشبكة زيادة في ملاءمة زراعة الأرز والقمح في خطوط العرض والارتفاعات الشمالية، وانخفاض في الملاءمة تقريبًا في كل مكان آخر (انظر الشكل 4). وهذا يتماشى مع دراسات أخرى تظهر زيادة الملاءمة في شمال أوروبا لمحاصيل القمح المعتمدة على الأمطار بسبب التغيرات في هطول الأمطار ودرجة الحرارة. لقد لوحظ أيضًا أن تقليل حدوث الصقيع يفيد في زيادة إنتاج المحاصيل. من منظور التكيف مع المناخ، وتوفير المياه، وتلوث المغذيات، قد يحتاج سلة الإنتاج على المستوى الإقليمي وتوزيع المحاصيل الأمثل إلى إعادة نظر كبيرة. . تفترض طريقتنا أن المحاصيل تتغذى بمياه الأمطار، ومع ذلك تقريبًا من الأراضي المزروعة و يحدث جزء من إنتاج المحاصيل العالمية تحت ظروف الري في المستقبل، تعني قيود المياه العذبة أن الأراضي الزراعية المروية في بعض المناطق قد تحتاج إلى العودة إلى إدارة تعتمد على الأمطار. بينما تتوسع في مجالات أخرى تلعب الريّة دورًا مهمًا في دعم إنتاج المحاصيل في المناطق الجافة، على الرغم من أن التغيرات المدفوعة بالطقس في سحب المياه للري على مستوى العالم كانت متواضعة. حتى الآن لقد توقعت دراسات متنوعة سيناريوهات للضغط المائي المستقبلي والطلب المتزايد من الصناعات المتنافسة وتغير استخدام الأراضي، مما من المحتمل أن يقيد توسيع الري ويقلل من موثوقية موارد المياه الزرقاء. استجابة المناخ للمزارع المعتمدة على الأمطار مقابل المزارع المروية ليست مفهومة جيدًا على نطاق عالمي وتستحق مزيدًا من الدراسة، خاصة في سياق التكيف مع تغير المناخ والتخفيف من الخسائر. . يجب أيضًا دمج توفر المياه الزرقاء للمحاصيل المروية في الأعمال المستقبلية حول هذا الموضوع .
تسلط نتائجنا الضوء على تغير توزيع المحاصيل بسبب تغير أنماط درجات الحرارة وهطول الأمطار، ومع ذلك فإنها تغفل آثار ارتفاع مستوى سطح البحر، والملوحة، وضغط المياه الناتج عن استخدامات الأراضي المتنافسة . في تحليلاتهم العالمية لأضرار تغير المناخ، يقدر المرجع النسبة المئوية لفقدان الأراضي الساحلية القابلة للزراعة لكل متر من ارتفاع مستوى سطح البحر لـ 140 منطقة، حيث وجد أن الآثار الفيزيائية لارتفاع مستوى سطح البحر تتركز بين الدول الجزرية الصغيرة في أوقيانوسيا والمحيط الهادئ. قد يؤدي ذلك إلى زيادة ملوحة التربة القابلة للزراعة في المناطق الداخلية والساحلية، وهو ضغط غير حيوي كبير على نباتات المحاصيل – حاليًا أكبر من من الأراضي الزراعية العالمية تتأثر سلبًا بضغط الملح .
تقديراتنا، التي تأخذ في الاعتبار التغيرات في المعايير متعددة العقود، لا تلتقط أيضًا بالكامل آثار درجات الحرارة المتغيرة والمتطرفة وأحداث هطول الأمطار. هناك أدلة تشير إلى أن المحاصيل قد تكون أكثر حساسية لمناخ متغير بشكل كبير على كوكب أكثر دفئًا في المتوسط مقارنة بمناخ يتغير تدريجيًا . في جنوب آسيا، يعد هطول الأمطار في موسم الرياح الموسمية الصيفية مسؤولًا عن ما يقرب من من التباين في شذوذ إنتاج الحبوب الغذائية الإجمالية في المنطقة . تم استخدام مولدات الطقس العشوائية أيضًا لإظهار أن زيادة التباين في درجات الحرارة أو هطول الأمطار في المملكة المتحدة وفرنسا كان لها أحيانًا تأثير أكبر على غلات القمح من التغيرات في المتوسطات، خاصة عند دمجها مع درجات حرارة موسمية متوسطة أكثر دفئًا .
لقد وجدنا أن تخفيضات في إمدادات السعرات الحرارية تتراوح بين 15 و ممكنة في الفترة من 2081 إلى 2100 تحت أسوأ سيناريو انبعاثات SSP5-8.5 (الشكل 5). العديد من الدول في أفريقيا جنوب الصحراء الكبرى، وجنوب آسيا، وأمريكا الشمالية، وأوقيانوسيا معرضة بشكل متزايد لخطر عدم القدرة على تلبية الطلب المحلي على الغذاء من خلال الإنتاج والاستيراد. وبالتالي، يجب النظر في تدابير التكيف بما في ذلك التكثيف المستدام، وتوسيع مناطق الزراعة، وتغيير أنماط التجارة أو التخفيف من الانبعاثات. في الواقع، تظهر نتائجنا لـ SSP1-2.6 و SSP2-4.5 التي تمثل سيناريوهات انبعاثات أقل (الشكل التكميلي 7) عجزًا أقل بكثير في السعرات الحرارية للعديد من الدول. من المهم أن نكرر أننا لم نفترض أي اتجاهات في تكرار الحصاد أو إجمالي مساحة الأراضي الزراعية القائمة، على الرغم من أن مساحة الأراضي المحصودة عالميًا نمت تقريبًا أربع مرات أسرع من إجمالي مساحة الأراضي الزراعية القائمة بين عامي 2000 و . في الواقع، نتوقع أن يستمر تكرار الحصاد وإجمالي مساحة الأراضي الزراعية القائمة في التغير، على الرغم من أنه ليس في نفس الاتجاه عبر جميع المناطق (على سبيل المثال، زاد تكرار حصاد المحاصيل من 1961 إلى 2011 في البرازيل والهند والصين ولكنه انخفض في العديد من الأماكن في أفريقيا ). قد تعوض توسيع المناطق المناسبة للزراعة بعض خسائر الغلة التي تم نمذجتها هنا. ومع ذلك، نعتقد أن تقديراتنا لخسائر الغلة والأضرار تعتبر محافظة حيث لم نأخذ في الاعتبار الآثار غير المباشرة لتغير المناخ على الأمن الغذائي مثل من خلال هطول الأمطار الشديد، والفيضانات، والحرائق، والآفات والأمراض.
من المحتمل أن تتأثر جودة المغذيات بتغير المناخ، ولكن لم يكن من الممكن تضمينها في هذا التحليل. تظهر الأبحاث آثارًا مختلطة لزيادة CO2، ودرجات الحرارة، والملوحة، والفيضانات، وضغط الجفاف على تراكم المغذيات في التربة، حيث تعتمد الآثار بشكل كبير على المغذيات، ونبات المحاصيل، ومستوى شدة التغيير . للمضي قدمًا، تدعم إدارة المغذيات والري المستدامة قدرة المزارعين على سد فجوات الغلة مع الحفاظ على جودة المغذيات أو زيادتها . عامل آخر يؤثر على الأمن الغذائي في المستقبل هو خسائر المحاصيل من الآفات والأمراض. يؤثر تغير المناخ بالفعل على نطاقات التوزيع، وحالة الآفات، وتزامن الآفات والأمراض الاقتصادية الهامة مع النباتات المضيفة . يجب دمج هذه الآثار في نمذجة استجابة الغلة المستقبلية.
أخيرًا، قمنا بالتحقيق في مصادر عدم اليقين التي تساهم في التباين حول استجابات الغلة العالمية المقدرة. وجدنا أن عدم اليقين الناتج عن اختيار النموذج هو الأكثر أهمية مقارنة بعدم اليقين الناتج عن العينة، والبيانات المفقودة، ونماذج المناخ العالمية. فيما يتعلق بعدم اليقين الناتج عن اختيار النموذج، قد لا تكون توقعات نماذج التجميع هي أفضل طريقة لتمثيل هذا التباين حيث أظهرت بعض النماذج أنها تتفوق على الأخرى. لقد اخترنا مواصفة نموذج التأثيرات المختلطة التي نعتقد أنها تحقق أفضل توازن بين الأداء التنبؤي والتكيف المفرط مع البيانات. نشعر بالراحة لرؤية توافق كامل للنموذج عبر العديد من المناطق في متوسط علامة توقعات الغلة المستقبلية لجميع المحاصيل باستثناء فول الصويا، على الرغم من أن هذه المناطق من التوافق لا تتداخل بالضرورة مع الأماكن التي تزرع فيها المحاصيل حاليًا (الشكل التكميلي 9). يعكس عدم اليقين الناتج عن نماذج المناخ العالمية عدم اليقين بشأن تأثير انبعاثات البشر على نظام المناخ. بينما يمكن توضيح هذا عدم اليقين العميق من خلال تضمين مجموعة من نماذج CMIP6 كمدخلات لنمذجة تأثير المناخ، قد لا تمثل مجموعات النماذج العينة الكاملة أو النظامية لاستجابة المناخ . أظهرت دراسات أخرى أنه بالنسبة لتوقعات استجابة المناخ المعتمدة على نماذج المناخ العالمية، يميل عدم اليقين الناتج عن نماذج المناخ العالمية والتباين الداخلي لنظام المناخ إلى الهيمنة على مصادر عدم اليقين الأخرى على المدى القريب، بينما في سيناريو أواخر القرن، يهيمن عدم اليقين
(على الرغم من أن عدم اليقين الناتج عن نماذج المناخ العالمية لا يزال مهمًا) . في تحليلنا لتوقعات أواخر القرن لسيناريو واحد، يهيمن عدم اليقين الناتج عن نماذج المناخ العالمية على عدم اليقين الناتج عن النموذج وعدم اليقين الناتج عن العينة. بالنسبة للتحليلات الميتا المستقبلية، نوصي بتوسيع مجموعة الدراسات وحجم العينة حيثما أمكن، أو اتخاذ نهج إعادة التقدير لوصف عدم اليقين الناتج عن استراتيجية أخذ العينات من البيانات.
تظل تحذيرات كبيرة بشأن النهج الذي اتخذناه في النمذجة الإحصائية هنا. ربما الأهم من ذلك، أن نقل الوظيفة يفترض ضمنيًا أن معلمات الوظيفة المقدرة في مواقع الدراسة تنتقل عبر الزمن والمكان . نظرًا لأن مجموعة فرعية فقط من جميع البلدان ممثلة في بيانات CGIAR (انظر الشكل التكميلي 1)، فإن هذا الافتراض قد لا يكون صحيحًا. على سبيل المثال، تحتوي أستراليا على تربة فقيرة بالمغذيات بشكل استثنائي نتيجة لمرورها بفترة أطول بكثير من التجوية والجفاف مقارنة بالقارات الأخرى، ويحدث جزء كبير من إنتاج المحاصيل في ظل ظروف شبه جافة وجفاف. تتخلف غلات المحاصيل الرئيسية في أستراليا من حيث متوسط الغلات المروية منذ الستينيات بشكل كبير عن بقية العالم باستثناء الغلات في أفريقيا . مقارنة بالمناطق ذات هطول الأمطار الأساسي الأعلى، تظهر المحاصيل المروية في أستراليا عمومًا حساسية أعلى للغلة تجاه هطول الأمطار، مما يعني أن زيادة صغيرة في هطول الأمطار تؤدي إلى زيادات نسبية أكبر في الغلة. وهذا يشير إلى أن نقل المعلمات العالمية لاستجابة الغلة لتغير هطول الأمطار قد لا يتنبأ بالاستجابات المحلية بشكل جيد، على الرغم من أن نماذجنا تأخذ في الاعتبار التفاعلات بين مستويات الأساس والتغيرات في هطول الأمطار، واستخدمنا تقديرات تأثيرات الدول العشوائية للتنبؤ بالتغيرات المحلية في الغلة. لم يتم التقاط عوامل مثل توفر المغذيات وتفاعلاتها مع العوامل المناخية هنا كما ذُكر سابقًا. وبالتالي، يتطلب نقل معلمات النموذج إلى إعدادات تكون فيها نقص المغذيات حادًا عناية إضافية. نؤكد أن استخدام نماذج التأثيرات المختلطة هو الأفضل مقارنة بنماذج OLS، حيث يساعد تضمين تأثيرات الدول العشوائية في معايرة تقدير الاستجابات لتلك الدول. علاوة على ذلك، نظهر أن استخدام نماذج التأثيرات المختلطة يؤدي إلى تقديرات تظهر تخفيضات أكثر حدة في السعرات الحرارية ونطاقات أكبر من عدم اليقين مقارنة بنموذج OLS الأصلي. نرى أن اختيار النموذج المناسب لملاءمة البيانات له آثار مهمة على تقدير استجابات المناخ على الإنتاجية الزراعية. لم نولي الكثير من الاهتمام لكيفية مقارنة هذه النتائج بنماذج عمليات المحاصيل، ومع ذلك، وجدت مراجعة حديثة لنماذج العمليات والنماذج الإحصائية أن الطريقتين تولدان حساسية غلة المحاصيل مشابهة لتغيرات المناخ علاوة على ذلك، على الرغم من أن التنبؤات الفردية غالبًا ما تنحرف بشكل كبير، إلا أن المتوسط المجمّع لنماذج العمليات يميل إلى الارتباط جيدًا مع تنبؤات النماذج الإحصائية. .

الخاتمة

في هذه الدراسة، استخدمنا التحليل الإحصائي الشامل لتناسب نماذج التأثيرات المختلطة لاستجابة إنتاج المحاصيل لتغير المناخ باستخدام قاعدة بيانات معتمدة. قمنا بتناسب خمسة مواصفات نموذجية وقدمنا تقديرات لتغيرات الإنتاج المستقبلية باستخدام توقعات درجات الحرارة وهطول الأمطار حتى عام 2100 من مجموعة CMIP6 المكونة من 23 نموذجًا من نماذج المناخ العالمية. استخدمنا هذه التقديرات لحساب إمدادات السعرات الحرارية الوطنية وقارناها بمقاييس الطلب على السعرات الحرارية لفهم آثارها على الأمن الغذائي العالمي. يقدر نموذجنا المفضل أن الإنتاج العالمي قد يتغير عن مستويات عام 2015 بـ (الذرة) (الأرز)، (فول الصويا) و (قمح) في تحت سيناريو الأعمال كالمعتاد SSP5-8.5، ولكن يتم تقليله إلى و على التوالي تحت سيناريو الانبعاثات المنخفضة SSP1-2.6. نحن نرى أن هذا يظهر فوائد كبيرة من تخفيف الانبعاثات. قمنا بتحديد عدم اليقين عبر خمسة أبعاد ووجدنا أنه على الرغم من النطاق الواسع للتغيرات المتوقعة في غلة فول الصويا، فإن نتائجنا تقدم تأكيدًا عامًا على أنماط التغير في توزيع محاصيل الأرز والقمح تحت تغير المناخ. بالنسبة لبعض أنواع المحاصيل والمناطق، مثل الدول المنتجة لفول الصويا في الولايات المتحدة والصين والبرازيل والأرجنتين، فإن عدم اليقين بشأن التغيرات المستقبلية في الغلة كبير جدًا لدرجة أن صانعي القرار يجب أن يتبنوا مبدأ الحذر للاستثمار في التكيف مع تغير المناخ في القطاع الزراعي. في غياب التحقق المحلي، نعتقد أن هذه التقديرات توفر توقعًا عامًا لتغيرات غلة المحاصيل الأساسية المستقبلية تحت تغير المناخ. في ظل عدم اليقين الكبير، فإن الإشارة واضحة بما فيه الكفاية حول الحاجة إلى سياسات وطنية ودولية للتanticipate مخاطر الأمن الغذائي الوطني. مع توفر بيانات تجريبية عن الغلات، وتغير تواريخ المحاصيل وأنظمة الزراعة بدقة أعلى وتكرارات زمنية، من الضروري التحقق من تقديرات التأثير وتحديث توقعات نماذج التقييم المتكامل المستخدمة في صنع السياسات. ستساعد التقييمات المنهجية لعدم اليقين مثل هذه صانعي السياسات والباحثين في مجال الأمن الغذائي على فهم العوامل المؤثرة في التباين والتحيزات في تمارين النمذجة الكبيرة المماثلة.

طرق

الأهداف والنطاق

قمنا بإجراء تحليل ميتا إحصائي لفهم التأثيرات المحتملة للمناخ على غلات المحاصيل وبالتالي الحصول على تقديرات للعجز المحتمل في السعرات الحرارية تحت ثلاثة سيناريوهات انبعاثات حتى نهاية القرن. للقيام بذلك، قمنا بتطبيق خمسة نماذج مختلفة بما في ذلك نماذج التأثيرات المختلطة لأخذ الارتباطات داخل وبين الدراسات في الاعتبار، وقمنا بتحديد الحجم النسبي لعدم اليقين الناتج عن اختيار النموذج، والعينة، والبيانات المفقودة، ونماذج المناخ العالمية من مجموعة CMIP6. قمنا بتحويل التغير المتوقع في الغلة الذي تم تقديره بواسطة النموذج المختلط المفضل إلى مقاييس للإمداد الوطني من السعرات الحرارية وقارناها بمقاييس الطلب على السعرات الحرارية لفهم التأثيرات المستقبلية على الأمن الغذائي.

البيانات، الاستخراج والفحص

تشمل بيانات CGIAR 13,890 تقديرًا نقطيًا لتغير العائد، وتغير درجة الحرارة، التغيير، التكيف في المزرعة، والمتغيرات الأخرى المستخرجة من 89 دراسة ذات منهجيات متنوعة يمكن تصنيفها بشكل عام إلى نماذج الانحدار الإحصائي أو نماذج العمليات البيوفيزيائية للمحاصيل. تظهر مراجعات نقاط القوة والضعف في المنهجين المختلفين أن النتائج قابلة للمقارنة عند تجميعها. . قمنا أولاً بمراجعة والتحقق من صحة كل من 8846 تقديرًا من 74 دراسة تتعلق بالمحاصيل الأربعة الرئيسية: الذرة، الأرز، فول الصويا، والقمح من خلال مراجعة الدراسات الأصلية (القائمة الكاملة في المعلومات التكميلية). أنشأنا قاعدة بيانات للتحقق والتنظيف لتسجيل القيم التي تم تحديدها من خلال عملية المراجعة هذه على أنها مفقودة أو
تم إدخالها بشكل خاطئ، واستبدلناها بالقيم الصحيحة من الدراسات. كما تحققنا من أن أي قيم مفقودة مسجلة في مجموعة بيانات CGIAR كانت قيمًا مفقودة حقيقية (أي ليست قيم صفرية).

درجات الحرارة وهطول الأمطار خلال موسم النمو الأساسي

لتقدير درجات حرارة موسم النمو الأساسية المحددة بالموقع، اعتمدنا على البيانات الشهرية العالمية بيانات درجة الحرارة الموزعة على الشبكة من وحدة الأبحاث المناخية (CRU TS v4.05) من عام 1950 إلى . ثم قمنا بتطبيق معلومات تقويم المحاصيل الموزعة حسب الدرجات حول تواريخ الزراعة والحصاد في عام 2000 لأربعة محاصيل من مجموعتين بيانات تكملان بعضهما البعض حوالي عام 2000 حول تواريخ زراعة المحاصيل العالمية والمساحات المزروعة بالمياه المروية والمطرية على مستوى العالم شهريًا تم حساب درجة حرارة موسم النمو الأساسية للمحاصيل لكل موقع من مواقع الشبكة عن طريق تحديد أشهر موسم النمو لكل محصول أولاً، ثم حساب متوسط درجات حرارة موسم النمو التاريخية السنوية لكل عام من 1950 إلى 2015، ثم حساب متوسط درجات حرارة موسم النمو عبر جميع السنوات في الفترة الأساسية المحددة لكل تقدير من CGIAR (مجموعة فرعية من فترة 1950-2015). أخيرًا، تم تجميع درجات حرارة موسم النمو الأساسية للمحاصيل إلى مستوى الدولة، من خلال أخذ متوسط مرجح لإنتاج المحاصيل لقيم الشبكة في كل دولة. للقيام بذلك، استخدمنا مجموعة بيانات مصفوفة بدرجة حول إنتاج المحاصيل تقريبًا .

تقدير البيانات

من بين 8846 تقديرًا نقطيًا في بيانات CGIAR، كانت هناك العديد من القيم للمتغيرات الرئيسية التي كانت مهمة نظريًا لنمذجة استجابة محصول المحاصيل لتغير درجة الحرارة مفقودة. وشملت هذه التغير في درجة الحرارة، وتغير هطول الأمطار، التغيير، تغيير العائد، وسنوات البداية والنهاية لفترة الأساس. نفترض أن الغياب بين المتغيرات الستة الرئيسية يجب أن يكون في الغالب غيابًا عشوائيًا، مع وجود بعض الاستثناءات المحتملة. حذف الملاحظات التي تحتوي على قيم مفقودة هو خيار واحد، ولكن تركيب نموذج على الحالات الكاملة فقط قد يؤدي إلى تحيز النتائج. لذلك، قمنا بتنفيذ إجراء تعويض رسمي باستخدام التعويض المتعدد بواسطة المعادلات المتسلسلة (MICE) لملء القيم المفقودة لخمس متغيرات: متوسط درجة حرارة موسم النمو الأساسي وهطول الأمطار، وتغير درجة الحرارة، وتغير العائد، وتغير هطول الأمطار و تعتبر MICE البيانات المفقودة كمصدر صريح للتباين يتم التوسط فيه من خلال محاكاة سلسلة ماركوف مونت كارلو (MCMC) استنادًا إلى مواصفات شرطية كاملة لتوزيعات الاحتمالات. اعتمدنا نهج الإحلال المتعدد متعدد المستويات الذي يأخذ في الاعتبار الهيكل العنقودي لبياناتنا. حددنا عدد مجموعات البيانات المملوءة بخمسة، و30 تكرارًا ليتم تشغيلها بالتوازي، استنادًا إلى الفحص البصري لأنماط التقارب. وجدنا أن المطابقة التنبؤية ذات المستويين لمحاكاة القيم المرصودة كانت الأفضل (الشكل التوضيحي 2).

توافق النموذج

قمنا بتناسب 5 نماذج مرشحة لكل محصول على حدة مع تغيير نسبة العائد كمتغير استجابة: نموذج OLS، ونموذجان GLMM، ونموذجان GAMM، كل منهما تم تناسبه مع تقاطعات عشوائية فقط (RI) ومع تقاطعات ومنحدرات عشوائية (RS). . تم تقدير هذه باستخدام الاحتمالية القصوى المقيدة (REML) باستخدام حزمة mgcv في . كما اختبرنا نماذج قامت بوزن التقديرات بناءً على عكس التباين في كل دراسة، كما يُوصى أحيانًا في التحليل التلوي الرسمي. ومع ذلك، يمكن أن يشمل التباين على مستوى الدراسة تباينًا ذا دلالة عبر تجارب منفصلة (أي مجموعات فريدة من التغيرات في درجة الحرارة، وهطول الأمطار و )، وكان من غير العملي فصل التركيبات الفريدة لجميع المتغيرات الثلاثة المستمرة إلى فئات منفصلة.
أكثر نماذج التحديد تعقيدًا (التي تشمل تقاطعات وانحدارات عشوائية) تسمح بتغير درجة الحرارة، وتغير هطول الأمطار و تتغير بشكل عشوائي لكل دراسة. تم كتابة مواصفات النموذج للمتغير الاستجابي مع دالة ارتباط الهوية لـ GLMM مع تقاطعات وانحدارات عشوائية في المعادلة (2) (المقدرة باستخدام مصطلحات متعددة الحدود المتعامدة) و لـ GAMM مع تقاطعات وانحدارات عشوائية في المعادلة (3). يُفترض أن دالة التوزيع طبيعية بناءً على فحص المتبقيات الملائمة في الشكل التوضيحي 10. تم تضمين مواصفة نموذج OLS المجمعة في المعادلة (4)، وتختلف عن الأدبيات السابقة. بطريقتين: يتم تقدير دوال الاستجابة بشكل منفصل لكل من المحاصيل الأربعة، ولا تشمل مصطلح التفاعل بين مصطلح تغيير درجة الحرارة التربيعي ودرجة الحرارة الأساسية.
المعلمات المضمنة: هو التغير النسبي في العائد من تقدير النقطة في البلاد ودرس ك. و هي التغيرات في درجة الحرارة (درجات) تركيز (جزء في المليون)، وهطول الأمطار (%) لتقدير النقطة و هي متوسط درجة حرارة موسم النمو وهطول الأمطار للبلد خلال فترة الأساس المحددة لتقدير النقطة و هي متغيرات وهمية تشير إلى مسار التمثيل الضوئي للنبات. هي متغير وهمي يشير إلى ما إذا كانت تقديرات النقاط تشمل التكيف في المزرعة (محددًا بالصفر في جميع التوقعات). نظرًا لأن دوال الاستجابة تُقدّر بشكل منفصل لكل محصول، يساوي 1 و يساوي 0 للأرز وفول الصويا والقمح، بينما يساوي 1 و يساوي 0 للذرة. بعد ذلك نفترض دالة مقعرة لأخذ في الاعتبار الأثر الإيجابي الهامشي المتناقص لـ من الشكل أين لـ محاصيل (مثل الأرز، فول الصويا، القمح) لـ محاصيل (مثل الذرة). الشروط : في مصطلحات الميل العشوائي تمثل قيم المتوسط للدراسة. المصطلح هو التقاطع العشوائي للبلد، هو التقاطع العشوائي للدراسة، هو ميل درجة الحرارة العشوائي المتغير حسب البلد، هل يتغير منحدر هطول الأمطار العشوائي حسب البلد، هو العشوائي ميل يختلف حسب البلد، و هو الخطأ المتبقي. المعلمات و في المعادلة (3) هي دوال سلسة لـ و على التوالي. كل دالة سلسة هي دالة انسيابية وهي مجموع دوال أساسية أصغر مع معامل تم تقييمه عند قيم .
قمنا بتشغيل هذه النماذج الخمسة المختلفة على كل من مجموعات البيانات الخمسة المعوضة، مما أسفر عن 25 مجموعة من التقديرات الملائمة لكل محصول. ثم قمنا بتطبيقها على 23 مجموعة من بيانات التنبؤ التي تحتوي على جميع مصطلحات النموذج، بما في ذلك درجة الحرارة، والهطول و تغيير المتغيرات المحسوبة من مجموعة CMIP6 المكونة من 23 نموذج مناخي عالمي مقدمة من WorldClim لأربعة فترات زمنية مستقبلية: 2021-2040، 2041-2060، 2061-2080، 2081-2100 .

تنبؤ النموذج

قمنا بإنشاء بيانات التنبؤ لكل من مصطلحات النموذج. تم تقدير درجات حرارة موسم النمو الأساسية وهطول الأمطار باستخدام نفس النهج المستخدم لحساب المتغير في مجموعة البيانات الرئيسية، ولكن لعام 2015 (الشكل التوضيحي 4). تم استرجاع قيم درجات الحرارة وهطول الأمطار الموزعة على الشبكة للفترات 2021-2040، 2041-2060، 2061-2080 و2081-2100 من 23 مجموعة بيانات GCM المصححة من الانحياز من WorldClim لنماذج SSP1-2.6، SSP2-4.5 وSSP5-8.5 عند الدقة المكانية قمنا بتقدير العوائد المتوقعة على كل من مجموعات بيانات GCM الـ 23 بشكل فردي. كانت الـ 23 GCM من مجموعة CMIP6 هي: ACCESSCM2، ACCESS-ESM1-5، CanESM5، CanESM5-CanOE، CMCC-ESM2، CNRM-CM6-1، CNRM-CM6-1-HR، CNRM-ESM2-1، EC-Earth3-Veg، EC-Earth3-Veg-LR، FIO-ESM-2-0، GISS-E2-1-G، GISS-E2-1-H، HadGEM3-GC31-LL، INM-CM4-8، INM-CM5-0، IPSL-CM6ALR، MIROC-ES2L، MIROC6، MPI-ESM1-2-HR، MPI-ESM1-2-LR، MRI-ESM2-0 و UKESM1-0-LL. تم استخدام نفس الـ 23 GCM لسيناريو SSP1-2.6 ولكن مع إضافة GFDL-ESM4. استخدمنا المتغيرات البيوكليمية لدرجة الحرارة المتوسطة السنوية (BIO1) وهطول الأمطار السنوي (BIO12) المحولة إلى متوسط هطول الأمطار الشهري.
للعالمية التغيير، استخدمنا بيانات عن العالمية السنوية تركيزات محسوبة من متوسط مجموعة النماذج المتعددة CMIP5 المكونة من 39 نموذجًا من نماذج المناخ العامة المسترجعة من مستكشف المناخ ، تم تحويله إلى تغييرات في التركيزات من مستويات عام 2015. التاريخية تم استنتاج المستويات من الملاحظات العالمية .
توقعنا تأثيرات العائدات العالمية الموزعة عبر كل من الفترات الزمنية الأربع باستخدام 5 تقديرات ملائمة مستنتجة. مواصفات النموذج المحاصيل (أي 100 مجموعة من النماذج)، وعلى كل من مخرجات نماذج المناخ العالمية الـ 23 (أي 575 مجموعة من التنبؤات لكل محصول، أو 2300 مجموعة من التنبؤات إجمالاً). للحصول على النتائج المجمعة في الشكل 4، والأشكال التكميلية 5 و6 و9، قمنا بتجميع تقديرات ملاءمة التنبؤ والخطأ المعياري بشكل متسلسل عبر كل تقديرات النماذج الخاصة بالتحليل لكل محصول وتقديرات النماذج الخاصة بنماذج المناخ العالمية. وبالتالي، اتبعنا نهج “تجميع الأخير” الذي يطبق قواعد روبين على التقدير النهائي المعني، حيث أن القيام بذلك يحافظ على تباين بين التحليلات ويأخذ في الاعتبار عدم الخطية في مكانها الصحيح. .
لا تتطلب دوال الاستجابة المودلة وتقديرات الملاءمة المتوقعة أن تمر عبر الأصل (مما يعني وجود تقاطع غير صفري)، ومع ذلك تتطلب النظرية عدم وجود تغيير في العائد مع عدم وجود تغيير في درجة الحرارة. قمنا بإجراء تعديلات بعد الحدث من خلال تقدير العوائد المتوقعة مرتين؛ مرة كما هو موضح أعلاه (أي، الأصلية) ومرة على بيانات التنبؤ المعدلة حيث درجة الحرارة، وهطول الأمطار و تم تعيين التغيرات على صفر. ثم قمنا بطرح التوقعات الأصلية من التغيرات في العائد المتوقعة عند عدم حدوث أي تغيير في متغيرات المناخ.

تقييم عدم اليقين

قمنا بتحديد أربعة مصادر مختلفة من عدم اليقين الناشئة عن اختيار النموذج، وأخذ العينات، والبيانات المفقودة، وتنوع النماذج المتعددة في CMIP6 عبر نماذج GCM، وفقًا للنهج في المرجع. قمنا بتطبيق النماذج والتنبؤ على 100 عينة من البلوك بوتستراب لكل مجموعة من خيارات النموذج، والمحاصيل، ونموذج المناخ العالمي، ومجموعة البيانات المعززة بالتعويض، مما أسفر عن 230,000 تقدير متوقع لتغير العائد العالمي الموزون بالإنتاج في الفترة من 2081 إلى 2100 (أي التقديرات المحسوبة كتغيرات في العائد العالمي بالنسب المئوية، موزونة بحجم إنتاج المحاصيل الموزعة جغرافياً). تم حظر البوتستراب بواسطة الدراسة، مما يعني أنه يتم سحب مجموعة عشوائية من الدراسات والتقديرات منها في كل عينة بوتستراب. ثم قمنا بتقدير التغير في متوسط العائد العالمي الموزون بإنتاج المحاصيل، مما أسفر عن 575 تقديرًا تمثل تركيبات فريدة من عينة بوتستراب الكتلية، مجموعة البيانات المعززة بالتقدير، اختيار النموذج ونماذج المناخ العالمية من مجموعة CMIP6.
قمنا بتوصيف عدم اليقين الناتج عن اختيار النموذج من خلال تثبيت مجموعة البيانات المعززة بالتعويض، وعينة البوتستراب، وبيانات مخرجات نموذج المناخ العالمي المستخدمة، مع تغيير مواصفات النموذج وحساب متوسط الانحراف المعياري في التغير العالمي في العائدات الموزونة عبر مواصفات النموذج. قمنا بتوصيف المصادر الأخرى لعدم اليقين بنفس الطريقة.

فجوة السعرات الحرارية في الغذاء

قمنا بتقدير إمدادات السعرات الحرارية وطلب السعرات الحرارية باستخدام بيانات منظمة الأغذية والزراعة حول إنتاج الدول على نطاق واسع، وكميات الاستيراد والتصدير، ونسبة المحاصيل المخصصة للغذاء، ومتطلبات الطاقة اليومية المتوسطة، والسعرات الحرارية من الطاقة المقدمة إلى نظام الغذاء حسب المحصول والدولة، وأنماط التجارة التفصيلية (انظر قسم “الطرق”). كما استخدمنا عوامل تحويل السعرات الحرارية للمحاصيل من الأدبيات بيانات العائد الموزعة ومساحة الهكتارات المحصودة حوالي 2015 من مجموعة بيانات عالمية جديدة عالية الدقة. وتوقعات السكان لـ SSP1-2.6 و SSP2-4.5 و SSP5-8.5 التي تم تطويرها بواسطة المعهد الدولي لتحليل النظم التطبيقية (IIASA) .
لتقدير الإنتاج الأساسي في عام 2015 من الذرة والأرز وفول الصويا والقمح بالأطنان حسب البلد، قمنا بضرب المساحات المحصودة (هكتار) في الغلات المحصودة (طن/هكتار) في مجموعة بيانات مكانية بدقة 5 دقائق. قمنا بتقدير حصة الإمدادات المحلية لكل دولة من كل محصول المخصصة لنظام الغذاء وحصة إنتاج كل دولة من كل محصول يتم تصديره، متوسطًا عبر الفترة من 2014 إلى 2016 باستخدام بيانات من جداول توازن الغذاء التابعة لمنظمة الأغذية والزراعة. قمنا بتقدير حصة صادرات كل دولة من كل محصول تُصدَّر إلى كل شريك تجاري، متوسطًا عبر الفترة من 2014 إلى 2016 باستخدام بيانات من مصفوفة التجارة التفصيلية لمنظمة الأغذية والزراعة. من هذه الحسابات استعدنا واردات كل دولة من كل محصول من خلال جمع الصادرات عبر الشركاء المستوردين. استخدمنا توافق منظمة الأغذية والزراعة بين المنتجات الغذائية في مصفوفة التجارة التفصيلية وبيانات توازن الغذاء (الجدول التكميلية 2). لتقدير الإمداد المحلي من الذرة والأرز وفول الصويا والقمح بالسعرات الحرارية، طبقنا عوامل تحويل السعرات الحرارية لكل محصول من الأدبيات على تقديرنا للإمداد المحلي من الإنتاج والواردات بالأطنان. لتقدير إجمالي العرض المحلي من السعرات الحرارية من جميع المحاصيل لكل دولة، قمنا بتعديل سعرات المحاصيل الأربعة الأساسية حسب حصتها من إجمالي السعرات الحرارية التي توفرها جميع المحاصيل في كل دولة. . لتقدير الإنتاج المستقبلي، طبقنا التغيرات النسبية المتوقعة في العائدات من نموذج الانحدار العشوائي GLMM على العائدات الأساسية الموزعة والهكتارات المحصودة كما هو موضح أعلاه. كررنا كل خطوة لتقدير إجمالي العرض المحلي من السعرات الحرارية من جميع المحاصيل في كل فترة زمنية مستقبلية.
تفترض هذه الخطوات أن حصة الصادرات من إنتاج البلاد وحصة الشركاء في التصدير تظل ثابتة في المستقبل. نستخدم بيانات عن مساحة الأراضي المحصودة. ، والتي تفترض أن تكرار الحصاد وإجمالي مساحة الأراضي الزراعية القائمة تظل ثابتة. تجعل هذه النتائج محافظة، حيث تظهر الاتجاهات التاريخية أن الأراضي المحصودة قد زادت في العقود القليلة الماضية. . نتائجنا أيضًا محافظة بسبب الاختيار الضيق لمنتجات المحاصيل الأساسية الممثلة في مجموعة بيانات العائدات الموزعة والهكتارات المحصودة. وبالتالي في تقديراتنا (الجدول التكميلي 2).
لتقدير الطلب الأساسي على السعرات الحرارية في عام 2015 حسب البلد، ضربنا متوسط الاحتياجات اليومية من الطاقة لكل فرد (ADER) على مدار 2014-2016 في أرقام السكان من FAOSTAT. . لتقدير الطلب المستقبلي على السعرات الحرارية مع الأخذ في الاعتبار نمو السكان ولكن مع الحفاظ على أنماط النظام الغذائي ثابتة، ضربنا ADER السنوي لكل فرد في توقعات SSP لعدد السكان على مستوى البلاد من IIASA المتوسطة عبر كل من الفترات المستقبلية البالغة 20 عامًا. . حسبنا الفجوة في السعرات الحرارية لكل بلد كإجمالي السعرات الحرارية المطروحة مطروحًا منها إجمالي السعرات الحرارية المطلوبة. تشير الفجوة الإيجابية في السعرات الحرارية إلى أن البلد غير قادر على تلبية الطلب الوطني على السعرات الحرارية من السعرات الحرارية التي توفرها الإنتاج المحلي والواردات. كما فعلنا ذلك للسنة الأساسية 2015 ولكل من الفترات الأربع البالغة 20 عامًا من 2021 إلى 2100، تمكنا من تحديد التغيير في فجوة السعرات الحرارية من 2015 إلى الفترات الزمنية المستقبلية.

توفر البيانات

البيانات والرموز المستخدمة في هذا التحليل متاحة علىhttps://github.com/christineklli/global-yield-impacts. تم تقديم تأثيرات العائدات المودلة للبلدان كبيانات تكملية.
تاريخ الاستلام: 22 مايو 2023؛ تاريخ القبول: 15 يناير 2025
تم النشر عبر الإنترنت: 22 يناير 2025

References

  1. Roson, R. & Sartori, M. Estimation of climate change damage functions for 140 regions in the GTAP 9 data base. J. Glob. Econ. Anal. 1(2), 78-115 (2016).
  2. Newell, R. G., Prest, B. C. & Sexton, S. E. The GDP-temperature relationship: Implications for climate change damages. J. Environ. Econ. Manag. 108, 102445 (2021).
  3. Liu, B. et al. Similar estimates of temperature impacts on global wheat yield by three independent methods. Nat. Clim. Change 6, 1130-1136 (2016).
  4. Challinor, A. J. et al. A meta-analysis of crop yield under climate change and adaptation. Nat. Clim. Change 4, 287-291 (2014).
  5. Kopp, R., Hsiang, S. & Oppenheimer, M. Empirically calibrating damage functions and considering stochasticity when integrated assessment models are used as decision tools. Impacts World 2013, International Conference on Climate Change Effects 12 (2013).
  6. Tol, R. S. J. The economic effects of climate change. J. Econ. Perspect. 23, 29-51 (2009).
  7. Tol, R. S. J. Correction and update: The economic effects of climate change. J. Econ. Perspect. 28, 221-226 (2014).
  8. Wang, X. et al. Emergent constraint on crop yield response to warmer temperature from field experiments. Nat. Sustain. 3, 908-916 (2020).
  9. Whetton, P. H., Grose, M. R. & Hennessy, K. J. A short history of the future: Australian climate projections 1987-2015. Clim. Serv. 2-3, 1-14 (2016).
  10. Nelson, J. P. & Kennedy, P. E. The use (and abuse) of meta-analysis in environmental and natural resource economics: An assessment. Environ. Resour. Econ. 42, 345-377 (2009).
  11. Howard, P. H. & Sterner, T. Few and not so far between: A meta-analysis of climate damage estimates. Environ. Resour. Econ. 68, 197-225 (2017).
  12. Moore, F. C., Baldos, U. L. C., Hertel, T. W. & Diaz, D. New science of climate change impacts on agriculture implies higher social cost of carbon. Nat. Commun. 9, 1 (2017).
  13. CGIAR. Agriculture Impacts. https://web.archive.org/web/20211204202518/; http://www.ag-impacts.org/ (2021).
  14. FAOSTAT. Food Balance Sheets. https://www.fao.org/faostat/en/#data/FBS (2023).
  15. Bell, A., Fairbrother, M. & Jones, K. Fixed and random effects models: Making an informed choice. Qual. Quant. 53, 1051-1074 (2019).
  16. Asseng, S., Foster, I. & Turner, N. C. The impact of temperature variability on wheat yields. Glob. Change Biol. 17, 997-1012 (2011).
  17. Porter, J. R. & Semenov, M. A. Crop responses to climatic variation. Philos. Trans. R. Soc. B Biol. Sci. 360, 2021-2035 (2005).
  18. Semenov, M. & Porter, J. Climatic variability and the modelling of crop yields. Agric. For. Meteorol. 73, 265-283 (1995).
  19. Moriondo, M., Giannakopoulos, C. & Bindi, M. Climate change impact assessment: The role of climate extremes in crop yield simulation. Clim. Change 104, 679-701 (2011).
  20. Springer, C. J. & Ward, J. K. Flowering time and elevated atmospheric CO 2. New Phytol. 176, 243-255 (2007).
  21. Wheeler, T. R., Craufurd, P. Q., Ellis, R. H., Porter, J. R. & Vara Prasad, P. Temperature variability and the yield of annual crops. Agric. Ecosyst. Environ. 82, 159-167 (2000).
  22. Rosenzweig, C., Iglesius, A., Yang, X. B., Epstein, P. & Chivian, E. Climate Change and Extreme Weather Events-Implications for Food Production, Plant Diseases, and Pests (NASA Publications, 2001).
  23. Hatfield, J. L. & Prueger, J. H. Temperature extremes: Effect on plant growth and development. Weather Clim. Extremes 10, 4-10 (2015).
  24. Craufurd, P. Q. & Wheeler, T. R. Climate change and the flowering time of annual crops. J. Exp. Bot. 60, 2529-2539 (2009).
  25. Lobell, D. B., Sibley, A. & Ivan Ortiz-Monasterio, J. Extreme heat effects on wheat senescence in India. Nat. Clim. Change 2, 186-189 (2012).
  26. Lobell, D. B., Baldos, U. L. C. & Hertel, T. W. Climate adaptation as mitigation: The case of agricultural investments. Environ. Res. Lett. 8, 015012 (2013).
  27. Masson-Delmotte, V. et al. IPCC, 2021: Summary for policymakers. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (2021).
  28. Riahi, K. et al. The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: An overview. Glob. Environ. Change 42, 153-168 (2017).
  29. FAOSTAT. Detailed Trade Matrix. https://www.fao.org/faostat/en/#data/TM (2023).
  30. Cassidy, E. S., West, P. C., Gerber, J. S. & Foley, J. A. Redefining agricultural yields: From tonnes to people nourished per hectare. Environ. Res. Lett. 8, 034015 (2013).
  31. Grogan, D., Frolking, S., Wisser, D., Prusevich, A. & Glidden, S. Global gridded crop harvested area, production, yield, and monthly physical area data circa 2015. Sci. Data 9, 15 (2022).
  32. Kc, S. & Lutz, W. The human core of the shared socioeconomic pathways: Population scenarios by age, sex and level of education for all countries to 2100. Glob. Environ. Change 42, 181-192 (2017).
  33. Tilman, D., Balzer, C., Hill, J. & Befort, B. L. Global food demand and the sustainable intensification of agriculture. Proc. Natl. Acad. Sci. 108, 20260-20264 (2011).
  34. Ivanovich, C. C., Sun, T., Gordon, D. R. & Ocko, I. B. Future warming from global food consumption. Nat. Clim. Change 1, 1-6 (2023).
  35. Ray, D. K. & Foley, J. A. Increasing global crop harvest frequency: Recent trends and future directions. Environ. Res. Lett. 8, 044041 (2013).
  36. Fahad, S. et al. Crop production under drought and heat stress: Plant responses and management options. Front. Plant Sci. 8, 1 (2017).
  37. Daryanto, S., Wang, L. & Jacinthe, P.-A. Global synthesis of drought effects on maize and wheat production. PLoS ONE 11, e0156362 (2016).
  38. Marothia, D. et al. Abiotic Stress in Plants (IntechOpen, 2020).
  39. Costa, M. V. J. D. et al. Combined drought and heat stress in rice: Responses, phenotyping and strategies to improve tolerance. Rice Sci. 28, 233-242 (2021).
  40. Hussain, H. A. et al. Interactive effects of drought and heat stresses on morpho-physiological attributes, yield, nutrient uptake and oxidative status in maize hybrids. Sci. Rep. 9, 3890 (2019).
  41. Jumrani, K. & Bhatia, V. S. Interactive effect of temperature and water stress on physiological and biochemical processes in soybean. Physiol. Mol. Biol. Plants 25, 667-681 (2019).
  42. Ostmeyer, T. et al. Impacts of heat, drought, and their interaction with nutrients on physiology, grain yield, and quality in field crops. Plant Physiol. Rep. 25, 549-568 (2020).
  43. Iqbal, M. M., Arif, M. & Khan, A. M. Climate Change Aspersions on Food Security of Pakistan 11 (2009).
  44. Lobell, D. B. & Burke, M. B. Why are agricultural impacts of climate change so uncertain? The importance of temperature relative to precipitation. Environ. Res. Lett. 3, 034007 (2008).
  45. Hausfather, Z. Explainer: What climate models tell us about future rainfall. Carbon Brief. https://www.carbonbrief.org/explaine r-what-climate-models-tell-us-about-future-rainfall/ (2022).
  46. Deryng, D., Conway, D., Ramankutty, N., Price, J. & Warren, R. Global crop yield response to extreme heat stress under multiple climate change futures. Environ. Res. Lett. 9, 034011 (2014).
  47. Hatfield, J. L. et al. Climate impacts on agriculture: Implications for crop production. Agron. J. 103, 351-370 (2011).
  48. FAOSTAT. Crops and Livestock Products. https://www.fao.org/faostat/en/#data/QCL (2023).
  49. Fahad, S. et al. Consequences of high temperature under changing climate optima for rice pollen characteristics-concepts and perspectives. Arch. Agron. Soil Sci. 64, 1473-1488 (2018).
  50. Fahad, S. et al. A combined application of biochar and phosphorus alleviates heat-induced adversities on physiological, agronomical and quality attributes of rice. Plant Physiol. Biochem. 103, 191-198 (2016).
  51. Bradford, J. B. et al. Future soil moisture and temperature extremes imply expanding suitability for rainfed agriculture in temperate drylands. Sci. Rep. 7, 12923 (2017).
  52. Tuninetti, M., Ridolfi, L. & Laio, F. Ever-increasing agricultural land and water productivity: A global multi-crop analysis. Environ. Res. Lett. 15, 09402 (2020).
  53. Gerten, D. et al. Global water availability and requirements for future food production. J. Hydrometeorol. 12, 885-899 (2011).
  54. Billen, G. et al. Reshaping the European agro-food system and closing its nitrogen cycle: The potential of combining dietary change, agroecology, and circularity. One Earth 4, 839-850 (2021).
  55. World-Bank. Water in Agriculture World Bank. https://www.worldbank.org/en/topic/water-in-agriculture (2023).
  56. Elliott, J. et al. Constraints and potentials of future irrigation water availability on agricultural production under climate change. Proc. Natl. Acad. Sci. 111, 3239-3244 (2014).
  57. Rosa, L. et al. Potential for sustainable irrigation expansion in a C warmer climate. Proc. Natl. Acad. Sci. 117, 29526-29534 (2020).
  58. Wisser, D. et al. Global irrigation water demand: Variability and uncertainties arising from agricultural and climate data sets. Geophys. Res. Lett. 35, 1 (2008).
  59. Ortiz-Bobea, A., Ault, T. R., Carrillo, C. M., Chambers, R. G. & Lobell, D. B. Anthropogenic climate change has slowed global agricultural productivity growth. Nat. Clim. Change 11, 306-312 (2021).
  60. WRI. Resource Watch Aqueduct Stress Projections. https://resourcewatch.org/data/explore (2023).
  61. Luck, M., Landis, M. & Gassert, F. Aqueduct Water Stress Projections: Decadal Projections of Water Supply and Demand Using CMIP5 GCMs 20 (2015).
  62. Puy, A., LoPiano, S. & Saltelli, A. Current models underestimate future irrigated areas. Geophys. Res. Lett. 47, e2020087360 (2020).
  63. Vanschoenwinkel, J. & Van Passel, S. Climate response of rainfed versus irrigated farms: The bias of farm heterogeneity in irrigation. Clim. Change 147, 225-234 (2018).
  64. Kompas, T., Che, T. N. & Grafton, R. Q. The Impact of Water and Heat Stress from Global Warming on Agricultural Productivity and Food Security (Global Commission on the Economics of Water, 2023).
  65. Roy, S., Chowdhury, N., Roy, S. & Chowdhury, N. Abiotic Stress in Plants (IntechOpen, 2020).
  66. Uçarlı, C. Abiotic Stress in Plants (IntechOpen, 2020).
  67. Fahad, S. et al. Phytohormones and plant responses to salinity stress: A review. Plant Growth Regul. 75, 391-404 (2015).
  68. Lal, M. Implications of climate change in sustained agricultural productivity in South Asia. Reg. Environ. Change 11, 79-94 (2011).
  69. Soares, J. C., Santos, C. S., Carvalho, S. M. P., Pintado, M. M. & Vasconcelos, M. W. Preserving the nutritional quality of crop plants under a changing climate: Importance and strategies. Plant Soil 443, 1-26 (2019).
  70. De Vries, W., Kros, J., Kroeze, C. & Seitzinger, S. P. Assessing planetary and regional nitrogen boundaries related to food security and adverse environmental impacts. Curr. Opin. Environ. Sustain. 5, 392-402 (2013).
  71. MacDonald, G. K., Bennett, E. M., Potter, P. A. & Ramankutty, N. Agronomic phosphorus imbalances across the world’s croplands. Proc. Natl. Acad. Sci. 108, 3086-3091 (2011).
  72. Zhang, C. et al. The role of nitrogen management in achieving global sustainable development goals. Resour. Conserv. Recycl. 201, 107304 (2024).
  73. Gu, B. et al. Cost-effective mitigation of nitrogen pollution from global croplands. Nature 613, 77-84 (2023).
  74. Qiao, L. et al. Assessing the contribution of nitrogen fertilizer and soil quality to yield gaps: A study for irrigated and rainfed maize in China. Field Crops Res. 273, 108304 (2021).
  75. Vanlauwe, B. & Dobermann, A. Sustainable intensification of agriculture in sub-Saharan Africa: First things first. Front. Agric. Sci. Eng. 7, 376-382 (2020).
  76. Cassman, K. G. & Dobermann, A. Nitrogen and the future of agriculture: 20 years on. Ambio 51, 17-24 (2022).
  77. Zhang, X. et al. Managing nitrogen for sustainable development. Nature 528, 51-59 (2015).
  78. Zhang, X. et al. Quantifying nutrient budgets for sustainable nutrient management. Glob. Biogeochem. Cycles 34, e2018006060 (2020).
  79. Juroszek, P. & von Tiedemann, A. Linking plant disease models to climate change scenarios to project future risks of crop diseases: A review. J. Plant Dis. Prot. 122, 3-15 (2015).
  80. Bradshaw, C. et al. Climate change in pest risk assessment: Interpretation and communication of uncertainties. EPPO Bull. 54, 4-19 (2024).
  81. Giorgi, F. Thirty years of regional climate modeling: Where are we and where are we going next? J. Geophys. Res. Atmos. 124, 5696-5723 (2019).
  82. Hawkins, E. & Sutton, R. The potential to narrow uncertainty in regional climate predictions. Bull. Am. Meteorol. Soc. 90, 10951108 (2009).
  83. Czajkowski, M. & ŠčasnÃý, M. Study on benefit transfer in an international setting. How to improve welfare estimates in the case of the countries’ income heterogeneity? Ecol. Econ. 69, 2409-2416 (2010).
  84. Muleke, A., Harrison, M. T., Yanotti, M. & Battaglia, M. Yield gains of irrigated crops in Australia have stalled: The dire need for adaptation to increasingly volatile weather and market conditions. Curr. Res. Environ. Sustain. 4, 100192 (2022).
  85. Ciscar, J.-C., Fisher-Vanden, K. & Lobell, D. B. Synthesis and review: An inter-method comparison of climate change impacts on agriculture. Environ. Res. Lett. 13, 070401 (2018).
  86. Lobell, D. B., Thau, D., Seifert, C., Engle, E. & Little, B. A scalable satellite-based crop yield mapper. Remote Sens. Environ. 164, 324-333 (2015).
  87. Roberts, M. J., Braun, N. O., Sinclair, T. R., Lobell, D. B. & Schlenker, W. Comparing and combining process-based crop models and statistical models with some implications for climate change. Environ. Res. Lett. 12, 095010 (2017).
  88. Moore, F. C., Baldos, U. L. C. & Hertel, T. Economic impacts of climate change on agriculture: A comparison of process-based and statistical yield models. Environ. Res. Lett. 12, 065008 (2017).
  89. Harris, I., Osborn, T. J., Jones, P. & Lister, D. Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data 7, 109 (2020).
  90. Sacks, W. J., Deryng, D., Foley, J. A. & Ramankutty, N. Crop planting dates: An analysis of global patterns. Glob. Ecol. Biogeogr. 19, 607-620 (2010).
  91. Portmann, F. T., Siebert, S. & Döll, P. MIRCA2000-Global monthly irrigated and rainfed crop areas around the year 2000: A new high-resolution data set for agricultural and hydrological modeling. Glob. Biogeochem. Cycles 24, 1 (2010).
  92. Monfreda, C., Ramankutty, N. & Foley, J. A. Farming the planet: 2. Geographic distribution of crop areas, yields, physiological types, and net primary production in the year 2000. Glob. Biogeochem. Cycles 22, 1 (2008).
  93. Graham, J. W. Missing data analysis: Making it work in the real world. Annu. Rev. Psychol. 60, 549-576 (2009).
  94. Schafer, J. L. & Graham, J. W. Missing data: Our view of the state of the art. Missing Data 31, 1 (2002).
  95. Van Buuren, S. & Groothuis-Oudshoorn, K. Mice: Multivariate imputation by chained equations in R. J. Stat. Softw. 67, 1 (2011).
  96. Yucel, R. M. Multiple imputation inference for multivariate multilevel continuous data with ignorable non-response. Philos. Trans. Ser. A Math. Phys. Eng. Sci. 366, 2389-2403 (2008).
  97. Heymans, M. W. & Eekhout, I. Applied Missing Data Analysis with SPSS and (R)Studio (2019).
  98. Schomaker, M. & Heumann, C. Model selection and model averaging after multiple imputation. Comput. Stat. Data Anal. 71, 758-770 (2014).
  99. Wood, S. N. Generalized Additive Models: An introduction with R 397 (2017).
  100. Harrer, M., Cuijpers, P., Furukawa, T. A. & Ebert, D. D. Doing Meta-analysis in R (2021).
  101. WorldClim. Future Climate, 10 Minutes Spatial Resolution—WorldClim 1 Documentation. https://www.worldclim.org/data/cmip 6/cmip6_clim10m.html (2023).
  102. KNMI, C.E. Monthly CMIP5 Scenario Runs. https://climexp.knmi.nl/start.cgi (2023).
  103. US Department of Commerce, N. Global Monitoring Laboratory—Carbon Cycle Greenhouse Gases. https://gml.noaa.gov/ccgg/tre nds/data.html (2023).
  104. White, I. R., Royston, P. & Wood, A. M. Multiple imputation using chained equations: Issues and guidance for practice. Stat. Med. 30, 377-399 (2011).
  105. Rubin, D. B. & Schenker, N. Multiple Imputation for Interval Estimation from Simple Random Samples with Ignorable Nonresponse 10 (1986).
  106. Fischer, E. M., Sedláěek, J., Hawkins, E. & Knutti, R. Models agree on forced response pattern of precipitation and temperature extremes. Geophys. Res. Lett. 41, 8554-8562 (2014).
  107. Wilcox, J. & Makowski, D. A meta-analysis of the predicted effects of climate change on wheat yields using simulation studies. Field Crop Res. 156, 180-190 (2014).

الشكر والتقدير

تم تمويل C.L. من قبل مركز التميز لتحليل مخاطر الأمن الحيوي. نشكر J. Baumgartner على تقديم نصائح برمجية.

مساهمات المؤلفين

A.R. وC.L. وJ.C. وT.K. تصوروا التحليل، قامت C.L. بإجراء التحليل وكتبت المسودة الأصلية. راجع جميع المؤلفين وحرروا المخطوطة.

الإعلانات

المصالح المتنافسة

يعلن المؤلفون عدم وجود مصالح متنافسة.

معلومات إضافية

المعلومات التكميلية النسخة عبر الإنترنت تحتوي على مواد تكميلية متاحة علىhttps://doi.org/1 .
يجب توجيه المراسلات وطلبات المواد إلى C.L.
معلومات إعادة الطبع والتصاريح متاحة علىwww.nature.com/reprints.
ملاحظة الناشر تظل Springer Nature محايدة فيما يتعلق بالمطالبات القضائية في الخرائط المنشورة والانتماءات المؤسسية.
الوصول المفتوح هذه المقالة مرخصة بموجب رخصة المشاع الإبداعي للاستخدام غير التجاري، والتي تسمح بأي استخدام غير تجاري، ومشاركة، وتوزيع، وإعادة إنتاج في أي وسيلة أو صيغة، طالما أنك تعطي الائتمان المناسب للمؤلفين الأصليين والمصدر، وتوفر رابطًا لرخصة المشاع الإبداعي، وتوضح إذا قمت بتعديل المادة المرخصة. ليس لديك إذن بموجب هذه الرخصة لمشاركة المواد المعدلة المشتقة من هذه المقالة أو أجزاء منها. الصور أو المواد الأخرى من طرف ثالث في هذه المقالة مشمولة في رخصة المشاع الإبداعي للمقالة، ما لم يُذكر خلاف ذلك في سطر ائتمان للمادة. إذا لم تكن المادة مشمولة في رخصة المشاع الإبداعي للمقالة واستخدامك المقصود غير مسموح به بموجب اللوائح القانونية أو يتجاوز الاستخدام المسموح به، ستحتاج إلى الحصول على إذن مباشرة من صاحب حقوق الطبع والنشر. لعرض نسخة من هذه الرخصة، قم بزيارةhttp://creativecommo ns.org/licenses/by-nc-nd/4.0/.
© المؤلفون 2025

  1. مدرسة العلوم الحيوية، مركز التميز لتحليل مخاطر الأمن الحيوي، جامعة ملبورن، ملبورن 3010، أستراليا. مدرسة الرياضيات والإحصاء، جامعة ملبورن، ملبورن 3010، أستراليا. مدرسة الزراعة، علوم الغذاء والبيئة، جامعة ملبورن، ملبورن 3010، أستراليا. البريد الإلكتروني:christinel3@unimelb.edu.au

Journal: Scientific Reports, Volume: 15, Issue: 1
DOI: https://doi.org/10.1038/s41598-025-87047-y
PMID: https://pubmed.ncbi.nlm.nih.gov/39843615
Publication Date: 2025-01-22

scientific reports

OPEN

Predicting changes in agricultural yields under climate change scenarios and their implications for global food security

Christine Li □ , James Camac , Andrew Robinson & Tom Kompas

Climate change has direct impacts on current and future agricultural productivity. Statistical metaanalysis models can be used to generate expectations of crop yield responses to climatic factors by pooling data from controlled experiments. However, methodological challenges in performing these meta-analyses, together with combined uncertainty from various sources, make it difficult to validate model results. We present updates to published estimates of crop yield responses to projected temperature, precipitation, and CO2 patterns and show that mixed effects models perform better than pooled OLS models on root mean squared error (RMSE) and explained deviance, despite the common usage of pooled OLS in previous meta-analyses. Based on our analysis, the use of pooled OLS may underestimate yield losses. We also use a block-bootstrapping approach to quantify uncertainty across multiple dimensions, including modeler choices, climate projections from the sixth Coupled Model Intercomparison Project (CMIP6), and emissions scenarios from Shared Socioeconomic Pathways (SSP). Our estimates show projected yield responses of – (maize), – (rice), – (soy), and 14% (wheat) from 2015 to 2080-2100 under the business-as-usual scenario of SSP5-8.5, which reduce to , and respectively under the lower emissions scenario of SSP1-2.6. Without mitigation and adaptation, countries in South Asia, sub-Saharan Africa, North America, and Oceania could become at risk of being unable to meet national calorie demand by the end of the century under the most severe emissions scenario.
Climate change is expected to directly impact agricultural production by reducing both crop yield and quality via changing patterns in temperature, water, gases and nutrients. Moreover, changing climate can also have indirect impacts on yields by altering impacts caused by pests, diseases and weeds. Statistical models of crop yield response to climate change can be estimated using data from controlled experiments. These models can be used to generate expectations of climate impacts in new settings such as future time periods or geographies with relatively scarce primary studies . Predictions from such models are highly relevant to climate policyclimate yield response functions have been included in a number of Integrated Assessment Models (IAMs) and Intergovernmental Panel on Climate Change (IPCC) assessments of the impacts of climate change on food production . However, it is difficult to validate the general usefulness of yield response functions applied to ‘out of sample settings. This difficulty is partly due to methodological challenges in estimating general relationships and transferring said relationships across settings, and partly because various sources of uncertainty throughout the modelling process can combine and propagate in confounding ways to modelled predictions. Estimating general relationships using data collected from studies-the technique of meta-analysis-is fraught with methodological challenges that can result in certain patterns in the data being over- or under-represented. A previous meta-analysis of climate impact estimates found strong evidence of duplication bias from betweenstudy correlation, as well as within-study correlation from the inclusion of multiple estimates . In the case of climate-yield responses, such challenges can lead to higher or lower yield responses than are warranted given the true structure of the data. Once general relationships have been estimated, how these estimates are transferred to new settings and interpreted in their context is also subject to many uncertainties. At a basic level, there is a large degree of variation across climate impact modelers’ data sampling strategies, choices to omit or impute missing values, and choices or assumptions related to the model specification and parameterization with input data. It is
helpful for modellers to clarify these ambiguities, which we do in this study. Going further, the trajectory of global economic development and the effect of human emissions on earth systems and climate is often referred to as irreducibly uncertain, where more information may not necessarily reduce this uncertainty . These ‘irreducible’ or ‘deep’ uncertainties cannot be collapsed, but we argue that impact modellers should aim to include inputs from a large range of climate models and emission scenarios, which will help users and policymakers better understand the range of possible impacts. The goal is to clarify the sources of uncertainty, making it easier to trust and use yield response predictions at policy-making levels.
Our study has three objectives: (1) to improve the estimation of yield responses to climatic factors such as temperature, precipitation and CO2, using data from an established crop yield response database (CGIAR data), (2) to decompose the range of uncertainty in resulting predictions into its separate sources, and (3) to provide policy-relevant estimates related to global food security. Our goal was to provide a comprehensive yet up to date estimation of global climate yield responses while also considering implications for food security and associated uncertainty under a greater set of emissions scenarios and time periods. First, we hypothesised that fitting mixed models to the CGIAR dataset would help to account for within-study correlation between multiple estimates in the dataset, resulting in better statistical model performance compared to pooled OLS models used in previous meta-analyses of the same dataset . We could then compare general expectations of crop yield responses to climate change obtained by following this approach, to the expectations derived from these older meta-analyses . Second, we used a bootstrap sampling technique repeated on the dataset across multiple dimensions of uncertainty, to see how different sources of uncertainty throughout the modelling process propagate through to estimated global yield responses. This allowed us to systematically quantify the share of variance that is attributable to data sampling, missing values, model specification, climate model input data from the sixth Coupled Model Intercomparison Project (CMIP6) multi-model ensemble of Global Circulation Models (GCMs). Finally, we estimated projected crop yield responses impacts using not only future projected temperatures but also future projected precipitation, which was not done in earlier meta-analyses fitted on the same dataset . We applied the outputs of the preferred response model to calculate country-level calorie gaps and domestic food security status for three emissions scenarios of varying severity.

Results

Modelled yield response functions

We began by refreshing the CGIAR database with screening and imputation of missing data . This database aggregates 74 studies related to maize, rice, soy and wheat (comprising of calories consumed globally in ) containing over 8800 point estimates of changes in crop yield across varying temperature, precipitation, and other factors. We did not add studies to the database; it therefore remains a non-exhaustive collection of information on yield responses to climatic factors. The CGIAR database has a nested or hierarchical structure in which multiple estimates can come from a single study; these estimates should be treated as correlated or non-independent. Similarly, many estimates are derived from the same region or country, and we also assume these to be correlated.
We fit statistical models for each crop separately (see “Methods” section for full specifications). Our five candidate models are: (1) Ordinary Least Squares (OLS) pooled model, (2) Generalised Linear Mixed Model (GLMM) with random intercepts, (3) GLMM with random intercepts and random slopes, (4) Generalised Additive Mixed Model (GAMM) with random intercepts and (5) GAMM with random intercepts and random slopes, plotted in Fig. 1. Model 1 is a pooled OLS model which treats point estimates from the same study and country as though they are independent, and are similar to the OLS models used in previous published metaanalysis on the same CGIAR dataset except that our response functions are separately estimated for each of the four crops, and do not include an interaction term between quadratic temperature change and baseline temperature (see “Methods” section). Models 2 and 4 include random intercepts for studies and countries in the dataset, where study random intercepts account for correlation among multiple estimates from the same study, and country random intercepts accounts for correlation between studies that include estimates for the same country . Including country random intercepts helps to capture non-weather factors associated with location, such as pest and disease susceptibility, soil fertility, and farming behaviour, and also allows for fitted estimates to be predicted on new countries not included in the CGIAR data. Models 3 and 5 include random intercepts for studies and countries but also random slopes for countries, which allows temperature, precipitation and CO2 functions to vary for each country (i.e. estimates a separate function for each country). Random effects allow for residual variance to partitioned among the hierarchical levels of countries and studies in the data. GAMMs are a non-parametric type of generalised linear model which allows the response variable (here yield change) to depend on smooth functions of model terms. Model terms were chosen following earlier literature and relevance to crop growth (see “Methods” section) . Response functions were calibrated to changes in mean temperature, precipitation and CO 2 conditions, and thus, do not represent impacts from extreme and variable temperatures and rainfall events including floods and droughts nor do they capture long term soil and nutrient fluctuations.
We found that fitted temperature-yield response functions are downward sloping for maize, rice and wheat, suggesting that temperature effects in isolation tend to have a yield-reducing effect on crops (Fig. 1a). This response is not seen in soybean. Precipitation-response functions describe a positive relationship between changes in rainfall and yields for all crops except rice, the latter may be due to rice being mostly irrigated (Fig. 1b). For visualisation purposes, these plots show response functions at the median baseline levels of temperature and precipitation, but do not show interactions between changes in temperature and precipitation, or combinations of different baseline levels with changes. What can also be seen from the plotted data points are multiple estimates from the same study, which may represent unique experiments testing responses to combinations of
Fig. 1. Marginal mean response functions ( ) for temperature change (a) and precipitation change (b), conditional on median baseline temperature, median baseline precipitation, and the absence of changes in precipitation (a) or changes in temperature (b), changes in and on-farm adaptation. Points depict study estimates derived from the CGIAR data. Error bars show 95% confidence interval pooling variance within and between 5 imputation-augmented fit estimates. Legend: GLM RI, Generalised Linear Mixed Model (GLMM) with random intercepts; GLM RS, GLMM with random intercepts and slopes; GAM RI, Generalised Additive Mixed Model (GAMM) with random intercepts; GAM RS, GAMM with random intercepts and slopes; LM, pooled Ordinary Least Squares (OLS) model. Supplementary Figure 3 shows some out-of-sample comparisons that were not included in the CGIAR database .
Crop Model Sampling GCMs Missing data Combined
Maize 10.9 8.67 4.52 8.58 32.7
Rice 19.3 12.0 9.46 12.4 53.1
Soy 51.1 48.2 18.6 47.5 165.0
Wheat 15.3 8.57 6.08 8.64 38.6
Table 1. Sources of uncertainty attributable to model choice, sampling, GCMs and imputation-augmented datasets (missing data), as measured by the average standard deviation of global production-weighted yield change (%). Combined uncertainty does not add to as uncertainty is not independent across sources.
temperature, precipitation and CO2. These multi-estimate studies raise the need for mixed effects modelling to account for correlation bias within-study.

Sources of uncertainty

We systematically quantified uncertainty by block-bootstrapping 100 samples of the data and comparing the resulting average standard deviation of estimated global yield responses across five dimensions. These dimensions span: (1) 100 block-bootstrap samples, blocking by study where a different set of studies are selected in each sample (i.e. uncertainty regarding the sampling strategy to compile the CGIAR dataset), (2) five different imputations of the CGIAR dataset to fill in missing values, using Multiple Imputation Chained Equations (MICE, see “Methods” section) (i.e. uncertainty regarding treatment of missing data), (3) five statistical models (i.e. uncertainty regarding the econometric specification), (4) 23 sets of temperature, precipitation and CO2 input data from the CMIP6 ensemble of 23 GCMs, for each of three emissions scenario (i.e. uncertainty regarding the effect of human emissions on the climate system) (see “Methods” section).
We found that uncertainty from model choice is equal to between 10 and of global agricultural yields for maize, rice and wheat and over for soybean (Table 1). All types of uncertainty and combined uncertainty are highest in soybean, followed by rice, wheat and maize. This roughly reflects the different sample sizes of non-imputed CGIAR point estimates related to each crop (maize: 3426; rice: 2776; soy: 788; wheat: 1972). High uncertainty around soybean results reflects the low sample size and considerable variation in these estimates. Variations across GCMs are lowest compared to other types of uncertainty across all crops. Figure 2 shows 575 model distributions of 2081-2100 impact estimates (pooling distributions across GCMs for better legibility).
Fig. 2. Distributions of predicted global production-weighted mean yield change in 2081-2100 from 575 model combinations for each crop ( 5 model specifications imputed datasets GCMs), pooled by GCM to represent only differences across model choice and imputation-augmented dataset. Legend: GLM_RI, Generalised Linear Mixed Model (GLMM) with random intercepts; GLM_RS, GLMM with random intercepts and slopes; GAM_RI, Generalised Additive Mixed Model (GAMM) with random intercepts; GAM_RS, GAMM with random intercepts and slopes; LM, pooled OLS model.
The pooled OLS model generates a narrower distribution of estimated responses compared to both GLMMs for maize, soy and wheat, and compared to GAM models for rice. This suggests that not accounting for study- and country-related correlations in the data may result in underestimates of prediction uncertainty.

Model performance

We evaluated model predictive performance by running k -fold cross validation with tenfolds. We found the GAMM and GLMM with random intercepts and random slopes performed best as measured by Root Mean Squared Error (RMSE), and the pooled OLS model performed the worst. Figure 3 shows this alongside explained deviance for each of the five model specifications (each pooled across five imputed fit estimates). The pooled OLS model fit performed the worst, with the highest mean model RMSE and least explained deviance across all five model specifications.
GLMM and GAMM models with random intercepts and slopes performed better compared to all the GLMM and GAMM models with random intercepts only. This is unsurprising given the structural variation in studies across the CGIAR database; allowing climate variables (temperature, precipitation and change) to vary randomly for each study helps extract ‘true’ relationships from each study. Yet the flexible nature of the nonparametric GAMM models (we do not restrict the number of knots in splines) and the shape of their response functions suggest that the GAMM with random intercepts and slopes could overfit the data. Of the five models, we prefer the GLMM model with random intercepts and slopes as it has a lower tendency to overfit the data compared to the GAMM model despite having similarly good predictive performance.

Global future projected yield changes

We used fitted estimates from the preferred model (GLMM with random intercepts and slopes) to estimate future projected yield changes on a grid ( at the equator) and converted these impacts into calorie measures. We used these measures to analyse risks of food insecurity at the country level from 2021 to 2100. We consider three emissions scenarios using the combinations of Shared Socioeconomic Pathways (SSPs) in CMIP6: : SSP1-2.6, SSP2-4.5 and SSP5-8.5. In line with the IPCC’s Sixth Assessment Report , SSP1-2.6 illustrates a scenario of sustainability with relatively low growth in population, consumption, and concentration of greenhouse gases (GHG), with a best estimate of average global surface temperature warming by 2081-2100 relative to the period 1850-1900. SSP2-4.5 is moderate population scenario, with medium challenges to mitigation and adaptation and warming by 2081-2100. SSP5-8.5 is a scenario of rapid economic growth alongside continued globalisation and fossil-fuelled development combined with high greenhouse gas concentrations, with warming by 2081-2100.
We show CMIP6-ensembled global predictions for the preferred model for SSP5-8.5 in Fig. 4 and for SSP12.6 and SSP2-4.5 in Supp. Fig. 5 and 6. The largest yield reductions are seen in maize, with losses of under SSP1-2.6 and under SSP5-8.5 in 2081-2100 (Tables 2, 3), which is consistent with other studies not included in the CGIAR dataset . Of the major global maize producers, the US and China are expected to suffer the largest yield reductions of and respectively (Table 3). Declining rice yields are expected in
Fig. 3. Model root mean squared error (RMSE) from tenfold cross validation (top) and deviance explained by each model (bottom) across 5 imputed dataset fit estimates. Black circles show model-specific mean RMSE and explained deviance, calculated across imputed datasets. Legend: GLM RI, Generalised Linear Mixed Model (GLMM) with random intercepts; GLM RS, GLMM with random intercepts and slopes; GAM RI, Generalised Additive Mixed Model (GAMM) with random intercepts; GAM RS, GAMM with random intercepts and slopes; LM, pooled OLS model.
the Middle East, South America, and South/South East Asia, worsening over the century with a global weighted mean yield loss of under SSP1-2.6 and by under SSP5-8.5 by 2081-2100. Rice yields are expected to decline in India and Bangladesh by around , while these losses are smaller in China ( ) and Indonesia ( ) (Table 3). Soy yields face considerable uncertainty with projections showing areas of net negative and positive change across regions of Brazil ( ) and the US ( ), while global mean yields show a net increase of under SSP1-2.6 but a net decrease of under SSP5-8.5. Wheat shows relatively small global yield decline of under SSP1-2.6 by 2081-2100, but large losses are expected globally ( ), in India ( ) and the US ( ) under SSP5-8.5 (Table 3).

Global projected food calorie gap

We converted projected global yield change estimates into national calorie supply and compared them to estimates of national calorie demand using data from the FAO and other literature . The calorie gap represents future national calorie demand that cannot be met by production and imports. Our results show reduced calorie supply in many countries in south and southeast Asia, sub-Saharan Africa, with most severe losses expected under SSP5-8.5 relative to baseline calories supplied in 2015 (Fig. 5a,b, Supp. Fig. 7a). In addition to these countries, North America and Oceania could become at risk of not being able to meet national food demand under SSP5-8.5 (Fig. 5c,d). We took a deliberately simple approach to illustrate the potential implications of estimated yield responses on global food security. This approach treats many dynamic factors as fixed, such as harvested areas, average daily energy requirements, the share of each country’s production going to each export partner, and non-staple crops’ share of a country’s diet. We thus assumed static patterns for crop distribution, harvest frequency, total standing cropland area, nations’ demographic compositions, export patterns and dietary patterns (see “Methods” section).
We also replicated calorie gap maps using the pooled OLS model (Supp. Fig. 7b,d). This shows many countries experiencing less severe calorie reductions occurring later in the century compared to the preferred GLMM model with random intercepts and slopes. This could be because multi-estimate studies with less severe yield
Fig. 4. Predicted yield changes derived from Generalised Linear Mixed Model (GLMM) with random intercepts and slopes pooled across 5 imputation models and 23 GCM prediction datasets for SSP5-8.5, shading indicates spatial distribution of crop production c. .
responses are biasing the pooled OLS results, or because negative yield responses are apparent within individual studies, but muted when data from those studies are combined (i.e. Simpson’s paradox).

Discussion

We offer a brief review of our results in the context of the broad and deep scientific literature covering the changing phenology of grain crops in response to elevated heat and drought conditions . Our results identify larger yield losses for maize and rice in many regions than previously estimated using the same CGIAR dataset, by roughly percentage points compared to . Wheat results are comparable to Ref. with a percentage point difference in either direction for most regions. Results from Ref. show slightly larger yield losses over central Asia, northern Africa, South America and western Australia. We believe these differences flow from methodological treatments of the data and the inclusion of precipitation change in our estimation of global gridded yield losses.
Crop Scenario 2021-2040 2041-2060 2061-2080 2081-2100
Maize SSP1-2.6 – 1.6 [- 9.0, 5.7] – 3.3 [- 10.5, 4.0] -4.0 [- 11.2, 3.3] -3.8 [- 11.1, 3.6]
SSP2-4.5 -1.6 [-9.0, 5.7] – 4.6 [- 11.9, 2.7] -7.2 [-14.8, 0.4] – 13.4 [- 17.0, – 1.2]
SSP5-8.5 – 2.0 [- 9.3, 5.3] -6.8 [- 14.3, 0.8] -12.9 [-22.3, -4.5] -22.2 [- 34.3, – 10.1]
Rice SSP1-2.6 0.5 [- 9.1, 10.0] – 1.2 [- 10.8, 8.4] – 2.2 [- 12.0, 7.5] – 2.7 [- 12.5, 7.1]
SSP2-4.5 1.0 [- 8.5, 10.5] – 0.9 [- 10.8, 9.0] – 2.7 [- 13.5, 8.0] – 4.3 [- 15.9, 7.2]
SSP5-8.5 1.2 [- 8.4, 10.7] – 1.8 [- 12.5, 8.9] – 5.8 [- 19.2, 7.7] -9.0 [- 26.8, 8.8]
Soy SSP1-2.6 2.9 [- 21.3, 27.0] 3.3 [- 20.5, 27.1] 2.4 [- 21.5, 26.3] 1.4 [- 22.7, 25.5]
SSP2-4.5 3.5 [- 20.5, 27.5] 5.3 [- 19.3, 30.0] 4.4 [- 22.4, 31.2] 2.6 [- 26.1, 31.4]
SSP5-8.5 4.5 [- 19.7, 28.6] 5.5 [- 21.4, 32.3] -0.6 [-36.0, 34.8] – 15.4 [- 69.2, 38.5]
Wheat SSP1-2.6 1.3 [- 11.0, 13.6] -0.4 [-12.3,11.5] – 1.3 [- 13.2, 10.6] – 1.5 [- 13.5, 10.5]
SSP2-4.5 1.7 [- 10.6, 13.9] -0.7 [-12.6,11.1] – 3.1 [- 15.1, 9.0] -4.9 [- 17.3, 7.5]
SSP5-8.5 1.6 [- 10.6, 13.7] – 2.4 [- 14.4, 9.7] – 7.8 [- 21.5, 6.0] – 14.1 [- 32.6, 4.3]
Table 2. Global production-weighted mean yield change (%) projections from Generalised Linear Mixed Model (GLMM) for SSP1-2.6 (top), SSP2-4.5 (middle) and SSP5-8.5 (bottom), central estimates bolded [95% confidence interval]. Weighting by gridded crop production values. Estimates ignore changes in areas not currently under cultivation.
Crop Largest producers SSP1-2.6 SS2-4.5 SSP5-8.5
Maize US -6.5 [-13.3, 0.3] – 12.6 [- 20.5, – 4.6] -26.0 [-39.3, -12.7]
China -4.7 [- 11.7, 2.2] – 10.1 [- 17.8, – 2.4] -24.7 [- 37.2, – 12.1]
Brazil 1.0 [- 7.5, 9.5] – 3.0 [- 10.7, 4.7] – 13.1 [- 22.6, – 3.7]
Argentina – 0.9 [- 8.3, 6.5] – 4.3 [- 11.6, 3.1] – 12.1 [- 20.8, – 3.5]
Rice China -5.6 [- 13.0, 1.8] -6.1 [- 15.1, 3.0] – 9.8 [- 24.6, 5.0]
India -7.2 [- 14.1, – 0.3] – 10.2 [- 18.6, – 1.7] – 21.3 [- 35.0, – 7.6]
Indonesia 7.4 [- 8.2, 23.0] 3.7 [- 11.7, 19.1] -6.5 [- 22.5, 9.6]
Bangladesh – 10.3 [- 19.8, – 0.7] – 11.9 [- 22.6, – 1.2] – 21.6 [- 36.4, – 6.9]
Soybean Brazil 0.4 [- 28.4, 29.1] 5.4 [- 20.0, 30.9] -5.6 [- 45.9, 34.8]
US – 3.5 [- 22.7, 15.8] – 6.0 [- 32.6, 20.7] – 30.9 [- 89.1, 27.3]
China 1.4 [- 20.7, 23.5] 2.1 [- 25.6, 29.7] – 17.2 [- 73.4, 39.0]
Argentina 2.9 [- 17.5, 23.3] 7.7 [- 14.0, 29.4] 2.9 [- 28.1, 33.8]
Wheat China -4.0 [- 12.9, 4.8] – 5.7 [- 15.3, 3.9] – 12.7 [- 29.2, 3.9]
India -4.6 [- 16.3, 7.1] – 8.5 [- 20.8, 3.8] – 21.7 [- 39.9, – 3.6]
Russia – 1.9 [- 14.3, 10.5] -4.9 [- 17.6, 7.7] – 13.3 [- 33.1, 6.4]
US -4.3 [- 14.5, 6.0] -7.2 [- 18.1, 3.8] – 14.4 [- 32.5, 3.6]
Canada -2.1 [- 13.0, 8.9] -3.4 [- 15.5, 8.8] -8.1 [- 31.0, 14.8]
Table 3. Production-weighted mean yield change projections for 2081-2100 from generalised linear mixed model (GLMM) random intercepts and slopes model for largest producers (%) for SSP1-2.6, SS2-4.5 and SS5-8.5, central estimates bolded [95% confidence interval]. Weighting by gridded crop production values ignores changes in areas not currently under cultivation. See supplementary data for all countries.
Compared to previous meta-analyses of the same CGIAR dataset , our inclusion of precipitation change and its interaction with temperature change and baseline precipitation levels resulted in larger global yield losses for maize and rice. This demonstrates the importance of including temperature and precipitation interactions in climate-agriculture impact modelling. High temperatures often co-occur with, or are caused by, low precipitation, and their interaction can create impacts not otherwise captured by considering temperature or precipitation in isolation from one another. Recent reviews of the interactions of combined heat and drought have found greater yield impacts on grain crops than individual stresses alone , as well as in interaction with nutrient availability . Rising mean temperatures cause more water to evaporate from the soil surface (evapotranspiration losses) and therefore higher crop water requirements . Precipitation is more spatially variable than temperature, and there is major disagreement between GCMs on the direction of precipitation change in many regions . This highlights the importance of including the range of variance in future rainfall projections as we have done here by using input data from 23 GCMs.
The relatively more severe yield losses for maize compared to wheat aligns with other comparative analyses in the literature. Reference also found maize to be more sensitive to drought compared to wheat, partially due to the lengthening Anthesis Silking Interval (ASI) leading to asynchrony of anthesis and silking which affects pollination and reproduction. They posited that differences in drought sensitivity could be attributed to the
Fig. 5. Food security implications estimated from Generalised Linear Mixed Model (GLMM) with random intercepts and slopes. Top: Reduction in total supply of food in calories relative to 2015 baseline (%) for SSP24.5 (a) and SSP5-8.5 (b). Bottom: Change in risk of not meeting national calorie demand from production and imports for SSP2-4.5 (c) and SSP5-8.5 (d). Maps for SSP1-2.6 are in Supp. Fig. 7.
origins of maize in less water-limited regions compared to wheat originating in dryland regions. Wheat has thus acquired adaptability traits including delayed senescence which can be useful for late-season drought, as well as deeper root systems to access sub-surface soil water . On the other hand, wheat has lower limit temperature thresholds ( ) than maize ( ) or soybean ( ) 46,47 , so we would expect declines or losses where these thresholds are surpassed where it is currently grown (see “Methods” section). However, many areas currently too cold for wheat may become viable for wheat up to the point of optimal temperatures for grain filling. A previous meta-analysis of maize and wheat drought responses identified that yields are more drought sensitive if water limited during the reproductive stage than vegetative stage of plant growth . Day and nighttime temperatures moving outside of the optimal temperature range for maximum grain yield is expected to directly affect grain yields (Supp. Table 1) . However, detailed differences in changing temperature and precipitation throughout the growing season cannot be picked up by our response functions here. Of all staple crops, maize has the current largest cultivated area . Assuming maize is grown in areas where the growing season temperatures are below but near optimal temperatures for development and production, further warming may push these locations out of the optimal range of growing temperatures.
Rice yield losses are larger than what has been estimated in previous meta-analyses using the same CGIAR dataset, and here yields are expected to decline in Africa, South America and Australia through to the end of the century in response to higher temperatures. This aligns with scientific understanding of the impact of high temperature stress on rice pollen, fertility and grain yield . Reference estimated yield declines at one degree of global warming for China ( ), Russia ( ), France ( ), Kansas in the US ( ), India ( ), comparable to our projections in 2041-2060 and 2061-2080 showing wheat yields declining by in China, in India, in the US, in Russia and rising by in France. Like , we also see rice yield increases in regions of high elevation and latitude (e.g. Tibetan Plateau, Greenland, Siberia, Andes). Our soy results are starkly different to those previously estimated using the same CGIAR dataset, in which project total crop failure in Africa and Brazil and southern Australia. Other studies have discussed the large degree of uncertainty surrounding soy yield climate responses, which may be due both to its higher critical temperature
tolerance to heat stress around anthesis (compared to maize and wheat) but also its lower limit temperature tolerance. This can be seen in similar studies predicting as many areas with decreased yields as there are with increased yields, particularly in the USA and Brazil . We also found a large band of uncertainty around soybean projections across all dimensions of model choice, sampling, GCMs and missing data.
Our maps of projected global gridded yield changes show expanding suitability for rice and wheat in the northern latitudes and elevations and declining suitability nearly everywhere else (see Fig. 4). This accords with other studies showing growing suitability in northern Europe for rainfed wheat crops due to changes in rainfall and temperature . The reduction of frost occurrence has also been observed to benefit crop yields . From the perspectives of climate adaptation, water saving and nutrient pollution, the regional scale production basket and optimal crop distribution may need to be significantly reconsidered . Our approach assumes that crops are fed by rainwater, yet almost of cultivated land and of global crop production occurs under irrigated conditions . Going forward, freshwater limitations mean that irrigated croplands in some regions may need to revert to rainfed management while expanding in others . Irrigation plays an important role in supporting crop production in drier regions, though weather-driven variability in global irrigation withdrawals has been modest at a global scale to date . Various studies have projected scenarios of future water stress and growing demand from competing industries and changing land use, which will likely constrain expansion of irrigation and diminish the reliability of blue water resources . The climate response of rainfed versus irrigated farms is not well understood on a global scale and deserves further study especially in the context of adaptation to climate change and mitigation of losses . Blue water availability for irrigated crops should also be integrated into future work on this topic .
Our results shed light on shifting crop distributions due to changing temperature and precipitation patterns, however it misses the effects of sea level rise, salinity and water stress from competing land uses . For their global analyses of climate change damages, Ref. estimate the percentage loss of arable coastal land by metre of sea level rise for 140 regions, finding that the physical effects of sea level rise are concentrated among small island states of Oceania and the Pacific. This may lead to increased salinity of inland and coastal arable soils, a significant abiotic stress for crop plants-currently greater than of global croplands are adversely affected by salt stress .
Our estimates, which consider changes in multi-decadal norms, also do not fully capture the effects of variable and extreme temperatures and rainfall events. There is evidence suggesting that crops may be more sensitive to a highly variable climate on a warmer planet on average than to a climate that changes gradually . In South Asia the summer monsoon rainfall is responsible for almost of the variability in total food grain production anomalies in the region . Stochastic weather generators have also been used to show that increased variability in temperature or precipitation in the UK and France sometimes had a larger effect on wheat yields than changes in averages, especially when combined with warmer mean seasonal temperature .
We have found that reductions in calorie supply of between 15 and are possible in 2081-2100 under the worst case emissions scenario of SSP5-8.5 (Fig. 5). Many countries in sub-Saharan Africa, south Asia, North America and Oceania are at heightened risk of being unable to meet domestic food demand through production and imports. Thus adaptation measures including sustainable intensification, expanding areas of cultivation, changing trade patterns or emissions mitigation should all be considered. Indeed, our results for SSP1-2.6 and SSP2-4.5 representing lower emissions scenarios (Supp. Fig. 7) show much smaller calorie deficits for many countries. It bears repeating that we have not assumed any trends in harvest frequency nor total standing cropland area, even though global harvested land area grew roughly four times faster than the total standing cropland area between 2000 and . In reality we would expect harvest frequency and total standing cropland area to continue changing, though not in the same direction across all regions (e.g. crop harvest frequency increased from 1961 to 2011 in Brazil, India and China but decreased in many places in Africa ). Expansion of suitable areas for cultivation may offset some of the yield losses modelled here. However, we believe our estimates of yield losses and damages to be conservative as we have not considered the indirect effects of climate change on food security such as through extreme rainfall, floods, wildfires, pests and diseases.
Nutrient quality is highly likely to be affected by climate change, but was not possible to include in this analysis. Research shows mixed effects of elevated CO2, temperature, salinity, waterlogging and drought stress on nutrient accumulation in soils, with effects being highly dependent on the nutrient, crop plant and intensity level of change . Moving forward, sustainable nutrient and irrigation management underpins farmers’ ability to close yield gaps while maintaining or increasing nutrient quality . Another factor affecting future food security are crop losses from pests and diseases. Climate change is already affecting the range distributions, pest status and synchrony of economically important pests and diseases with host plants . These impacts should be integrated in future modelling of yield responses.
Finally, we have interrogated the sources of uncertainty contributing to variance around estimated global yield responses. We found that uncertainty from model choice is most significant compared to uncertainty from sampling, missing data and GCMs. Regarding model choice uncertainty, ensembling model predictions may not be the best way to represent this variation since some models were shown to outperform others. We have chosen the mixed effects model specification that we believe best achieves the trade-off between predictive performance and over-fitting the data. We are comforted by seeing full model agreement across many regions in the mean sign of future yield projections for all crops except for soy, though these areas of agreement do not necessarily overlap with where crops are currently grown (Supp. Fig. 9). Uncertainty from GCMs reflects uncertainty about the effect of human emissions on the climate system. While this deep uncertainty can be clarified by including a range of CMIP6 models as inputs to climate impact modelling, model ensembles may not represent the complete or systematic sample of the climate response . Other studies have shown that for GCM-based projections of climate response, uncertainty attributable to GCMs and the internal variability of the climate system tends to dominate other sources of uncertainty in the near-term, whereas in the late-century scenario uncertainty
dominates (though GCM uncertainty remains significant) . In our analysis of late-century projections for a single scenario, GCM uncertainty is dominated by model uncertainty and sampling uncertainty. For future metaanalyses, we recommend expanding the set of studies and sample size where possible, or to take a bootstrapping approach to characterise uncertainty from data sampling strategy.
Significant caveats remain on the approach we have taken to statistical modelling here. Perhaps most significantly, function transfer implicitly assumes that estimated function parameters at study sites transfer across time and space . Since only a subset of all countries is represented in the CGIAR data (see Supp. Fig. 1), this assumption obviously may not hold. For example, Australia has exceptionally nutrient poor soils from undergoing a much longer period of weathering and drying than other continents, and a large proportion of crop production takes place under semi-arid and drought conditions. Australian average rainfed yields of key grain crops since the 1960s severely lag the rest of the world with the exception of yields in Africa . Compared to zones with higher baseline rainfall, rainfed crops in Australia generally display higher yield sensitivity to rainfall, meaning that a small increase in rainfall results in relatively higher increases in yield. This suggests that transferring global parameters for yield response to precipitation change may not predict local responses well, even though our models account for interactions between baseline levels and level changes in rainfall, and we used estimated country random effects to predict local yield changes. Factors such as nutrient availability and their interactions with climatic factors have not been captured here as mentioned earlier. Thus transferring model parameters to settings in which nutrient deficiencies are particularly acute requires extra care. We maintain that the use of mixed effects models is preferrable to OLS models, as the inclusion of random country effects helps to calibrate the estimation of responses for those countries. Moreover, we show that the use of mixed effects models leads to estimates showing more severe calorie reductions and larger ranges of uncertainty compared to the original OLS model. We contend that choosing the right model to fit the data has important implications for estimating climate responses on agricultural productivity. We have not given much attention to how these results compare to crop process models, however a recent review of process and statistical models found that the two methods generate similar crop yield sensitivities to changes in climate . Further, although individual predictions often deviate widely, the ensembled average of process models tends to correlate well with statistical model predictions .

Conclusion

In this study, we used statistical meta-analysis to fit mixed effects models of crop yield responses to climate change using an established database. We fit five model specifications and estimated future yield changes using temperature and precipitation projections to 2100 from a CMIP6 ensemble of 23 GCMs. We used these estimates to calculate national calorie supply and compared them to measures of calorie demand to understand their implications for global food security. Our preferred model estimates that global yields may change from 2015 levels by (maize), (rice), (soy) and (wheat) in under the business-as-usual scenario of SSP5-8.5, but reduce to and respectively under the lower emissions scenario of SSP1-2.6. We argue that this demonstrates substantial benefits from emissions mitigation. We quantified uncertainty across five dimensions and found that despite a wide range for projected soybean yield changes, our results offer overall validation of changing patterns for rice and wheat crop distributions under climate change. For some crop species and regions, such as the soybean-producing nations of US, China, Brazil and Argentina, uncertainty regarding future yield changes is so large that decision-makers should adopt a precautionary principle to invest in climate change adaptation in the agricultural sector. In the absence of local validation we believe these estimates provide a general expectation of future staple crop yield changes under climate change. Amidst considerable uncertainty, the signal is sufficiently clear on the need for national and intergovernmental policies to anticipate national food security risks. As empirical data on yields, changing crop calendars and cropping systems become available at higher resolutions and time frequencies, it is critical to validate impact estimates and update IAM projections used in policymaking . Systematic evaluations of uncertainty such as these will help food security policymakers and researchers understand the drivers of variation and biases in similar large-scale modelling exercises.

Methods

Objectives and scope

We performed statistical meta-analysis to understand potential climate impacts on crop yields and thus obtain estimates of potential calorie deficits under three emissions scenarios to the end of the century. To do this we fitted five model specifications including mixed effects models to account for correlation within and between studies, and quantified the relative magnitude of uncertainty attributable to model choice, sampling, missing data and GCMs from the CMIP6 ensemble. We converted projected yield change estimated by the preferred mixed model into measures of national calorie supply and compared them against measures of calorie demand to understand future impacts on food security.

Data, extraction and screening

CGIAR data includes 13,890 point estimates of yield change, temperature change, change, on-farm adaptation, and other variables extracted from 89 studies with varied methodologies that can be broadly categorised into either statistical regression models or crop biophysical process models . Reviews of the strengths and weaknesses of the two different methodologies show that the results are comparable when aggregated . We first reviewed and validated each of the 8846 estimates from 74 studies that are related to the four major crops maize, rice, soy, and wheat by reviewing the original studies (full list in Supplementary Information). We built a validation and cleaning database to manually record values identified from this review process as missing or
mis-entered, and replaced them with correct values from the studies. We also verified that any missing values recorded in the CGIAR dataset were true missing values (i.e., not zero values).

Baseline growing-season temperatures and precipitation

To estimate location-specific baseline growing-season temperatures, we drew on global monthly degree gridded temperature data from the Climatic Research Unit (CRU TS v4.05) from 1950 to . We then applied degree gridded crop calendar information on planting and harvesting dates in 2000 for the four crops from two complementary data sets circa 2000 on global crop planting dates and global monthly irrigated and rain-fed crop areas . Crop baseline growing-season temperature was calculated for each grid cell location by first identifying the growing-season months for each crop, and then calculating average annual historical growing-season temperatures for each year from 1950 to 2015, and then calculating average growing-season temperatures across all years in the baseline period specified for each CGIAR estimate (a subset of the 19502015 period). Finally, crop baseline growing-season temperatures were aggregated to the country level, by taking a crop production-weighted average of grid cell values in each country. To do this we used a degree gridded data set on crop production circa .

Data imputation

Of the 8846 point estimates in the CGIAR data, many values for key variables that were theoretically important for modelling crop yield responses to temperature change were missing. These included temperature change, precipitation change, change, yield change, and start and end years for the baseline period. We assume that missingness among the six key variables should be mostly missing at random, with possibly a few exceptions . Listwise deletion of observations with missing values is one option, but fitting a model on complete cases only may bias the results. We therefore implemented a formal imputation procedure using Multiple Imputation by Chained Equations (MICE) to fill in missing values for five variables: baseline average growing-season temperature and precipitation, temperature change, yield change, precipitation change and change. MICE treats missing data as an explicit source of variation to be averaged over by Markov chain Monte Carlo (MCMC) simulation based on a fully conditional specification of probability distributions . We adopted a multilevel Multiple Imputation approach that takes into account the clustered structure of our data . We set the number of imputed datasets to five, and 30 iterations to be run in parallel, based on visual inspection of convergence patterns. We found that two-level predictive mean matching simulated observed values best (Supp. Fig. 2).

Model fitting

We fit 5 candidate models for each crop separately with percentage yield change as the response variable: OLS model, two GLMM and two GAMM models each fit with random intercepts only (RI) and with random intercepts and slopes (RS) . These were estimated with Restricted Maximum Likelihood (REML) using the mgcv package in . We also tested models that weighted estimates by the inverse of variance in each study, as sometimes recommended in formal meta-analysis. However, study-wide variance can encompass meaningful variance across separate experiments (i.e. unique combinations of changes in temperature, precipitation and ), and it was impractical to separate unique combinations of all three continuous variables into discrete categories.
The most complex model specification (that includes random intercepts and slopes) allows temperature change, precipitation change and change to vary randomly for each study. Model specifications for the response variable with an identity link function are written for GLMM with random intercepts and slopes in Eq. (2) (estimated with orthogonal polynomial terms) and for GAMM with random intercepts and slopes in Eq. (3). The distribution function is assumed normal based on inspection of fitted residuals in Supp. Fig. 10. The pooled OLS model specification is included in Eq. (4), and differs from earlier literature in two ways: response functions are separately estimated for each of the four crops, and do not include an interaction term between the quadratic temperature change term and baseline temperature.
Included parameters: is the percentage change in yield from point estimate in country and study k. and are the changes in temperature (degrees), concentration (ppm), and precipitation (%) for point estimate and are mean growing-season temperature and precipitation for country over the baseline period specified for point estimate and are dummy variables indicating the crop photosynthetic metabolic pathway, and is a dummy variable indicating whether the point estimate includes on-farm adaptation (set to zero in all projections). As the response functions are estimated separately for each crop, is equal to 1 and equal to 0 for rice, soy and wheat, while is equal to 1 and equal to 0 for maize. Following , we assume a concave function to account for the diminishing marginal positive effect of of the form where for crops (i.e. rice, soy wheat) and for crops (i.e. maize). Terms : in the random slope terms represent the study-mean values. The term is the random country intercept, is the random study intercept, is the random temperature slope varying by country, is the random precipitation slope varying by country, is the random slope varying by country, and is the residual error. Parameters and in Eq. (3) are smooth functions of and respectively. Each smooth function is a spline which is the sum of smaller basis functions with coefficient evaluated at the values of .
We ran these five different model specifications on each of the five imputed datasets, yielding 25 combinations of fitted estimates per crop. We then fit these to 23 sets of prediction data containing all model terms, including temperature, precipitation and change variables calculated from the CMIP6 ensemble of 23 GCMs provided by WorldClim for four future time periods: 2021-2040, 2041-2060, 2061-2080, 2081-2100 .

Model prediction

We created prediction data for each of the model terms. Baseline growing-season temperatures and precipitation were estimated using the same approach used to calculate the variable in the main data set, but for 2015 (Supp. Fig. 4). Gridded temperature and precipitation values for 2021-2040, 2041-2060, 2061-2080 and 2081-2100 were retrieved from 23 WorldClim bias-corrected CMIP6 GCM datasets for SSP1-2.6, SSP2-4.5 and SSP5-8.5 at a spatial resolution . We estimated projected yields on each of the 23 GCM datasets individually. The 23 GCMs from the CMIP6 ensemble were: ACCESSCM2, ACCESS-ESM1-5, CanESM5, CanESM5-CanOE, CMCC-ESM2, CNRM-CM6-1, CNRM-CM6-1-HR, CNRM-ESM2-1, EC-Earth3-Veg, EC-Earth3-Veg-LR, FIO-ESM-2-0, GISS-E2-1-G, GISS-E2-1-H, HadGEM3-GC31-LL, INM-CM4-8, INM-CM5-0, IPSL-CM6ALR, MIROC-ES2L, MIROC6, MPI-ESM1-2-HR, MPI-ESM1-2-LR, MRI-ESM2-0 and UKESM1-0-LL. The same 23 GCMs was used for scenario SSP1-2.6 but with the addition of GFDL-ESM4. We used the bioclimatic variables for annual mean temperature (BIO1) and annual precipitation (BIO12) converted to average monthly precipitation.
For global change, we used data on annual global concentrations calculated from the multi-model CMIP5 ensemble mean of 39 GCMs retrieved from Climate Explorer , converted to changes in concentrations from 2015 levels. Historical levels were drawn from global observations .
We predicted global gridded yield impacts across each of the four time periods using 5 imputed fit estimates model specifications crops (i.e. 100 model combinations), and on each of the 23 GCM outputs (i.e. 575 prediction combinations per crop, or 2300 prediction combinations in total). To obtain pooled results in Fig. 4, Supp. Figs. 5, 6 and 9, we pooled prediction fit and standard error estimates sequentially across each crop model specification’s imputation-specific model predictions and GCM-specific model predictions. We thus followed a ‘pool last’ approach that applies Rubin’s rules to the final estimate of interest, as doing so maintains betweenimputation variability and accounts for non-linearities in their proper place .
Modelled response functions and predicted fit estimates do not necessarily go through the origin (meaning a non-zero intercept exists), yet theory requires zero yield change with zero temperature change . We performed post-hoc adjustments by estimating projected yields twice; once as described above (i.e., original) and once on adjusted prediction data where temperature, precipitation and changes are set equal to zero. We then subtracted the original predictions from yield changes predicted on zero change in climate variables.

Evaluating uncertainty

We quantified four different sources of uncertainty arising from model choice, sampling, missing data and CMIP6 multi-model variation across GCMs, following the approach in Ref. . We fit and predicted on 100 block bootstrap samples per combination of model choice, crop, GCM and imputation-augmented dataset, yielding 230,000 predicted estimates of global production-weighted yield change in 2081-2100 (i.e. estimates calculated as global mean yield changes in percentage terms, weighted by gridded crop production volumes c. ). Bootstraps were blocked by study, meaning that a random set of studies and estimates therein are drawn in each bootstrap sample. We then estimated the crop production-weighted global mean yield change, yielding 575 estimates representing unique combinations of block bootstrap sample, imputation-augmented dataset, model choice and GCMs from the CMIP6 ensemble.
We characterised uncertainty from model choice by holding constant the imputation-augmented dataset, bootstrap sample and GCM output data used, varying the model specification and calculating the average standard deviation in global weighted yield change across model specifications. We characterised the other sources of uncertainty in the same way.

Food calorie gap

We estimated calorie supply and calorie demand using FAO data on country-scale production, import and export quantities and the share of crops allocated to food, average daily energy requirements, calories of energy supplied to the food system by crop and country, and detailed trade patterns (see “Methods” section) . We also used crop calorie conversion factors from the literature , gridded yield data and hectares harvested c . 2015 from a new high resolution global spatial dataset and population projections for SSP1-2.6, SSP2-4.5 and SSP5-8.5 developed by the International Institute for Applied Systems Analysis (IIASA) .
To estimate baseline production in the year 2015 of maize, rice, soy and wheat in tons by country, we multiplied gridded hectares harvested (ha) by gridded yields (t/ha) in a spatial dataset with 5 -min resolution . We estimated the share of each country’s domestic supply of each crop allocated to the food system and the share of each country’s production of each crop that is exported, averaged across 2014-2016 using data from the FAO Food Balance Sheets . We estimated the share of each country’s exports of each crop that are exported to each trade partner, averaged across 2014-2016 using data from the FAO Detailed Trade Matrix . From this calculation we recovered each country’s imports of each crop by summing exports across importing partners. We used the FAO’s concordance between food products in the Detailed Trade Matrix and Food Balance Sheets (Supplementary Table 2). To estimate the domestic supply of maize, rice, soy and wheat in calories we applied calorie conversion factors for each crop from the literature to our estimate of domestic supply from production and imports in tons . To estimate the total domestic supply of calories from all crops for each country, we scaled the four staple crop calories by their share of total calories supplied by all crops per country . To estimate future production, we applied the gridded projected yield percentage changes from the GLMM random slopes model to baseline gridded yields and hectares harvested as described above. We repeated each of the steps to estimate total domestic supply of calories from all crops in each future time period.
These steps assume that the export share of country production and export partner share remain fixed in the future. We use data on harvested area data c. , which assumes that harvest frequency and total standing cropland area remain fixed. These make our results conservative, as historical trends show that harvested land has increased in the last few decades . Our results are also conservative due to the narrower selection of staple crop products represented in the gridded yields and hectares harvested dataset and thus in our estimates (Supplementary Table 2).
To estimate baseline calorie demand in the year 2015 by country, we multiplied annual per capita average daily energy requirements (ADER) averaged over 2014-2016 by population figures from FAOSTAT . To estimate future calorie demand taking into account population growth but keeping dietary patterns fixed, we multiplied annual per capita ADER by SSP projections of country-level population counts from IIASA averaged across each of the 20 -year future periods . We calculated the calorie gap for each country as the supply of total calories subtracted from total calories demanded. A positive calorie gap indicates that a country is unable to meet national calorie demand from calories supplied by domestic production and imports. As we did this for the baseline year of 2015 and for each of the four 20-year periods from 2021 to 2100, we were able to determine the change in calorie gap from 2015 to future time periods.

Data availability

Data and code used for this analysis are provided at https://github.com/christineklli/global-yield-impacts. Mod elled country yield impacts are provided as Supplementary Data.
Received: 22 May 2023; Accepted: 15 January 2025
Published online: 22 January 2025

References

  1. Roson, R. & Sartori, M. Estimation of climate change damage functions for 140 regions in the GTAP 9 data base. J. Glob. Econ. Anal. 1(2), 78-115 (2016).
  2. Newell, R. G., Prest, B. C. & Sexton, S. E. The GDP-temperature relationship: Implications for climate change damages. J. Environ. Econ. Manag. 108, 102445 (2021).
  3. Liu, B. et al. Similar estimates of temperature impacts on global wheat yield by three independent methods. Nat. Clim. Change 6, 1130-1136 (2016).
  4. Challinor, A. J. et al. A meta-analysis of crop yield under climate change and adaptation. Nat. Clim. Change 4, 287-291 (2014).
  5. Kopp, R., Hsiang, S. & Oppenheimer, M. Empirically calibrating damage functions and considering stochasticity when integrated assessment models are used as decision tools. Impacts World 2013, International Conference on Climate Change Effects 12 (2013).
  6. Tol, R. S. J. The economic effects of climate change. J. Econ. Perspect. 23, 29-51 (2009).
  7. Tol, R. S. J. Correction and update: The economic effects of climate change. J. Econ. Perspect. 28, 221-226 (2014).
  8. Wang, X. et al. Emergent constraint on crop yield response to warmer temperature from field experiments. Nat. Sustain. 3, 908-916 (2020).
  9. Whetton, P. H., Grose, M. R. & Hennessy, K. J. A short history of the future: Australian climate projections 1987-2015. Clim. Serv. 2-3, 1-14 (2016).
  10. Nelson, J. P. & Kennedy, P. E. The use (and abuse) of meta-analysis in environmental and natural resource economics: An assessment. Environ. Resour. Econ. 42, 345-377 (2009).
  11. Howard, P. H. & Sterner, T. Few and not so far between: A meta-analysis of climate damage estimates. Environ. Resour. Econ. 68, 197-225 (2017).
  12. Moore, F. C., Baldos, U. L. C., Hertel, T. W. & Diaz, D. New science of climate change impacts on agriculture implies higher social cost of carbon. Nat. Commun. 9, 1 (2017).
  13. CGIAR. Agriculture Impacts. https://web.archive.org/web/20211204202518/; http://www.ag-impacts.org/ (2021).
  14. FAOSTAT. Food Balance Sheets. https://www.fao.org/faostat/en/#data/FBS (2023).
  15. Bell, A., Fairbrother, M. & Jones, K. Fixed and random effects models: Making an informed choice. Qual. Quant. 53, 1051-1074 (2019).
  16. Asseng, S., Foster, I. & Turner, N. C. The impact of temperature variability on wheat yields. Glob. Change Biol. 17, 997-1012 (2011).
  17. Porter, J. R. & Semenov, M. A. Crop responses to climatic variation. Philos. Trans. R. Soc. B Biol. Sci. 360, 2021-2035 (2005).
  18. Semenov, M. & Porter, J. Climatic variability and the modelling of crop yields. Agric. For. Meteorol. 73, 265-283 (1995).
  19. Moriondo, M., Giannakopoulos, C. & Bindi, M. Climate change impact assessment: The role of climate extremes in crop yield simulation. Clim. Change 104, 679-701 (2011).
  20. Springer, C. J. & Ward, J. K. Flowering time and elevated atmospheric CO 2. New Phytol. 176, 243-255 (2007).
  21. Wheeler, T. R., Craufurd, P. Q., Ellis, R. H., Porter, J. R. & Vara Prasad, P. Temperature variability and the yield of annual crops. Agric. Ecosyst. Environ. 82, 159-167 (2000).
  22. Rosenzweig, C., Iglesius, A., Yang, X. B., Epstein, P. & Chivian, E. Climate Change and Extreme Weather Events-Implications for Food Production, Plant Diseases, and Pests (NASA Publications, 2001).
  23. Hatfield, J. L. & Prueger, J. H. Temperature extremes: Effect on plant growth and development. Weather Clim. Extremes 10, 4-10 (2015).
  24. Craufurd, P. Q. & Wheeler, T. R. Climate change and the flowering time of annual crops. J. Exp. Bot. 60, 2529-2539 (2009).
  25. Lobell, D. B., Sibley, A. & Ivan Ortiz-Monasterio, J. Extreme heat effects on wheat senescence in India. Nat. Clim. Change 2, 186-189 (2012).
  26. Lobell, D. B., Baldos, U. L. C. & Hertel, T. W. Climate adaptation as mitigation: The case of agricultural investments. Environ. Res. Lett. 8, 015012 (2013).
  27. Masson-Delmotte, V. et al. IPCC, 2021: Summary for policymakers. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (2021).
  28. Riahi, K. et al. The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: An overview. Glob. Environ. Change 42, 153-168 (2017).
  29. FAOSTAT. Detailed Trade Matrix. https://www.fao.org/faostat/en/#data/TM (2023).
  30. Cassidy, E. S., West, P. C., Gerber, J. S. & Foley, J. A. Redefining agricultural yields: From tonnes to people nourished per hectare. Environ. Res. Lett. 8, 034015 (2013).
  31. Grogan, D., Frolking, S., Wisser, D., Prusevich, A. & Glidden, S. Global gridded crop harvested area, production, yield, and monthly physical area data circa 2015. Sci. Data 9, 15 (2022).
  32. Kc, S. & Lutz, W. The human core of the shared socioeconomic pathways: Population scenarios by age, sex and level of education for all countries to 2100. Glob. Environ. Change 42, 181-192 (2017).
  33. Tilman, D., Balzer, C., Hill, J. & Befort, B. L. Global food demand and the sustainable intensification of agriculture. Proc. Natl. Acad. Sci. 108, 20260-20264 (2011).
  34. Ivanovich, C. C., Sun, T., Gordon, D. R. & Ocko, I. B. Future warming from global food consumption. Nat. Clim. Change 1, 1-6 (2023).
  35. Ray, D. K. & Foley, J. A. Increasing global crop harvest frequency: Recent trends and future directions. Environ. Res. Lett. 8, 044041 (2013).
  36. Fahad, S. et al. Crop production under drought and heat stress: Plant responses and management options. Front. Plant Sci. 8, 1 (2017).
  37. Daryanto, S., Wang, L. & Jacinthe, P.-A. Global synthesis of drought effects on maize and wheat production. PLoS ONE 11, e0156362 (2016).
  38. Marothia, D. et al. Abiotic Stress in Plants (IntechOpen, 2020).
  39. Costa, M. V. J. D. et al. Combined drought and heat stress in rice: Responses, phenotyping and strategies to improve tolerance. Rice Sci. 28, 233-242 (2021).
  40. Hussain, H. A. et al. Interactive effects of drought and heat stresses on morpho-physiological attributes, yield, nutrient uptake and oxidative status in maize hybrids. Sci. Rep. 9, 3890 (2019).
  41. Jumrani, K. & Bhatia, V. S. Interactive effect of temperature and water stress on physiological and biochemical processes in soybean. Physiol. Mol. Biol. Plants 25, 667-681 (2019).
  42. Ostmeyer, T. et al. Impacts of heat, drought, and their interaction with nutrients on physiology, grain yield, and quality in field crops. Plant Physiol. Rep. 25, 549-568 (2020).
  43. Iqbal, M. M., Arif, M. & Khan, A. M. Climate Change Aspersions on Food Security of Pakistan 11 (2009).
  44. Lobell, D. B. & Burke, M. B. Why are agricultural impacts of climate change so uncertain? The importance of temperature relative to precipitation. Environ. Res. Lett. 3, 034007 (2008).
  45. Hausfather, Z. Explainer: What climate models tell us about future rainfall. Carbon Brief. https://www.carbonbrief.org/explaine r-what-climate-models-tell-us-about-future-rainfall/ (2022).
  46. Deryng, D., Conway, D., Ramankutty, N., Price, J. & Warren, R. Global crop yield response to extreme heat stress under multiple climate change futures. Environ. Res. Lett. 9, 034011 (2014).
  47. Hatfield, J. L. et al. Climate impacts on agriculture: Implications for crop production. Agron. J. 103, 351-370 (2011).
  48. FAOSTAT. Crops and Livestock Products. https://www.fao.org/faostat/en/#data/QCL (2023).
  49. Fahad, S. et al. Consequences of high temperature under changing climate optima for rice pollen characteristics-concepts and perspectives. Arch. Agron. Soil Sci. 64, 1473-1488 (2018).
  50. Fahad, S. et al. A combined application of biochar and phosphorus alleviates heat-induced adversities on physiological, agronomical and quality attributes of rice. Plant Physiol. Biochem. 103, 191-198 (2016).
  51. Bradford, J. B. et al. Future soil moisture and temperature extremes imply expanding suitability for rainfed agriculture in temperate drylands. Sci. Rep. 7, 12923 (2017).
  52. Tuninetti, M., Ridolfi, L. & Laio, F. Ever-increasing agricultural land and water productivity: A global multi-crop analysis. Environ. Res. Lett. 15, 09402 (2020).
  53. Gerten, D. et al. Global water availability and requirements for future food production. J. Hydrometeorol. 12, 885-899 (2011).
  54. Billen, G. et al. Reshaping the European agro-food system and closing its nitrogen cycle: The potential of combining dietary change, agroecology, and circularity. One Earth 4, 839-850 (2021).
  55. World-Bank. Water in Agriculture World Bank. https://www.worldbank.org/en/topic/water-in-agriculture (2023).
  56. Elliott, J. et al. Constraints and potentials of future irrigation water availability on agricultural production under climate change. Proc. Natl. Acad. Sci. 111, 3239-3244 (2014).
  57. Rosa, L. et al. Potential for sustainable irrigation expansion in a C warmer climate. Proc. Natl. Acad. Sci. 117, 29526-29534 (2020).
  58. Wisser, D. et al. Global irrigation water demand: Variability and uncertainties arising from agricultural and climate data sets. Geophys. Res. Lett. 35, 1 (2008).
  59. Ortiz-Bobea, A., Ault, T. R., Carrillo, C. M., Chambers, R. G. & Lobell, D. B. Anthropogenic climate change has slowed global agricultural productivity growth. Nat. Clim. Change 11, 306-312 (2021).
  60. WRI. Resource Watch Aqueduct Stress Projections. https://resourcewatch.org/data/explore (2023).
  61. Luck, M., Landis, M. & Gassert, F. Aqueduct Water Stress Projections: Decadal Projections of Water Supply and Demand Using CMIP5 GCMs 20 (2015).
  62. Puy, A., LoPiano, S. & Saltelli, A. Current models underestimate future irrigated areas. Geophys. Res. Lett. 47, e2020087360 (2020).
  63. Vanschoenwinkel, J. & Van Passel, S. Climate response of rainfed versus irrigated farms: The bias of farm heterogeneity in irrigation. Clim. Change 147, 225-234 (2018).
  64. Kompas, T., Che, T. N. & Grafton, R. Q. The Impact of Water and Heat Stress from Global Warming on Agricultural Productivity and Food Security (Global Commission on the Economics of Water, 2023).
  65. Roy, S., Chowdhury, N., Roy, S. & Chowdhury, N. Abiotic Stress in Plants (IntechOpen, 2020).
  66. Uçarlı, C. Abiotic Stress in Plants (IntechOpen, 2020).
  67. Fahad, S. et al. Phytohormones and plant responses to salinity stress: A review. Plant Growth Regul. 75, 391-404 (2015).
  68. Lal, M. Implications of climate change in sustained agricultural productivity in South Asia. Reg. Environ. Change 11, 79-94 (2011).
  69. Soares, J. C., Santos, C. S., Carvalho, S. M. P., Pintado, M. M. & Vasconcelos, M. W. Preserving the nutritional quality of crop plants under a changing climate: Importance and strategies. Plant Soil 443, 1-26 (2019).
  70. De Vries, W., Kros, J., Kroeze, C. & Seitzinger, S. P. Assessing planetary and regional nitrogen boundaries related to food security and adverse environmental impacts. Curr. Opin. Environ. Sustain. 5, 392-402 (2013).
  71. MacDonald, G. K., Bennett, E. M., Potter, P. A. & Ramankutty, N. Agronomic phosphorus imbalances across the world’s croplands. Proc. Natl. Acad. Sci. 108, 3086-3091 (2011).
  72. Zhang, C. et al. The role of nitrogen management in achieving global sustainable development goals. Resour. Conserv. Recycl. 201, 107304 (2024).
  73. Gu, B. et al. Cost-effective mitigation of nitrogen pollution from global croplands. Nature 613, 77-84 (2023).
  74. Qiao, L. et al. Assessing the contribution of nitrogen fertilizer and soil quality to yield gaps: A study for irrigated and rainfed maize in China. Field Crops Res. 273, 108304 (2021).
  75. Vanlauwe, B. & Dobermann, A. Sustainable intensification of agriculture in sub-Saharan Africa: First things first. Front. Agric. Sci. Eng. 7, 376-382 (2020).
  76. Cassman, K. G. & Dobermann, A. Nitrogen and the future of agriculture: 20 years on. Ambio 51, 17-24 (2022).
  77. Zhang, X. et al. Managing nitrogen for sustainable development. Nature 528, 51-59 (2015).
  78. Zhang, X. et al. Quantifying nutrient budgets for sustainable nutrient management. Glob. Biogeochem. Cycles 34, e2018006060 (2020).
  79. Juroszek, P. & von Tiedemann, A. Linking plant disease models to climate change scenarios to project future risks of crop diseases: A review. J. Plant Dis. Prot. 122, 3-15 (2015).
  80. Bradshaw, C. et al. Climate change in pest risk assessment: Interpretation and communication of uncertainties. EPPO Bull. 54, 4-19 (2024).
  81. Giorgi, F. Thirty years of regional climate modeling: Where are we and where are we going next? J. Geophys. Res. Atmos. 124, 5696-5723 (2019).
  82. Hawkins, E. & Sutton, R. The potential to narrow uncertainty in regional climate predictions. Bull. Am. Meteorol. Soc. 90, 10951108 (2009).
  83. Czajkowski, M. & ŠčasnÃý, M. Study on benefit transfer in an international setting. How to improve welfare estimates in the case of the countries’ income heterogeneity? Ecol. Econ. 69, 2409-2416 (2010).
  84. Muleke, A., Harrison, M. T., Yanotti, M. & Battaglia, M. Yield gains of irrigated crops in Australia have stalled: The dire need for adaptation to increasingly volatile weather and market conditions. Curr. Res. Environ. Sustain. 4, 100192 (2022).
  85. Ciscar, J.-C., Fisher-Vanden, K. & Lobell, D. B. Synthesis and review: An inter-method comparison of climate change impacts on agriculture. Environ. Res. Lett. 13, 070401 (2018).
  86. Lobell, D. B., Thau, D., Seifert, C., Engle, E. & Little, B. A scalable satellite-based crop yield mapper. Remote Sens. Environ. 164, 324-333 (2015).
  87. Roberts, M. J., Braun, N. O., Sinclair, T. R., Lobell, D. B. & Schlenker, W. Comparing and combining process-based crop models and statistical models with some implications for climate change. Environ. Res. Lett. 12, 095010 (2017).
  88. Moore, F. C., Baldos, U. L. C. & Hertel, T. Economic impacts of climate change on agriculture: A comparison of process-based and statistical yield models. Environ. Res. Lett. 12, 065008 (2017).
  89. Harris, I., Osborn, T. J., Jones, P. & Lister, D. Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data 7, 109 (2020).
  90. Sacks, W. J., Deryng, D., Foley, J. A. & Ramankutty, N. Crop planting dates: An analysis of global patterns. Glob. Ecol. Biogeogr. 19, 607-620 (2010).
  91. Portmann, F. T., Siebert, S. & Döll, P. MIRCA2000-Global monthly irrigated and rainfed crop areas around the year 2000: A new high-resolution data set for agricultural and hydrological modeling. Glob. Biogeochem. Cycles 24, 1 (2010).
  92. Monfreda, C., Ramankutty, N. & Foley, J. A. Farming the planet: 2. Geographic distribution of crop areas, yields, physiological types, and net primary production in the year 2000. Glob. Biogeochem. Cycles 22, 1 (2008).
  93. Graham, J. W. Missing data analysis: Making it work in the real world. Annu. Rev. Psychol. 60, 549-576 (2009).
  94. Schafer, J. L. & Graham, J. W. Missing data: Our view of the state of the art. Missing Data 31, 1 (2002).
  95. Van Buuren, S. & Groothuis-Oudshoorn, K. Mice: Multivariate imputation by chained equations in R. J. Stat. Softw. 67, 1 (2011).
  96. Yucel, R. M. Multiple imputation inference for multivariate multilevel continuous data with ignorable non-response. Philos. Trans. Ser. A Math. Phys. Eng. Sci. 366, 2389-2403 (2008).
  97. Heymans, M. W. & Eekhout, I. Applied Missing Data Analysis with SPSS and (R)Studio (2019).
  98. Schomaker, M. & Heumann, C. Model selection and model averaging after multiple imputation. Comput. Stat. Data Anal. 71, 758-770 (2014).
  99. Wood, S. N. Generalized Additive Models: An introduction with R 397 (2017).
  100. Harrer, M., Cuijpers, P., Furukawa, T. A. & Ebert, D. D. Doing Meta-analysis in R (2021).
  101. WorldClim. Future Climate, 10 Minutes Spatial Resolution—WorldClim 1 Documentation. https://www.worldclim.org/data/cmip 6/cmip6_clim10m.html (2023).
  102. KNMI, C.E. Monthly CMIP5 Scenario Runs. https://climexp.knmi.nl/start.cgi (2023).
  103. US Department of Commerce, N. Global Monitoring Laboratory—Carbon Cycle Greenhouse Gases. https://gml.noaa.gov/ccgg/tre nds/data.html (2023).
  104. White, I. R., Royston, P. & Wood, A. M. Multiple imputation using chained equations: Issues and guidance for practice. Stat. Med. 30, 377-399 (2011).
  105. Rubin, D. B. & Schenker, N. Multiple Imputation for Interval Estimation from Simple Random Samples with Ignorable Nonresponse 10 (1986).
  106. Fischer, E. M., Sedláěek, J., Hawkins, E. & Knutti, R. Models agree on forced response pattern of precipitation and temperature extremes. Geophys. Res. Lett. 41, 8554-8562 (2014).
  107. Wilcox, J. & Makowski, D. A meta-analysis of the predicted effects of climate change on wheat yields using simulation studies. Field Crop Res. 156, 180-190 (2014).

Acknowledgements

C.L. was funded by the Centre of Excellence for Biosecurity Risk Analysis. We thank J. Baumgartner for providing programming advice.

Author contributions

A.R., C.L., J.C., and T.K. conceived the analysis, C.L. conducted the analysis and wrote the original draft. All authors reviewed and edited the manuscript.

Declarations

Competing interests

The authors declare no competing interests.

Additional information

Supplementary Information The online version contains supplementary material available at https://doi.org/1 .
Correspondence and requests for materials should be addressed to C.L.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommo ns.org/licenses/by-nc-nd/4.0/.
© The Author(s) 2025

  1. School of BioSciences, Centre of Excellence for Biosecurity Risk Analysis, University of Melbourne, Melbourne 3010, Australia. School of Mathematics and Statistics, University of Melbourne, Melbourne 3010, Australia. School of Agriculture, Food and Ecosystem Sciences, University of Melbourne, Melbourne 3010, Australia. email: christinel3@unimelb.edu.au