Kaium Md Abdul, Ahmed Md Sharif, Habib-Ur-Rahman Muhammad, Islam Md Saidul, Ratry Yeasmin Akter, Helal Md Mostofa Uddin, Siddquy Muhammad Ali Fardoush, Haque Most Moslema, Raza Ahsan, Mansour Fatma, Alotaibi Majed, El Sabagh Ayman, Roetter Reimund P
Department of Crop Science and Technology, University of Rajshahi, Rajshahi, 6205, Bangladesh.
Institute of Natural Resources Research and Development, Rajshahi, 6206, Bangladesh.
Sci Rep. 2025 Jul 16;15(1):25882. doi: 10.1038/s41598-025-09820-3.
Climate change is causing more frequent and extraordinary extreme weather events that are already negatively affecting crop production. There is a need for improved climate risk assessment by developing smart adaptation strategies for sustainable future crop production. This study aims to assess yield impacts of extreme temperatures and rainfall variability on wheat, and winter and summer season-planted maize in northwestern Bangladesh. Utilizing a machine learning approach, future yield patterns were predicted for these crops under various climate change scenarios. Additionally, the study developed adaptation strategies focusing on prediction of optimum sowing windows for wheat and maize to minimize climate risk-related yield losses jeopardizing food security. A fuzzy logical model was applied, incorporating a set of fuzzy rules to estimate the probable yields of wheat and maize (winter and summer growing seasons). Key climatic variables (temperature and rainfall) were added as model inputs, enabling the model to handle uncertainty and nonlinear interactions in the climate-yield relationship. Findings demonstrated that climate change has significant negative impacts at the different phenological stages of both wheat and maize (winter and summer seasons), with yield levels generally showing notable declines. Only small variations in optimal temperature and rainfall patterns affected crop yields significantly. Moreover, maize summer yield was consistently lower than maize winter as the temperature prevails high during the maize summer season (April to July). The study found that the wheat crop, maize winter, and maize summer have as optimal planting windows November 1-7, November 1-10, and February 20 - March 7, respectively. Such adaptation would ensure maximum yield and effective reduction of climate change risks. Outcomes of this study contribute to multiple Sustainable Development Goals (SDGs), especially three; zero hunger (SDG2), climate action (SDG13), and life on land (SDG14). These adaptations identified in this study can support policymakers and stakeholders to combat the impact of extreme climate - and achieving optimal yield. The approach is also applicable to other regions of the country and similar monsoon climates.
气候变化正导致更频繁和异常的极端天气事件，这些事件已经对作物生产产生了负面影响。有必要通过制定智能适应策略来改善气候风险评估，以实现未来作物的可持续生产。本研究旨在评估极端温度和降雨变化对孟加拉国西北部小麦、冬播玉米和夏播玉米产量的影响。利用机器学习方法，预测了这些作物在各种气候变化情景下的未来产量模式。此外，该研究制定了适应策略，重点是预测小麦和玉米的最佳播种窗口，以尽量减少危及粮食安全的与气候风险相关的产量损失。应用了模糊逻辑模型，纳入了一组模糊规则来估计小麦和玉米（冬夏生长季）的可能产量。关键气候变量（温度和降雨）作为模型输入，使模型能够处理气候-产量关系中的不确定性和非线性相互作用。研究结果表明，气候变化对小麦和玉米（冬夏季节）的不同物候阶段都有显著负面影响，产量水平普遍显著下降。最佳温度和降雨模式的微小变化都会对作物产量产生重大影响。此外，由于玉米夏季（4月至7月）温度普遍较高，玉米夏季产量一直低于冬季产量。研究发现，小麦、冬播玉米和夏播玉米的最佳种植窗口分别为11月1日至7日、11月1日至10日和2月20日至3月7日。这种适应措施将确保产量最大化并有效降低气候变化风险。本研究的结果有助于实现多个可持续发展目标（SDG），特别是三个目标：零饥饿（SDG2）、气候行动（SDG13）和陆地生物（SDG14）。本研究确定的这些适应措施可以支持政策制定者和利益相关者应对极端气候的影响并实现最佳产量。该方法也适用于该国其他地区和类似的季风气候。
