Compound and cascading droughts and heatwaves decrease maize yields by nearly half in Sinaloa, Mexico

Climate change intensifies extreme weather events like droughts and heatwaves, significantly impacting global food security and prices. These events, individually and concurrently, severely affect crop growth and yield. Even mild events during vegetative stages can drastically reduce productivity and farmer income. Previous research shows maize yield losses of up to 40% due to water scarcity. Compound events (simultaneous drought and heatwave) exacerbate yield reductions compared to single events. While many studies use statistical methods or machine learning to assess the impact of compound droughts and heatwaves, few utilize process-based models to consider single, compound, and cascading events throughout the growing period. This study uses the WOFOST crop growth simulation model and ERA5 meteorological data (1990-2022) to evaluate these impacts on maize yields in Sinaloa, Mexico – a major maize-producing region historically vulnerable to these events. The choice of Sinaloa allows for detailed analysis incorporating soil data, crop information, and agricultural practices to understand real-world impacts. The Standardized Precipitation Index (SPI-3) identifies droughts, while a threshold method identifies heatwaves. The study compares maize yield estimations under normal, drought, heatwave, compound (drought and heatwave simultaneously), and cascading (drought followed by heatwave) conditions. Pearson correlation assesses the relationship between hazards and yields. Although a case study, the methodology and results offer broader applicability.

Existing literature extensively documents the individual impacts of drought and heatwaves on crop yields, particularly maize. Studies show significant yield reductions (up to 40%) due to water scarcity and drought. The compounding effect of simultaneous drought and heatwave is also recognized, leading to greater losses than individual events. However, there's a gap in research systematically analyzing the impacts of both single and compound/cascading events on maize using process-based crop models. Many studies employ statistical methods or machine learning, neglecting the temporal dynamics and combined effects of these events throughout the growing season. Some research highlights the increasing frequency of compound droughts and heatwaves globally, but region-specific studies, particularly for Mexico, are limited. This research addresses this gap by focusing on Sinaloa, Mexico, a region with significant historical exposure to these extreme events, enabling a more comprehensive understanding of their impacts on maize production.

This study utilized the WOFOST crop growth simulation model to estimate maize yields in Sinaloa, Mexico, from 1990 to 2022. The model was driven by ERA5 reanalysis meteorological data, incorporating soil data, crop information, and agricultural practices to enhance realism. Droughts were identified using the Standardized Precipitation Index (SPI-3) with a 3-month accumulation period, categorizing events as moderate, severe, or extreme based on SPI-3 values. Heatwaves were defined as periods where daily maximum temperatures exceeded the 90th percentile for at least three consecutive days, using a 31-day moving average to smooth thresholds. Compound events (CDH) were defined as simultaneous occurrence of drought and heatwave in the same region, while cascading events (CaDH) were defined as drought followed by heatwave consecutively. The analysis focused on events occurring within the maize growing season (November to May). The WOFOST model simulated yields under normal, drought, heatwave, compound, and cascading conditions. Pearson correlation coefficients assessed the relationships between the lowest SPI-3 value (representing the most severe drought during the growing season) and annual maize yields, and between the number of heatwave events and annual yields. The model considered consistent parameters each year (planting date, thermal time, fertilizer application, soil properties), isolating the impact of weather conditions on yield variation.

The analysis identified 27 distinct drought events and 75 heatwave events in Sinaloa from 1990 to 2022. Of these, 15 drought and 9 heatwave events occurred during the maize growing season. Extreme drought events reduced maize yields by 25% compared to normal conditions. Heatwaves alone had a negligible effect. However, compound and cascading drought-heatwave events significantly amplified yield reductions. The three most severe yield reductions occurred during the extreme drought of 2010, the cascading event of 1998, and the compound event of 1999. A strong positive correlation (r=0.77) was found between maize yields and SPI-3, indicating a significant impact of drought on yields. The correlation between heatwaves and yields was weak. The WOFOST model simulated lower yields during drought events, with the lowest yields (2.4-2.8 t/ha) occurring during the extreme drought of 2010 and the compound/cascading events of 1998 and 1999. Yields during compound events dropped significantly to 2.7 t/ha, and cascading events resulted in yields of 2.8 t/ha. Despite the occurrence of heatwaves, yields remained comparable or even higher than normal during some years, indicating maize's relative resilience to heat under sufficient water conditions. The study's findings are consistent with historical records of severe agricultural impacts in Sinaloa during the 1990s and 2010-2012, which were also identified as periods of severe drought in the analysis. The critical periods during 1998-1999 and 2010-2012, marked by recurrent adverse impacts on agriculture and significantly reduced staple grain production (up to 50%) align with the model's simulations of low yields. The impact of drought on yield was attributed to water stress, as maize's assimilation capacity is only marginally affected by high temperatures. The timing of drought events within the maize growth cycle (vegetative vs. reproductive stage) also influenced yield losses.

The study's findings highlight the significant impact of compound and cascading drought and heatwave events on maize yields in Sinaloa, exceeding the impact of single events. The strong positive correlation between SPI-3 and maize yields underscores the vulnerability of maize production to drought. The relatively weak correlation with heatwaves suggests that water scarcity is a more significant factor than heat stress under the conditions examined. The results emphasize the limitations of focusing solely on single extreme events, as compound and cascading events can lead to much more severe agricultural losses. The findings are relevant to other agricultural regions susceptible to similar compound climate hazards, emphasizing the need for climate-resilient agricultural practices. The severe yield reductions observed during compound and cascading events necessitate the development and implementation of adaptive strategies to mitigate food insecurity in vulnerable areas. The study demonstrates the effectiveness of process-based models like WOFOST in assessing the combined impacts of multiple climate hazards on crop production. This is crucial for informing evidence-based decision-making in agricultural planning and adaptation strategies.

This study demonstrates that compound and cascading drought and heatwave events have a significantly greater negative impact on maize yields in Sinaloa, Mexico, than single events. The use of the WOFOST model provides valuable insights into the combined effects of these hazards. Future research could explore other regions and crops to assess the generalizability of these findings and further investigate the interaction between drought and heat stress at various growth stages. The development of climate-resilient agricultural strategies is vital to ensure food security in the face of increasing climate variability.

This study focuses solely on Sinaloa, Mexico, limiting the generalizability of the findings to other regions with different climatic conditions, soil types, and agricultural practices. The study uses the WOFOST model which, while sophisticated, is still a simplification of complex ecological processes. The accuracy of the model's predictions depends on the accuracy of input data and parameterizations. Future research could investigate the potential for improvement by incorporating more fine-scale climate information and local-specific agronomic data.