Assessing the impact of climate variability on maize yields in the different regions of Ghana—A machine learning perspective

temperature of 27.53˚C and a precipitation total of 1262.70 mm. This region clearly benefits from higher precipitation, as it is a significant factor for achieving higher yields. In the Upper East region (Fig 3i), maize yield increases with average temperatures between 28.98˚C and 29.03˚C and precipitation amounts ranging from 1124.53 mm to 1142.28 mm. The optimal maize yield is observed when the average temperature is 28.98˚C, and the precipitation mea- sures 1142.28 mm. Similarly, in the Upper West region (Fig 3j), maize yield increases with average temperatures ranging from 28.68˚C to 28.92˚C and precipitation between 1168.38 mm and 1242.41 mm. The optimal maize yield in the region is achieved at an average tempera- ture of 28.68˚C and a precipitation total of 1242.41 mm. In the Northern region (Fig 3h), a maximum maize yield is produced with average temperatures ranging from 28.58˚C to 28.79˚C and precipitation ranging from 1030.30 mm to 1166.60 mm. The data in Fig 3h clearly Fig 2. Trend of maize yield in each of the regions of Ghana. https://doi.org/10.1371/journal.pone.0305762.g002 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 14 / 24

illustrates the beneficial effect of higher average temperatures on maize yield in this region. In the Western region (Fig 3g), maximum maize yields are obtained when average temperatures range from 26.72˚C to 27.33˚C and precipitation amounts fall between 1223.53 mm and 1663.42 mm. The optimal combination occurs at an average temperature of 26.92˚C and a pre- cipitation total of 1223.53 mm, resulting in the maximum maize yield. Based on the findings from Fig 3, it can be concluded that increased precipitation is benefi- cial for maize yield in Ghana. Additionally, the optimal range of average temperatures for maize yield varies depending on the region, ranging from 26.00˚C to 28.00˚C. Conversely, lower average temperatures and less precipitation do not contribute to the production of a good maize yield, as evidenced by the data presented in these figures. Fig 3. Combined impact of average temperature and precipitation on maize yield in the different regions of Ghana. https://doi.org/10.1371/journal.pone.0305762.g003 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 15 / 24

Machine learning models implementation Using 5-fold cross-validation and Grid Search, the results of the stacking ensemble learning model, which includes the optimal parameters of the base models (Gradient Boosting, Linear Regression, Decision Tree, Support Vector Machine, and Random Forest), are shown in Table 5. Table 5 provides a thorough summary of these ideal parameter values, showing the precise setups that produced the best results when used with the stacking ensemble building. Utilizing the unique capabilities of each base model, the stacking ensemble learning employs the prediction of the optimal outcomes of each base model to build the final prediction. Each region’s unique ideal parameters were achieved (see Table 5), and the optimal parameter for each region was identified. This will be useful in predicting maize yield based on climate variables. Results of the performance metrics for the models The statistics in Table 6 demonstrate that, in terms of MSE and the MAE, the LR and the SVM models perform much worse than other single models. As a result, the nonlinear relationship between the climate factors and maize production makes it difficult to predict using either the linear regression model or the support vector machine. The performance of various machine learning models in predicting maize yield across different regions demonstrates notable varia- tions. By evaluating the MSE and the MAE, it is evident that the linear regression model dem- onstrates excellent performance in the Ashanti region. Additionally, the decision tree

outperforms other models in the Eastern region. In the Central region, the random forest model emerges as the most effective. Meanwhile, the support vector machine model shows effectiveness in the Northern and Brong Ahafo regions. The gradient boosting model also stands out as the optimal choice in the Volta and Central regions. The Stacking ensemble model also emerged as a top performer in regions many regions like Western, Upper East, Upper West, and Greater Accra, achieving the lowest MSE and MAE values. It is essential to highlight that the regions, particularly Greater Accra, Upper East, and Upper West, exhibit relatively high coefficients of variation, as shown in Table 3. This Table 5. The results of the optimal parameters for each base model after performing Grid Search. Regions Optimal parameters values of each base model after Grid Search with 5-fold CV DT RF SVM GB max depth min sample_split min_sample_leaf n_estimators max depth min sample_split min_sample_leaf C Kernel Degree n_estimators Learning rate Ashanti 5 10 4 100 15 5 4 1 poly 3 100 0.01 Eastern 5 10 4 200 5 10 1 1 rbf 2 100 0.01 Volta 5 2 4 100 15 10 2 1 sigmoid 2 100 0.01 Central 5 10 4 500 5 10 1 1 linear 2 100 0.01 Brong Ahafo 15 2 1 100 5 2 1 0.1 poly 3 100 0.01 Greater Accra 5 5 2 100 10 10 1 0.1 poly 2 100 0.01 Western 5 10 4 200 5 10 4 1 rbf 2 100 0.01 Northern 5 2 4 100 10 10 1 10 poly 2 100 0.01 Upper east 5 2 1 200 10 10 4 1 sigmoid 2 100 0.01 Upper west

15 2 2 100 15 10 4 10 poly 4 100 0.01 https://doi.org/10.1371/journal.pone.0305762.t005 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 16 / 24

observation emphasizes the stacking ensemble model’s capability to effectively manage data instances characterized by significant variability in these regions. However, due to the nature of the stacking ensemble learning model, where the deficiencies of underperforming models are compensated by the strengths of other models, the MSE of some base models tended to be slightly lower compared to the MSE of the stacking ensemble learning model in some regions. This emphasizes the importance of careful consideration when choosing the base models for the stacking ensemble approach. As indicated by the findings in Table 6, it becomes apparent that the stacking ensemble learning model excels in predicting maize yield based on diverse cli- matic variables across various regions in Ghana. Our findings corroborate the impactful per- formance of the Stacking ensemble model as documented in the study by LI et al. [35]. Moreover, our observations underscore its proficiency in effectively handling small datasets [34], reaffirming its practical utility in diverse scenarios. The stacking ensemble learning model has also proved its effectiveness in handling nonlinear data as suggested by Wang et al. [51]. Feature importance of the climate variables in predicting maize yield Maize yield in Ghana have been experiencing an upward trend as shown in Fig 2. This increas- ing trend can be attributed to the favorable climatic conditions and optimal policies in maize cultivation. Fig 4 illustrates the feature importance of climate variables in forecasting maize

yield. Since random forest performs better across all regions, we base our feature importance calculations for the climatic variables on the random forest Regressor. The statistics below show that CO2 is a key factor in predicting maize output throughout all of Ghana’s regions, Table 6. Performance metric of machine learning models. REGIONS METRIC MACHINE LEARNING MODELS LR DT RF SVM GB STACKING ASHANTI MSE 0.208 0.258 0.254 0.219 0.241 0.300 MAE 0.328 0.376 0.377 0.336 0.357 0.497 EASTERN MSE 0.051 0.041 0.049 0.053 0.044 0.052 MAE 0.148 0.133 0.145 0.155 0.135 0.143 VOLTA MSE 0.148 0.170 0.152 0.145 0.135 0.047 MAE 0.258 0.280 0.262 0.272 0.238 0.281 WESTERN MSE 0.028 0.028 0.027 0.026 0.027 0.025 MAE 0.116 0.118 0.112 0.111 0.115 0.109 CENTRAL MSE 0.244 0.214 0.153 0.178 0.148 0.200 MAE 0.381 0.374 0.307 0.350 0.311 0.401 NORTHERN MSE 0.085 0.088 0.083 0.082 0.085 0.083 MAE 0.220 0.226 0.222 0.220 0.225 0.224 UPPER WEST MSE 0.160 0.278 0.143 0.136 0.157 0.133 MAE 0.332 0.434 0.299 0.279 0.328 0.270 UPPER EAST MSE 2.259 2.915 1.268 0.953 1.178 0.866 MAE 1.080 1.328 0.818 0.725 0.865 0.725 BRONG-AHAFO MSE 0.209 0.282 0.211 0.165 0.232 0.215 MAE 0.441 0.515 0.443 0.359 0.449 0.410 GREATER ACCRA MSE 0.397 0.079 0.090 0.077 0.085 0.065 MAE 0.500 0.225 0.246 0.245 0.246 0.244 Note: The numbers in bold indicate the best performing model in the particular region. https://doi.org/10.1371/journal.pone.0305762.t006 PLOS ONE

Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 17 / 24

having a strong positive impact on maize yield. This supports the findings of Baffour-Ata et al. [28] which found that the benefit of maize yield across the climate variables is attributed to the positive impact of CO2. All the climate variables are crucial and have a positive impact on maize yield with none showing negative importance. However, some of them have a stronger impact on maize yield in the regions. The Greater Accra region, the Brong Ahafo region, and the Upper east region all consider precipitation to be a crucial factor in predicting maize yield. Minimum temperature has less impact in the regions that are part of the Guinea savanna zone Fig 4. The feature importance of the climate variables in predicting maize yield in each region of Ghana. https://doi.org/10.1371/journal.pone.0305762.g004 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 18 / 24

(Upper West region, Upper East region, and Northern region) and the Rain Forest zone (Ashanti region, Eastern region), but has a strong impact on maize yield in the transitional zone, as evidenced by the prominent position of the feature in the Brong Ahafo region. Addi- tionally, the Brong Ahafo and western areas’ temperatures are more significant compared with those in the rest of the nation. The upward trend in maize yield can also be attributed to the advancements in agricultural technology and practices. Farmers have gained access to better-quality seeds, improved irriga- tion systems, and modern farming machinery, all of which have contributed to higher maize yields. Additionally, the government’s focus on agricultural development and support pro- grams has had a positive impact. Through initiatives such as subsidized fertilizer distribution, extension services, and training programs, farmers have been empowered with knowledge and resources to enhance their farming techniques, leading to increased yields. However, the maize yield in the northern region of Ghana (Fig 2h) has been experiencing a downward trend since 2016, therefore, future research should look at the cause of the declining maize yield in the region. The rise in maize yield across all regions in Ghana is not only significant from an agricultural standpoint but also holds economic and food security implications. Higher yields contribute to increased food production, reducing dependency on imports, and ensuring food self-sufficiency. Moreover, surplus maize production can be sold in markets, generating

income for farmers, and stimulating economic growth at both local and national levels. Discussion Summary of Key findings The agricultural sector of Ghana is heavily dependent on maize production, which also has a significant impact on both economic growth and food security. However, variabilities in maize output across the nation’s regions can pose complications for farmers and decision-makers. We go over the main findings and implications of our assessments on maize yield, climate var- iables, and their importance in Ghana. The Greater Accra region shows a high degree of fluctuation in maize production when assessed, which suggests a lack of consistency over time. The Western region, on the other hand, has the lowest coefficient of variation, indicating maize output that is more dependable and predictable. Notably, the Greater Accra region also has a lower average yield in compari- son to other regions, suggesting possible factors affecting maize farming there. Further investi- gation is required to determine and address the variables affecting maize production in the Greater Accra region considering this study. Even though there are variabilities in maize yield in all regions, all of Ghana’s regions have seen a substantial increase in maize output, according to our data. The development of agricultural technology, better farming methods, and govern- ment support programs are all responsible for this growing tendency. Productivity has signifi- cantly increased because of improved irrigation systems [28], access of high-quality seeds, and

sophisticated equipment. Furthermore, measures like subsidized fertilizer delivery, extension services, and training courses have helped farmers become more independent and improved their farming practices [39]. The increase in maize output has implications for the economy and food security, lowering reliance on imports and promoting regional and national eco- nomic growth. Among the climate variables, precipitation patterns, and temperature have significant impact on the productivity of maize. The higher average temperatures are brought on by the Guinea Savannah Zone’s connection to the Upper West, Upper East, and Northern areas. The average temperature in these regions was higher than the average for the rest of the regions, revealing its position as the warmest zone in the nation with less precipitation. The variability PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 19 / 24

in temperature in this region is small indicating how the temperature in the regions vary less over time. The region with the highest variability is the Volta region which experiences signifi- cant variations in temperatures. In all regions of Ghana, carbon dioxide (CO2) also plays a sig- nificant role in predicting maize production. In some areas, additional climate factors like precipitation and minimum temperature are also quite crucial. For instance, the Brong Ahafo, Greater Accra, and Upper East regions place a significant emphasis on precipitation. When compared to the Guinea savanna and Rainforest zones, the relevance of lowest temperature is more obvious in the transitional zone (such as Brong Ahafo). Policy implications Our research focuses on maize yield variability in all the regions in Ghana and its contributing components (climate variables), with the employment of machine learning algorithms. We employed machine learning algorithms because of the abilities in handling nonlinear data. Pol- icymakers should enhance agricultural technologies, farming techniques, and assistance pro- grams to maintain maize production. They should also be aware of the factors influencing maize output in Greater Accra and investigate climate-resilient farming methods. As maize yields increase across the country, these initiatives will also help to mitigate Ghana’s agricul- ture sector’s low average yield and significant yield variability. Overproduction and market sales that arise from this produce economic opportunities and food self-sufficiency, but it’s

important to take sustainability and climate change into account. Although greater production is a good thing, it is crucial to prioritize sustainable agriculture and the reduction of green- house gases, ensuring that climate-smart agriculture practices reduce the sector’s carbon foot- print and boost climate resilience. The information highlights the need to modify agricultural practices to local weather patterns to improve crop output and highlights the complexity of connections between climate and maize yield. Since Ghanaian maize yield assessments provide information on regional variability, yield determinants, and climate variables, this emphasizes the value of localized agriculture that considers regional weather variations and crop growth. The findings of these analyses show the importance of implementing targeted programs and regulations to address local issues, giving Ghana the chance to boost maize production, food security, economic development, and environmental sustainability. This study’s conclusions are based on analysis and facts based on our data span. To further understand Ghanaian maize yield patterns, new data and studies may require additional analy- sis. This study lays the groundwork for evidence-based decision-making and measures to boost maize production and agricultural resilience in the country employing the stacking ensemble learning model as an optimal predictive model. Conclusion This study sought to assess the impact of climate variability on all the regions in Ghana’s maize

yield and to develop an accurate prediction model by employing five machine learning tech- niques (SVM, GB, RF, DT and LR) to establish a high-precision prediction model using the Stacking Ensemble Learning model. Our model is optimized using hyperparametric optimiza- tion and validated using 5-fold cross-validation to improve its generalizability. We found that CO2 is a crucial factor in predicting maize yield in all regions of Ghana. Precipitation is impor- tant for maize yield in the Greater Accra, Brong Ahafo, and Upper East regions. Minimum temperature is less significant in the Guinea savanna and Rain Forest zones shows a greater impact in the transitional zone, particularly in the Brong Ahafo region. The Stacking model showed higher accuracy and robustness compared to SVM and linear regression models, as confirmed by MSE and MAE evaluation metrics. Stacking outperformed other models in the PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 20 / 24

western and Upper East regions, as well as in the Ashanti region, except for slightly higher MSE in the Eastern region. The Stacking model provides a reliable and accurate approach for predicting maize yield based on climate variables, especially with limited data. There exist certain constraints within this investigation that pave the way for future explo- ration. Primarily, the data at our disposal is constrained, potentially impacting the model’s per- formance. Consequently, future research endeavors could leverage extensive datasets to enhance the accuracy of machine learning models in this field. Moreover, specific machine learning models, such as Linear Regression, are susceptible to multicollinearity. Given that the temperature variables (minimum temperature, maximum temperature, average temperature) may contribute to multicollinearity, alternative regression models should be considered. For instance, one might explore penalized linear regression models like Ridge (L2 regularization), Lasso (L1 regularization), and/or Elastic Net (a combination of L1 and L2 penalties). Author Contributions Conceptualization: Samuel Asante Gyamerah, Frank Baffour-Ata. Data curation: Henry Ofoe Agbi-Kaeser. Formal analysis: Samuel Asante Gyamerah, Clement Asare, Henry Ofoe Agbi-Kaeser. Investigation: Clement Asare. Methodology: Samuel Asante Gyamerah, Clement Asare, Henry Ofoe Agbi-Kaeser. Software: Henry Ofoe Agbi-Kaeser. Supervision: Samuel Asante Gyamerah. Validation: Samuel Asante Gyamerah, Frank Baffour-Ata. Visualization: Samuel Asante Gyamerah.

Writing – original draft: Clement Asare, Henry Ofoe Agbi-Kaeser. Writing – review & editing: Samuel Asante Gyamerah, Frank Baffour-Ata. References 1. J. R. Porter et al., “Food security and food production systems,” 2014. 2. Ahmed M. and Hassan F., “Cumulative effect of temperature and solar radiation on wheat yield,” Not Bot Horti Agrobot Cluj Napoca, vol. 39, pp. 146–152, 2011. 3. Tunde A. M., Usman B. A., and Olawepo V. O., “Effects of climatic variables on crop production in Patigi L. G. A., Kwara State, Nigeria,” Journal of Geography and Regional Planning, vol. 4, no. 14, pp. 695– 700, 2011, [Online]. http://www.academicjournals.org/JGRP 4. Maitah M. et al., “Assessment and Prediction of Maize Production Considering Climate Change by Extreme Learning Machine in Czechia,” Agronomy, vol. 11, no. 11, p. 2344, Nov. 2021. https://doi.org/ 10.3390/agronomy11112344 5. Makowski D., Marajo-Petitzon E., Durand J.-L., and Ben-Ari T., “Quantitative synthesis of temperature, CO2, rainfall, and adaptation effects on global crop yields,” European Journal of Agronomy, vol. 115, p. 126041, Apr. 2020. https://doi.org/10.1016/j.eja.2020.126041 6. Bhat K. A. et al., “Low Temperature Stress Tolerance: An Insight Into the Omics Approaches for Legume Crops,” Front Plant Sci, vol. 13, Jun. 2022. https://doi.org/10.3389/fpls.2022.888710 PMID: 35720588 7. Lizaso J. I. et al., “Impact of high temperatures in maize: Phenology and yield components,” Field Crops Res, vol. 216, pp. 129–140, Feb. 2018. https://doi.org/10.1016/j.fcr.2017.11.013

8. Kaur G., Singh G., Motavalli P. P., Nelson K. A., Orlowski J. M., and Golden B. R., “Impacts and man- agement strategies for crop production in waterlogged or flooded soils: A review,” Agron J, vol. 112, no. 3, pp. 1475–1501, May 2020. https://doi.org/10.1002/agj2.20093 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 21 / 24

9. Kuwayama Y., Thompson A., Bernknopf R., Zaitchik B., and Vail P., “Estimating the Impact of Drought on Agriculture Using the U.S. Drought Monitor,” Am J Agric Econ, vol. 101, no. 1, pp. 193–210, Jan. 2019. https://doi.org/10.1093/ajae/aay037 10. Baffour-Ata F., Tabi J. S., Sangber-Dery A., Etu-Mantey E. E., and Asamoah D. K., “Effect of rainfall and temperature variability on maize yield in the Asante Akim North District, Ghana,” Current Research in Environmental Sustainability, vol. 5, p. 100222, 2023. https://doi.org/10.1016/j.crsust. 2023.100222 11. S. K. Assem, “Maize, tropical (Zea mays L.),” Agrobacterium Protocols: Volume 1, pp. 119–134, 2015. 12. Cudjoe G. P., Antwi-Agyei P., and Gyampoh B. A., “The Effect of Climate Variability on Maize Produc- tion in the Ejura-Sekyedumase Municipality, Ghana,” Climate, vol. 9, no. 10, p. 145, Sep. 2021. https:// doi.org/10.3390/cli9100145 13. B. Darfour and K. A. Rosentrater, “MAIZE IN GHANA: AN OVERVIEW OF CULTIVATION TO PRO- CESSING,” in 2016 ASABE International Meeting, American Society of Agricultural and Biological Engi- neers, Jul. 2016. 14. Fagariba C. J., Song S., and Baoro S. K. G. S., “Climate change in Upper East Region of Ghana; chal- lenges existing in farming practices and new mitigation policies,” Open Agric, vol. 3, no. 1, pp. 524– 536, Dec. 2018. https://doi.org/10.1515/opag-2018-0057 15. Peters M. K. and Kusimi J. M., “Changes in wetland and other landscape elements of the Keta Municipal area of Ghana,” J Coast Conserv, vol. 27, no. 1, p. 1, Feb. 2023. https://doi.org/10.1007/s11852-022- 00928-6

16. R. Wassmann et al., “Chapter 3 Regional Vulnerability of Climate Change Impacts on Asian Rice Pro- duction and Scope for Adaptation,” 2009, pp. 91–133. 17. Wang J., Wang E., Yang X., Zhang F., and Yin H., “Increased yield potential of wheat-maize cropping system in the North China Plain by climate change adaptation,” Clim Change, vol. 113, no. 3–4, pp. 825–840, Aug. 2012. https://doi.org/10.1007/s10584-011-0385-1 18. Sommer R. et al., “Impact of climate change on wheat productivity in Central Asia,” Agric Ecosyst Envi- ron, vol. 178, pp. 78–99, Sep. 2013. https://doi.org/10.1016/j.agee.2013.06.011 19. Guntukula R., “Assessing the impact of climate change on Indian agriculture: Evidence from major crop yields,” J Public Aff, vol. 20, no. 1, Feb. 2020. https://doi.org/10.1002/pa.2040 20. Aryal J. P., Sapkota T. B., Khurana R., Khatri-Chhetri A., Rahut D. B., and Jat M. L., “Climate change and agriculture in South Asia: adaptation options in smallholder production systems,” Environ Dev Sus- tain, vol. 22, no. 6, pp. 5045–5075, Aug. 2020. https://doi.org/10.1007/s10668-019-00414-4 21. Tack J., Barkley A., and Nalley L. L., “Effect of warming temperatures on US wheat yields,” Proceedings of the National Academy of Sciences, vol. 112, no. 22, pp. 6931–6936, Jun. 2015. https://doi.org/10. 1073/pnas.1415181112 PMID: 25964323 22. Hsiang S. et al., “Estimating economic damage from climate change in the United States,” Science (1979), vol. 356, no. 6345, pp. 1362–1369, Jun. 2017. https://doi.org/10.1126/science.aal4369 PMID: 28663496

23. Yu C., Miao R., and Khanna M., “Maladaptation of US corn and soybeans to a changing climate,” Sci Rep, vol. 11, pp. 1–12, 2021. 24. Gachene C. K. K., Karuma A. N., and Baaru M. W., “Climate Change and Crop Yield in Sub-Saharan Africa,” in Sustainable Intensification to Advance Food Security and Enhance Climate Resilience in Africa, Cham: Springer International Publishing, 2015, pp. 165–183. 25. Atiah W. A., Amekudzi L. K., Akum R. A., Quansah E., Antwi-Agyei P., and Danuor S. K., “Climate variability and impacts on maize (Zea mays) yield in Ghana, West Africa,” Quarterly Journal of the Royal Meteorological Society, vol. 148, no. 742, pp. 185–198, Jan. 2022. https://doi.org/10.1002/qj. 4199 26. Acquah H. de-G. and Kyei C. K., “THE EFFECTS OF CLIMATIC VARIABLES AND CROP AREA ON MAIZE YIELD AND VARIABILITY IN GHANA,” Russ J Agric Socioecon Sci, vol. 10, no. 10, pp. 10–13, Oct. 2012. https://doi.org/10.18551/rjoas.2012-10.02 27. Adjei V. and Kyerematen R., “Impacts of Changing Climate on Maize Production in the Transitional Zone of Ghana,” Am J Clim Change, vol. 07, no. 03, pp. 463–476, 2018. https://doi.org/10.4236/ajcc. 2018.73028 28. Baffour-Ata F., Antwi-Agyei P., Nkiaka E., Dougill A. J., Anning A. K., and Kwakye S. O., “Effect of cli- mate variability on yields of selected staple food crops in northern Ghana,” J Agric Food Res, vol. 6, p. 100205, Dec. 2021. https://doi.org/10.1016/j.jafr.2021.100205 29. Lobell D. B. and Burke M. B., “On the use of statistical models to predict crop yield responses to climate

change,” Agric For Meteorol, vol. 150, no. 11, pp. 1443–1452, Oct. 2010. https://doi.org/10.1016/j. agrformet.2010.07.008 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 22 / 24

30. Gyamerah S. A., Ngare P., and Ikpe D., “Probabilistic forecasting of crop yields via quantile random for- est and Epanechnikov Kernel function,” Agric For Meteorol, vol. 280, p. 107808, Jan. 2020. https://doi. org/10.1016/j.agrformet.2019.107808 31. Crane-Droesch A., “Machine learning methods for crop yield prediction and climate change impact assessment in agriculture,” Environmental Research Letters, vol. 13, no. 11, p. 114003, Oct. 2018. https://doi.org/10.1088/1748-9326/aae159 32. Appiah-Badu N. K. A., Missah Y. M., Amekudzi L. K., Ussiph N., Frimpong T., and Ahene E., “Rainfall Prediction Using Machine Learning Algorithms for the Various Ecological Zones of Ghana,” IEEE Access, vol. 10, pp. 5069–5082, 2022. https://doi.org/10.1109/ACCESS.2021.3139312 33. Cedric L. S. et al., “Crops yield prediction based on machine learning models: Case of West African countries,” Smart Agricultural Technology, vol. 2, p. 100049, Dec. 2022. https://doi.org/10.1016/j. atech.2022.100049 34. Lee J., Kim J., and Ko W., “Day-Ahead Electric Load Forecasting for the Residential Building with a Small-Size Dataset Based on a Self-Organizing Map and a Stacking Ensemble Learning Method,” Applied Sciences, vol. 9, no. 6, p. 1231, Mar. 2019. https://doi.org/10.3390/app9061231 35. LI Q., XU S., ZHUANG J., LIU J., ZHOU Y., and ZHANG Z., “Ensemble Learning Prediction of Soybean Yields in China Based on Meteorological Data,” J Integr Agric, Feb. 2023. https://doi.org/10.1016/j.jia. 2023.02.011

36. Abdul-Razak M. and Kruse S., “The adaptive capacity of smallholder farmers to climate change in the Northern Region of Ghana,” Clim Risk Manag, vol. 17, pp. 104–122, 2017. https://doi.org/10.1016/j. crm.2017.06.001 37. Owusu-Sekyere J. D., Alhassan M., and Nyarko B. K., “ARPN Journal of Agricultural and Biological Sci- ence ASSESSMENT OF CLIMATE SHIFT AND CROP YIELDS IN THE CAPE COAST AREA IN THE CENTRAL REGION OF GHANA,” vol. 6, no. 2, 2011, [Online]. www.arpnjournals.com 38. Armah F. A., Odoi J. O., Yengoh G. T., Obiri S., Yawson D. O., and Afrifa E. K. A., “Food security and climate change in drought-sensitive savanna zones of Ghana,” Mitig Adapt Strateg Glob Chang, vol. 16, no. 3, pp. 291–306, Mar. 2011. https://doi.org/10.1007/s11027-010-9263-9 39. Tachie-Obeng E., Akponikpè P. B. I., and Adiku S., “Considering effective adaptation options to impacts of climate change for maize production in Ghana,” Environ Dev, vol. 5, pp. 131–145, Jan. 2013. https:// doi.org/10.1016/j.envdev.2012.11.008 40. Lu X., Zhou W., Ding X., Shi X., Luan B., and Li M., “Ensemble Learning Regression for Estimating Unconfined Compressive Strength of Cemented Paste Backfill,” IEEE Access, vol. 7, pp. 72125– 72133, 2019. https://doi.org/10.1109/ACCESS.2019.2918177 41. Feng L. et al., “Alfalfa Yield Prediction Using UAV-Based Hyperspectral Imagery and Ensemble Learning,” Remote Sens (Basel), vol. 12, no. 12, p. 2028, Jun. 2020. https://doi.org/10.3390/ rs12122028 42. Gyamerah S. A., Asare C., Mintah D., Appiah B., and Kayode F. A., “Exploring the optimal climate con-

ditions for a maximum maize production in Ghana: Implications for food security,” Smart Agricultural Technology, vol. 6, p. 100370, Dec. 2023. https://doi.org/10.1016/j.atech.2023.100370 43. Preety K., Prasad A. K., Varma A. K., and El-Askary H., “Accuracy Assessment, Comparative Perfor- mance, and Enhancement of Public Domain Digital Elevation Models (ASTER 30 m, SRTM 30 m, CAR- TOSAT 30 m, SRTM 90 m, MERIT 90 m, and TanDEM-X 90 m) Using DGPS,” Remote Sens (Basel), vol. 14, no. 6, p. 1334, Mar. 2022. https://doi.org/10.3390/rs14061334 44. A. C. Tan and D. Gilbert, “Ensemble machine learning on gene expression data for cancer classifica- tion,” 2003. 45. Ceamanos X., Waske B., Benediktsson J. A., Chanussot J., Fauvel M., and Sveinsson J. R., “A classi- fier ensemble based on fusion of support vector machines for classifying hyperspectral data,” Int J Image Data Fusion, vol. 1, no. 4, pp. 293–307, Dec. 2010. https://doi.org/10.1080/19479832.2010. 485935 46. Qi C. and Tang X., “A hybrid ensemble method for improved prediction of slope stability,” Int J Numer Anal Methods Geomech, vol. 42, no. 15, pp. 1823–1839, Oct. 2018. https://doi.org/10.1002/nag. 2834 47. Meharie M. G., Mengesha W. J., Gariy Z. A., and Mutuku R. N. N., “Application of stacking ensemble machine learning algorithm in predicting the cost of highway construction projects,” Engineering, Con- struction and Architectural Management, vol. 29, no. 7, pp. 2836–2853, Aug. 2022. https://doi.org/10. 1108/ECAM-02-2020-0128

48. A. A. Neloy, H. M. S. Haque, and Md. M. Ul Islam, “Ensemble Learning Based Rental Apartment Price Prediction Model by Categorical Features Factoring,” in Proceedings of the 2019 11th Interna- tional Conference on Machine Learning and Computing, New York, NY, USA: ACM, Feb. 2019, pp. 350–356. PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 23 / 24

49. Twumasi C. and Twumasi J., “Machine learning algorithms for forecasting and backcasting blood demand data with missing values and outliers: A study of Tema General Hospital of Ghana,” Int J Fore- cast, vol. 38, no. 3, pp. 1258–1277, Jul. 2022. https://doi.org/10.1016/j.ijforecast.2021.10.008 50. Panigrahi B., Kathala K. C. R., and Sujatha M., “A Machine Learning-Based Comparative Approach to Predict the Crop Yield Using Supervised Learning With Regression Models,” Procedia Comput Sci, vol. 218, pp. 2684–2693, 2023. https://doi.org/10.1016/j.procs.2023.01.241 51. Wang S., Wu Y., Li R., and Wang X., “Remote sensing-based retrieval of soil moisture content using stacking ensemble learning models,” Land Degrad Dev, vol. 34, no. 3, pp. 911–925, Feb. 2023. https:// doi.org/10.1002/ldr.4505 PLOS ONE Impact of climate variability on maize yields in Ghana PLOS ONE | https://doi.org/10.1371/journal.pone.0305762 June 25, 2024 24 / 24