Machine-Learning-Based Climate Impact Analysis for Agriculture

In this study, a climate impact analysis on agricultural yield has been carried out by using a curated experimental dataset which consists of 1,500 crop-year observations for five crops in 15 Indian states for the study period 2005–2024 using machine learning approaches. The modeling of the behavior of photoclimatic yield for four types of climate exposure, two seasonal structures, and five types of vegetation and vegetal status and management buffers was carried out. Well defined differences in temperature, precipitation, soil moisture, stress index and yield were found in the descriptive results. A negative correlation between yield and temperature and climate stress factors, and a positive correlation between yield and soil moisture, were found. Looking at moisture retention, stress timing and duration is important and not only rainfall depth for a season alone, as rainfall alone was weak. The modelling results showed that the Gradient Boosting Regressor had highest test statistics results; RMSE (0.2325 t/ha), MAE (0.1826 t/ha), R2 (0.9712) and MAPE (5.31%). Ridge regression did did not diverge much while random forest exhibited higher error. Some of the large factors were identified as baseline yield, soil moisture, climate stress index, NDVI, year, carbon dioxide, irrigation and temperature. Technically speaking, the result has no problem: “Climate does not effect one variable each other”. This is reflected in temperature, water and plant & regional capacity indicators. The approach is descriptive and is limited in its informativeness. The data is simulated and should only be utilized as an experimental validation data set, and not field data. The next steps for the future would be to replace the experimental record with the Before where they use it for policy or insurance purposes, and to include district level mapping as well as uncertainty intervals and explainable output from the model. Also, with the increased capacity of this work, the exposition of the climate adaptation possibilities (expansion of irrigation, heat-tolerant cultivars, sowing date change and soil-water conservation) is also improved. These additions would move the model to more geographical, applied and disciplined assessment work.

[1]     J. Jägermeyr, C. Müller, A. C. Ruane, J. Elliott, J. Balkovič, O. Castillo, et al., “Climate impacts on global agriculture emerge earlier in new generation of climate and crop models,” Nature Food, vol. 2, pp. 873–885, 2021, doi: 10.1038/s43016-021-00400-y. [2]     A. Ortiz-Bobea, T. R. Ault, C. M. Carrillo, R. G. Chambers, and D. B. Lobell, “Anthropogenic climate change has slowed global agricultural productivity growth,” Nature Climate Change, vol. 11, pp. 306–312, 2021, doi: 10.1038/s41558-021-01000-1. [3]     E. E. Rezaei, S. Siebert, and C. Müller, “Integrated assessment of climate change impacts on crop yields and crop water productivity in Europe,” Agricultural Systems, vol. 193, Art. no. 103210, 2021, doi: 10.1016/j.agsy.2021.103210. [4]     G. S. Malhi, M. Kaur, and P. Kaushik, “Impact of climate change on agriculture and its mitigation strategies: A review,” Sustainability, vol. 13, no. 3, Art. no. 1318, 2021, doi: 10.3390/su13031318. [5]     D. Ozdemir, “The impact of climate change on agricultural productivity in Asian countries: A heterogeneous panel data approach,” Environmental Science and Pollution Research, vol. 29, no. 6, pp. 8205–8217, 2022, doi: 10.1007/s11356-021-16291-2. [6]     M. Rashid, B. S. Bari, Y. Yusup, M. A. Kamaruddin, and N. Khan, “A comprehensive review of crop yield prediction using machine learning approaches with special emphasis on palm oil yield prediction,” IEEE Access, vol. 9, pp. 63406–63439, 2021, doi: 10.1109/ACCESS.2021.3075159. [7]     L. Benos, A. C. Tagarakis, G. Dolias, R. Berruto, D. Kateris, and D. Bochtis, “Machine learning in agriculture: A comprehensive updated review,” Sensors, vol. 21, no. 11, Art. no. 3758, 2021, doi: 10.3390/s21113758. [8]     K. Alibabaei, P. D. Gaspar, and T. M. Lima, “Crop yield estimation using deep learning based on climate big data and irrigation scheduling,” Energies, vol. 14, no. 11, Art. no. 3004, 2021, doi: 10.3390/en14113004. [9]     M. G. Ziliani, M. U. Altaf, B. Aragon, R. Houborg, T. E. Franz, Y. Lu, J. Sheffield, I. Hoteit, and M. F. McCabe, “Early season prediction of within-field crop yield variability by assimilating CubeSat data into a crop model,” Agricultural and Forest Meteorology, vol. 313, Art. no. 108736, 2022, doi: 10.1016/j.agrformet.2021.108736. [10]   P. Muruganantham, S. Wibowo, S. Grandhi, N. H. Samrat, and N. Islam, “A systematic literature review on crop yield prediction with deep learning and remote sensing,” Remote Sensing, vol. 14, no. 9, Art. no. 1990, 2022, doi: 10.3390/rs14091990.