Skip to main navigation menu Skip to main content Skip to site footer
Plant Science Today (PST)

Research Articles

Vol. 13 No. sp9 (2026): Recent Advances in Agriculture

District-level cotton yield prediction using temporal trends and climate covariates in Telangana, India (1998–2019)

DOI
https://doi.org/10.14719/pst.17929
Submitted
3 September 2026
Published
10-11-2026

Abstract

District-level cotton yield combines long-term technological change with shorter-term variation associated with climate, management, pests and measurement error. We evaluated trend-assisted reconstruction of observed cotton lint yield for 6 Telangana districts from 1998–2019 (132 district-year observations). Daily India Meteorological Department (IMD) rainfall and temperature series were rebuilt into June–September rainfall totals and mean temperature. September-ending standardised precipitation index (SPI-3) was calibrated separately for each district from July–September rainfall using a zero-adjusted two-parameter gamma distribution over 1951–2019. Standardised precipitation evapotranspiration index (SPEI) was excluded because the previously supplied approximation did not follow a recognised SPEI procedure. Nine machine-learning models predicted deviations from district-specific yield trends; raw yield was reconstructed by adding the predicted deviation to the trend. Under retrospective conditional leave-one-district-out validation (LODO), Extra Trees achieved pooled raw-yield R² = 0.531, mean district raw-yield R² = 0.405 ± 0.139 and root mean square error (RMSE) = 0.891 quintals per hectare (q/ha). The trend-only baseline already achieved pooled raw-yield R² = 0.514 and RMSE = 0.906 q/ha, so Extra Trees improved squared-error performance by only 3.3 %. Residual-scale R² was 0.033 pooled and 0.019 ± 0.129 across districts. In forward validation using only earlier years, ridge regression had RMSE = 1.192 q/ha versus 1.210 q/ha for the trend baseline, but mean district raw-yield R² was −0.281. Correct SPI calibration identified 19 of 132 observations (14.4 %) below −1.0. The results support limited trend-assisted reconstruction of historical raw yield, but not strong climate-only prediction or causal drought-loss attribution.

References

  1. 1. Economics, Statistics and Evaluation Division, Department of Agriculture and Farmers Welfare, Government of India. Crop production statistics: area, production and yield. New Delhi: Government of India; 2026. https://data.desagri.gov.in/
  2. 2. Cotton Corporation of India. Annual report 2019–20. Mumbai: Cotton Corporation of India; 2020. https://cotcorp.org.in/
  3. 3. Parthasarathy B, Pant GB. Seasonal relationships between Indian summer monsoon rainfall and the Southern Oscillation. J Climatol. 1985;5(4):369–78. https://doi.org/10.1002/joc.3370050404
  4. 4. Krishna Kumar K, Rupa Kumar K, Ashrit RG, Deshpande NR, Hansen JW. Climate impacts on Indian agriculture. Int J Climatol.2004;24(11):1375–93. https://doi.org/10.1002/joc.1081
  5. 5. Mall RK, Singh R, Gupta A, Srinivasan G, Rathore LS. Impact of climate change on Indian agriculture: a review. Clim Change. 2006;78:445–78. https://doi.org/10.1007/s10584-005-9042-x
  6. 6. Bhavani P, Chakravarthi V, Roy PS, Joshi PK, Chandrasekar K. Long-term agricultural performance and climate variability for drought assessment: a regional study from Telangana and Andhra Pradesh states, India. Geomat Nat Hazards Risk. 2017;8(2):822–40. https://doi.org/10.1080/19475705.2016.1271831
  7. 7. Van Klompenburg T, Kassahun A, Catal C. Crop yield prediction using machine learning: a systematic literature review. Comput Electron Agric. 2020;177:105709. https://doi.org/10.1016/j.compag.2020.105709
  8. 8. Jeong JH, Resop JP, Mueller ND, Fleisher DH, Yun K, Butler EE, et al. Random forests for global and regional crop yield predictions. PLoS One. 2016;11(6):e0156571. https://doi.org/10.1371/journal.pone.0156571
  9. 9. Roberts DR, Bahn V, Ciuti S, Boyce MS, Elith J, Guillera-Arroita G, et al. Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure. Ecography.2017;40(8):913–29. https://doi.org/10.1111/ecog.02881
  10. 10. India Meteorological Department. Daily gridded rainfall data, 0.25° × 0.25°. Pune: Climate Research and Services; 2025. https://imdpune.gov.in/cmpg/Griddata/Rainfall_25_NetCDF.html
  11. 11. India Meteorological Department. Daily gridded maximum and minimum temperature data, 1° × 1°. Pune: Climate Research and Services; 2025. https://imdpune.gov.in/cmpg/Griddata/
  12. 12. Pai DS, Sridhar L, Rajeevan M, Sreejith OP, Satbhai NS, Mukhopadhyay B. Development of a new high spatial resolution (0.25° × 0.25°) long-period daily gridded rainfall data set over India and its comparison with existing data sets. Mausam. 2014;65(1):1–18.
  13. 13. Srivastava AK, Rajeevan M, Kshirsagar SR. Development of a high resolution daily gridded temperature data set (1969–2005) for the Indian region. Atmos Sci Lett. 2009;10(4):249–54. https://doi.org/10.1002/asl.232
  14. 14. McKee TB, Doesken NJ, Kleist J. The relationship of drought frequency and duration to time scales. In: Proceedings of the 8th Conference on Applied Climatology; 1993 Jan 17–22; Anaheim, California. Boston: American Meteorological Society; 1993. p. 179–84.
  15. 15. Breiman L. Random forests. Mach Learn. 2001;45:5–32. https://doi.org/10.1023/A:1010933404324
  16. 16. Friedman JH. Greedy function approximation: a gradient boosting machine. Ann Stat. 2001;29(5):1189–232. https://doi.org/10.1214/aos/1013203451
  17. 17. Zou H, Hastie T. Regularization and variable selection via the elastic net. J R Stat Soc Series B Stat Methodol. 2005;67(2):301–20. https://doi.org/10.1111/j.1467-9868.2005.00503.x
  18. 18. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: machine learning in Python. J Mach Learn Res. 2011; 12:2825–30.

Downloads

Download data is not yet available.