Improving Agricultural Productivity in Indonesia Using Xgboost Precipitation Prediction with Downscaling and Bias Correction

Climate change has negatively impacted agriculture, particularly by altering planting and harvesting times. This has led to crop failures, reduced productivity, and increased production costs for farmers. Understanding current climate trends is crucial for farmers to predict suitable planting times and locations. This study focuses on the precipitation aspect of climate, which is vital for plant growth. It uses historical climate data from BNU-ESM and reanalysis data from the RCP8.5 and ERA-5 climate models. The BNU-ESM data, with a spatial resolution of $2.79^{\circ} \times 2.81^{\circ}$, required downscaling to match the $0.25^{\circ} \times 0.25^{\circ}$ resolution of the ERA5 data. This alignment allowed for accurate model training and precipitation forecasting. The study found that future changes in rainfall will impact planting and harvesting schedules, necessitating adjustments in farming practices. This research highlights the need for the agricultural sector to adapt to climate change and underscores the importance of understanding climate variability and utilizing climate models for future agricultural planning.