Abstract
With climate change becoming increasingly evident, understanding how it affects local weather patterns is crucial. Such kind of studies provide valuable insights into how rainfall and temperature variability are changing over time, helping communities adapt to shifting climate conditions with respect to their agricultural practices. This study examines the trends in rainfall and temperature in Muthalamada, Kerala, using statistical trend analysis and forecasting using SARIMAX modeling techniques. Rainfall data spanning from 1981 to 2022 indicates an upward trend in both Chulliyar and Meenkara regions, with varying patterns across seasons. The wettest month consistently was July, while the southwest monsoon contributed significantly to annual rainfall. Predictions for the next decade suggest an increase in rainfall for both regions, except for winter. Temperature analysis reveals an upward trend in maximum and minimum temperatures across all seasons, with strong implications for mango cultivation. Maximum temperatures are anticipated to rise slightly in summer and winter but decrease in post-monsoon and pre-monsoon seasons, affecting mango flowering and fruit set. Conversely, minimum temperatures are expected to decrease across all seasons. Relative humidity is projected to increase during post-monsoon and southwest monsoon but decrease in other seasons, impacting pest populations. The performance of the SARIMAX model was evaluated using various metrics such as correlation coefficient (r), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Root Mean Squared Error (RMSE). The analysis revealed overall accurate predictions for rainfall, temperature, and humidity, with some variability across seasons. Model performance evaluation indicates accurate predictions for rainfall, temperature, and humidity, with implications for mango farming. Implementation of appropriate mitigation strategies is crucial for Muthalamada mango farmers to adapt to climate variability and enhance production and export potential.