Assessing Future Climate Trends: Downscaling Maximum Temperature for Water and Agricultural Management

Climate change represents a serious challenge to agricultural systems around the world, as increasing temperatures and changing rainfall patterns impact crop yields, water supply, and ecosystems. Accurate forecasts of future daily maximum temperatures (Tmax) are vital for evaluating how vulnerable agricultural systems are to climate change. Rising Tmax can result in heat stress for crops, heightened water use in crops, diminished yields, and alterations in crop developmental timelines. Grasping the projected Tmax is crucial for recognizing potential threats to crop production, maintaining food security, and developing sound agricultural policies. To analyze upcoming climate changes, Global Circulation Models (GCMs) are beneficial. Nonetheless, GCMs reveal broad climate trends but do not capture local variations in Tmax that influence agriculture. For example, in Aurangabad, local Tmax variations have a substantial effect on crop development, water needs, and harvest yields. To address this issue, downscaling methods are useful. These methods transform broad-scale data into more precise, local Tmax figures. This study employs the Random Forest (RF) algorithm, an effective machine learning approach, to statistically downscale Tmax projections for Aurangabad, India, utilizing the CMIP5 (Coupled Model Intercomparison Project Phase 5) CanESM2 (Canadian Earth System Model, Version 2) GCM, which is a Canadian climate model that furnishes global climatic data for future projections. The Random Forest algorithm functions by identifying patterns from historical data, allowing it to make future predictions, which makes it well-suited for intricate, non-linear relationships within climate information. The CanESM2 model was selected for its expansive coverage and its demonstrated capability to yield precise regional climate forecasts, making it well-suited for this research. The CanESM2 model generates future climatic information on a global scale. By combining observed Tmax data with pertinent large-scale climate variables from the CMIP5 CanESM2 model, the Random Forest model was created and validated. Following successful calibration and validation of the model, it was applied to downscale future Tmax scenarios under three Representative Concentration Pathways (RCPs): RCP 2.6, RCP 4.5, and RCP 8.5 for three future time frames the 2020s, 2050s, and 2080s. The findings reveal a considerable warming trend in Tmax across all scenarios, with the most notable warming expected under the RCP 8.5 scenario compared to the baseline period of 1961-2005. These results underscore the need for adapting agricultural practices to future climatic conditions, assisting local farmers and policymakers in preparing for the rising challenges presented by climate change. These results offer essential insights for agricultural stakeholders in Aurangabad to evaluate the potential effects of climate change on crop production, devise strategies to mitigate heat stress, and adopt climate-smart agricultural practices to bolster resilience and assure food security in the region.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.