Identifying high resolution benchmark data needs and Novel data-driven methodologies for Climate Downscaling

Abstract

We address the essential role of information retrieval in enhancing climate downscaling, focusing on the need for high-resolution datasets and the application of deep learning models. We explore the requirements for acquiring detailed spatial and temporal climate data, crucial for accurate local forecasts, and discuss how deep learning (DL) techniques can significantly improve downscaling precision by modelling the complex relationships between climate variables. Additionally, we examine the specific challenges related to the retrieval of relevant climatic data, emphasizing methods for efficient data extraction and utilization to support advanced model training. This research underscores an integrated approach, combining information retrieval, deep learning, and climate science to refine the process of climate downscaling, aiming to produce more accurate and actionable local climate projections.