Accurate information about rainfall is crucial for understanding risks like floods and droughts. Scientists recently created a new model using methods called Random Forest and Wavelet Multi-Resolution Analysis (RF-WMRA). The goal was to improve data about rainfall in the area along the Sichuan-Tibet Railway from 1961 to 2020. This model used data from the European Centre for Medium-Range Weather Forecasts and a global satellite map of rainfall. It also considered factors like altitude, humidity, temperature, and wind speed. In the end, they got a set of rainfall data with high detail (0. 1° resolution).

The new data was compared to old data (ERA5 and GSMaP) to see which was better. The new data did well in tests like Correlation Coefficient (0. 89), Root Mean Squared Error (1. 65 mm), Relative Bias (-0. 01%), Mean Absolute Error (0. 42 mm), and Kling-Gupta efficiency (0. 80). After adjustments, these numbers got even better! This tells us that the new model worked well. For big rain events, the new data was the best among the three sets.

This new data can help us understand how geological disasters happen and respond to climate change. It's important for studying topics like floods and droughts.

Weather Data Refreshed: A New Look at Rainfall

Actions