Comparing and Optimizing Four Machine Learning Approaches to Radar-Based Quantitative Precipitation Estimation

• Published in: Remote sensing (Basel, Switzerland) Vol. 16; no. 24; p. 4713
• Main Authors: Liu, Miaomiao, Zuo, Juncheng, Tan, Jianguo, Liu, Dongwei
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.12.2024
• Subjects: Accuracy; Correlation coefficient; Correlation coefficients; Data collection; Data points; Decision trees; Deep learning; Estimates; KAN deep learning; Learning algorithms; Machine learning; machine learning method; Moisture content; multivariable; Neural networks; Observational learning; Optimization; Performance evaluation; Precipitation; Quality control; Radar; Radar systems; radar-based quantitative precipitation estimation; Rain; Rainfall; Rainfall intensity; random forest; Reflectance; Support vector machines; Water; Water content; Wavelet transforms; China

To improve radar-based quantitative precipitation estimation (QPE) methods, this study investigated the relationship between radar reflectivity (Z) and hourly rainfall intensity (R) using data from 289 precipitation events in Shanghai between September 2020 and March 2024. Two Z-R relationship model...