Evaluation of sub-hourly MRMS quantitative precipitation estimates in mountainous terrain using machine learning

Precipitation gage networks are often sparse or nonexistent in mountainous regions, resulting in a problematic data gap when accurate local observations are required. Quantitative precipitation estimates (QPEs) approximate precipitation from remote sensing data, gage networks, and climate models. These datasets are spatially continuous but are subject to various sources of measurement error, especially in complex terrain. In recent decades, QPEs have been improving in accuracy and resolution, but there is no comprehensive method of estimating uncertainty, and error models are often tested in specific regions during a small number of events. The Multi-Radar Multi-Sensor (MRMS) product incorporates radar, climate model, and gage data at a high spatiotemporal resolution for the contiguous United States. The goal of this study is to provide a framework for understanding the uncertainty of MRMS in mountainous areas with limited observations in the mountains of Colorado.