Mesonet‐Based Ground Validation of IMERG‐V07 and GSMaP‐V08 Over a Semi‐Arid Dryland Ecosystem in West Texas

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Climate change has caused highly uncertain precipitation regimes and associated hydrometeorological hazards over global drylands. Accurate precipitation measurements are vital for monitoring dryland ecohydrology. Satellite precipitation products (SPPs), including the Integrated Multi‐satellitE Retrievals for GPM (Global Precipitation Measurement) (IMERG‐V07) and the Global Satellite Mapping of Precipitation (GSMaP‐V08), provide quasi‐global coverage but require regional‐scale ground validation to assess errors and reliability, over drylands. We evaluated IMERG‐V07 and GSMaP‐V08 using measurements from 107 West Texas Mesonet (WTM) sites from 2018 to 2022. Precipitation over the semi‐arid WTM domain is concentrated during the May–October wet season. The observed southeast–to‐northwest precipitation gradient (i.e., higher to lower) exists primarily due to dryline‐induced convergence and orographic lifting. At monthly scale, IMERG (GSMaP) overestimates (underestimates), with GSMaP demonstrating greater overall accuracy. SPPs perform best during the peak precipitation months (May, September, and October) and reproduce interannual fluctuations well. The GSMaP performs better on monthly and annual timescales. The mean bootstrap 95% confidence interval for correlation coefficient were [0.759, 0.825] and [0.837, 0.892] for IMERG and GSMaP, respectively. Monthly accumulations of SPPs exhibit intensity dependent biases in high‐intensity precipitation regimes. For near‐real‐time (NRT) detection of extreme precipitation, IMERG‐Early shows higher sensitivity via higher detection but also more false alarms. The GSMaP‐NRT provides balanced event detection, with lower false alarms and higher critical success index in most cases. However, the GSMaP reliability‐flag and IMERG quality‐index were also found to be decisive factors. Results highlight the need for intensity‐ and timescale‐sensitive validation to improve SPP algorithms for advanced precipitation monitoring and forecasting over drylands.
Funding Information
  • Earth Sciences Division (80NSSC22K1786)

This publication has 52 references indexed in Scilit: