GPM IMERG V07B and V06B: Evaluation Using Ground‐Based Radar Observations and Application in Global Mesoscale Convective System Tracking

Abstract This study evaluates the Integrated Multi‐satellitE Retrievals for Global Precipitation Measurement (IMERG) Version 07B (V07B) against its predecessor V06B to enable a seamless transition of the widely used global Mesoscale convective system tracking data set from V06B to V07B. We use ground‐based radar observations from five climatically diverse regions, the contiguous United States, the central Amazon, the mountains of central Argentina, the equatorial Indian Ocean, and northern Australia, across multiple temporal (1–3 hr) and spatial (0.1°–0.25°) scales. An updated global MCS tracking data set is generated by combining IMERG V07B with satellite infrared brightness temperature. Relative to the radar benchmark, IMERG exhibits a pronounced land‐ocean contrast in its biases: over most land regions, both versions underestimate heavy (10–25 mm hr −1 ) and intense (>25 mm hr −1 ) precipitation, with the bias larger in V07B than in V06B. Over the ocean, V07B improves the intense precipitation retrieval. The global spatial distribution of MCS precipitation contribution is consistent between versions, but MCSs tracked in V07B have lower mean rain rates and heavy rain volume ratios than those in V06B. Our results further suggest that, over the regions and at the spatiotemporal scales evaluated, the apparent overestimation of heavy‐to‐intense precipitation in kilometer‐scale models relative to IMERG may partly reflect low biases in IMERG. This does not imply that km‐scale models are free of bias; rather, the magnitude of model overestimation inferred from IMERG‐based evaluations should be interpreted with caution. These findings provide a critical context for applying IMERG V07B in extreme precipitation, MCS climatology and for evaluating next‐generation convection‐permitting models.