PNNL

Abstract

Mesoscale convective systems (MCSs) produce more than half of tropical rainfall and are central to the global hydrologic cycle. As climate warms, environments favorable for MCSs may become more common, yet limited observational records hamper understanding of how MCSs respond to variations and changes in their environments. Here, we evaluate how well MCSs are represented in ERA5, a widely used global high-resolution reanalysis product. Using PyFLEXTRKR, which jointly tracks top-of-atmosphere infrared brightness temperature and surface precipitation, we identified MCSs in ERA5 and compared them with those identified in satellite observations using the same detection algorithm. This comparison spans 2007-2020 using hourly data at 0.25° horizontal resolution focusing over the tropics. Results showed that ERA5 reproduces observed brightness-temperature statistics and captures the geographic distribution, frequency, and seasonal and diurnal cycles of MCS cold cloud shields. In contrast, ERA5’s rain-rate distribution is biased, with too much light rain and too little heavy rain, yielding systematic underestimation of MCS precipitation intensity and their contribution to tropical rainfall by 25%-34% in key regions. Consequently, ERA5 is well suited for analyzing MCS cold cloud-shield climatology and evolution, but it should be used with caution for event-level precipitation features and is unsuitable for quantifying the geospatial distribution of MCS precipitation. These findings clarify when ERA5 can eff ectively support studies of potential changes in MCS behavior in a warming climate.