A 40-year Near-Global Mesoscale Convective System Database (1985-2024)

This database provides a near-global record of Mesoscale Convective Systems (MCSs), identified and tracked via the KFyAO algorithm (as part of the MCSMIP project). The database spans four decades, utilizing two primary satellite brightness temperature (Tb) sources: the Cloud Archive User Service (CLAUS) (1985–2008) and the NCEP/Climate Prediction Center (NCEP/CPC) (2001–2024).

To ensure long-term temporal homogeneity across the transition from CLAUS to NCEP/CPC data, a latitude-stratified, seasonally varying adjustment was applied. This correction minimizes systematic discrepancies between the two source datasets, resulting in a consistent 40-year climate record suitable for long-term variability and trend analysis.

The database is prepared in two complementary formats designed for interoperability and ease of use:
Grid-level (netCDF): These self-describing files contain the binary MCS masks and tracking information. Following the MCSMIP structural standards, each detected system is assigned a unique global identifier (ID). At each time step, all pixels belonging to a specific MCS are labeled with its global ID, while non-MCS pixels are assigned missing values (NaN). These files maintain the original spatial grid of the input Tb data, facilitating direct comparison with other satellite-derived products.
Object-level (CSV): These files provide a concise summary of the physical characteristics of each individual MCS. Key metrics include the global ID, timestamp, centroid coordinates, lifetime, size, intensity, associated mean/max/total precipitation.