Description:
A database of CSV files, containing 2007-2017 GridRad MESH cell characteristics co-located with GOES-12/GOES-13 or GOES-16 embedded cold spot detections, other IR-derived updraft characteristics, SPC reports, and MERRA-2 or ERA5 convective parameters analyzed in this Journal of Artificial Intelligence for the Earth Systems paper. Columns 1-5 contain spatio-temporal information of each cell centroid. Column 6 is the 95th-percentile of the GridRad Maximum Expected Size of Hail cell maxima. Columns 7-18 contain automated infrared-based overshooting cloud top detection probability, and many other fields produced by NASA Langley Research Center convective cloud characterization algorithms. In addition, the files contain other MESH cell characteristics (e.g., area of the cell exceeding 10, 25, 50, and 75 mm), reanalysis convective parameters interpolated to the GOES satellite pixels, elevation, viewing zenith angle, the fractional resolution of a NADIR pixel relative to the current match. Missing values are -888 or left blank.

Publication Title:
Scarino, B. R., Itterly, K. F., Bedka, K. M., Homeyer C. R., Allen J. A., Bang S. D., Cecil, D. (2023). Deriving Severe Hail Likelihood from Satellite Observations and Model Reanalysis Parameters using a Deep Neural Network. Artificial Intelligence for the Earth Systems. Accepted pending revision.

Data Set/Document Primary Contact:
benjamin.r.scarino@nasa.gov

+ Access Data 

Page Revision Date: 05/30/2023 Revision 1 (REVISED DATA SET DEPLOYED THIS DATE)