Warm season rainfall as a function of longitude and time of day over the U.S

We have used national composite images of radar reflectivity over the U.S. to create a climatology of warm season rainfall as a function of longitude and time of day. The climatology is based on five years of data (1996-2000) from the months of May through August using the national mosaic of radar reflectivity produced by Weather Services International (WSI) Corporation. Known as the NOWrad Master Sector, this product is of national scale, and is suited for large domain studies. Every 15 minutes, reflectivity data from the WSR-88D radar network is mapped by WSI to a ~2 km grid in increments of 5 dBZ, undergoing three steps of quality control. For our study, we used a global Z-R relationship to convert from reflectivity factor to estimated rainrate. The estimated rainrates were averaged between 30 and 48 deg N and multiplied by the time step (0.25 h) to obtain the cumulative rainfall for a particular longitude and latitude band. Rainfall data was further distinguished by time of day.

As expected, the cumulative rainfall increases from west to east, starting with a minimum over the semi-arid Rocky Mountain cordillera (115-105 W) and ending with a broad maximum near the eastern boundary of our study area. Even with a single Z-R relationship, absolute and relative rainfall totals are close to those derived by independent rain gauge analyses.

Plotting rainfall with respect to time of day brings out the importance of the diurnal cycle in the warm season. Almost all precipitation west of 105 W is produced in the early to late afternoon, while the time of maximum precipitation shifts to later periods as one moves further east. The plots suggest a daily initiation of convection in the early afternoon over the Rocky Mountains, with eastward progression of heavy precipitation all the way to the Appalachian Mountains over a 48 hour period. If this pattern could be better understood and represented in numerical models, this could lead to improvements in both climate models and 1-2 day forecasts of heavy precipitation over the U.S.