A Study of Rainfall Spatial Correlation using Radar Observation

Rainfall spatial correlation is a key factor for driving the relationship between a “point” quantity and its corresponding “area-averaged” quantity in rain fields. In an application, the relationship of rainrate variance between a “point” value and its corresponding “area-averaged” one was used to retrieve the rainrate variance in a radar beam for the Nonuniform Beamfilling correction of spaceborne radar rainfall measurements. In another application, the relationship between the “point” and the “area-averaged” was also applied to get the areal representative error of raingage in the comparison of rainfall measurements between radar and raingage.

Since the resolution of the radar rain data used in the former published rainfall spatial correlation researches was about 1km, and this could not exhibit the rain variation within the small scales less than 1 km, so the high resolution data was needed and was applied to this work. We conducted rain cloud observations with a high resolution of 125m in the summer of 2000 using a 3cm-meteorological radar, and obtained the high resolution reflectivity data from rainfall fields in three rainfall systems. Using these data, average reflectivity, strength of nonuniformity described by an excursion (the difference between maximum and minimum) of reflectivity herein, were calculated at the different spatial resolutions of 0.5km, 1km, 2km, 3km and 4km. The calculation results indicate that the nonuniformity is significant, even within 1km size, the excursion above 10dB is common. Then rainfall spatial correlation function was calculated using high resolution data and the simulated 1 km (low) resolution data respectively, which showed how small scale rainfall variability affects rainfall spatial correlation.