Is there a weekly cycle of warm season precipitation?

The recent work of Bell et al. (JGR 2008) documented a midweek increase

in warm season (June-August) rainfall over portions of the United States as

deduced from the Tropical Rainfall Measuring Mission (TRMM) data from 1998

to 2005. It was hypothesized that anthropogenic air pollution is the

principle cause of the observed weekly cycle of a Tuesday/Wednesday maxima

and a Saturday/Sunday minima.

The authors of this study have published several dynamically-based climatologies

of summer rainfall over the United States using a national composited radar

data set from the WSR-88D radar network. Since ground-based radars offer

some advantages to the TRMM data, a study was started to compare and confirm

the results of the satellite-based study. Using a 12-year data set statistics

of radar-derived rainfall are calculated and stratified by day of week. Several

statistical methods are used to determine if significant differences exists

between days. Initial results tended to confirm the findings of Bell et al.,

especially over their most statistically significant area, the east-central

U.S. To further check our results the data were resampled randomly into virtual

"Tuesdays" and "Saturdays". The random resampling produced patterns that are

similar in scale and magnitude to the true Tuesday versus Saturday rainfall

differences. Work will continue to better understand the relationship between

the local PDFs of rainfall amount and the occurrence of large coherent regions

with apparent day-of-week differences.