ESTIMATION OF DAILY AREA-AVERAGE RAINFALL DURING THE CAPE EXPERIMENT IN CENTRAL FLORIDA

The principal component of the water cycle over land areas is precipit ation. Knowledge of the accuracy of areal precipitation estimates, the refore, is imperative. The manifold observations of hydrometeorologica l quantities in the Convection and Precipitation/Electrification Exper iment (CaPE) that took place in central Florida during the summer of 1 991 have provided an opportunity to examine the various water fluxes. This paper deals with only one of them, the daily area-average rainfal l as derived from rain gauges. The theory for spatial sampling errors for random gauge design is extended to include rain gauge bias and ran dom errors. The requirement for randomly located gauges turns out to b e quite well met with the approximately 100 gauges available each day over the 17 300-km(2) CaPE study area. Rain gauge bias, due mainly to wind effects, is estimated to be 6% based on a previous study and rand om measurement error is calculated from seven pairs of collocated gaug es. For the category of daily area-average rainfall less than 1 mm, th e standard error was found to be 0.27 mm or a 53% variation with respe ct to the category mean. For the highest rainfall category, 11-15 mm, the standard error was 1.49 mm or 11%. For this application, the stand ard error is essentially a consequence of the limited number of gauges . Given that the rain gauges are randomly distributed and arithmetic a veraging is used to estimate the daily area-average rainfall, applicat ion of the appropriate standard error provides a useful measure of the best accuracy achievable in assessing the daily water budget of this area.