Abstract:
The efficacy of low-resolution infrared (IR) satellite data for the estimation of the spatial variation of rainfall is examined. Three analysis techniques were applied to 41 NOAA satellite images of the Southern Alps of New Zealand. Estimated rainfall is compared with surface measurements from 35 sites. Single-channel histograms of cloud-top temperature are shown to have limited application but demonstrate the diurnal variation of cloud cover. Plots of the fraction of cloud amount below four selected temperatures in each of 14 areas across the Alps showed strong orographic dependence for low and middle cloud. Three simple regression models of rainfall dependence on cloudiness and position along a transect crossing the Main Divide show stronger correlation with location than with cloudiness.