Spatially Averaged versus Point Precipitation in Monongahela Basin:Statistical Distinctions for Forecasting

Abstract

The predictand of a probabilistic quantitative precipitation forecast (PQPF) may be either a point precipitation amount or a spatially averaged precipitation (SAP) amount. At the current state of the art, it is the SAP amount (the volume of water accumulated over an area during a period) that is most predictable. This case study compares the climatic PQPFs of the two predictands within a river basin in the Appalachians, then highlights similarities and distinctions of which the forecasters should be aware. Empirical relations reveal whether or not a given statistic of the point precipitation amount is (i) locally invariant, that is, does not vary appreciably within some area so that a single estimate (e.g., a spatial average) can approximate the statistic at every point within the area, and (ii) amenable to averaging, that is, can be averaged over some area to obtain an approximation to the statistic of the SAP amount. The study also illustrates the effect of elevation on the statistics of point precipitation and highlights seasonal differences. The conclusions point to a need for local climatic guidance to help forecasters in calibrating PQPFs.