Laserfiche WebLink
<br />1. Introduct ion <br /> <br />The statistical characteristics of rainfall are important considerations <br /> <br />in many climatological, agricultural, hydrological, and weather modification <br />applications. Of the various statistical properties, the variability of <br />rainfall is, perhaps, the most important, especially in the evaluation of <br />precipitation augmentation experiments in which rainfall is the primary <br />response variable. Estimates of the natural variability of rainfall are, <br />however, compounded by vari ance components that are introduced by the <br /> <br />method of measurement, namely those due to sampl ing and measurement errors. <br /> <br />It is the purpose of this papE~r to examine the sampling variance of raingage <br /> <br />networks, the most commonly used precipitation measurement system, and <br /> <br />quantitatively relate this variance to both the density of raingd.ges and the <br />characteristics of the isohyetal pattern. <br /> <br />The variance due to sampl ing t",esults from the necessity of estimclting areal <br /> <br />rainfall values from point measurements of a storm footprint that contains <br />spatial gradients. This is especially true for convective storms which <br /> <br />are of primary interest in this study. The estimation procedure is usually <br /> <br />based on the assumption that a point measurement by a precipitatjion gage is <br />represent at ive of the area that surrounds it. Th us, depend i ng on just how <br />the storm passes over the gagE! network, the precipitation is usually either <br /> <br />under or overestimated. <br /> <br />A number of investigators (e.g., Light 2, Linsley and Kohler, 1951; <br />McGuinness, 1963; Huff, 1970; Woodley et al., 1975; and Eddy, 1976) have <br /> <br />2L ight, P., 1947: Hydrol ogi c aspects of thunderstorm rainf all. Hydrometeor <br />Report No.5, U.S. Weather BurE~au-Corps of Engineers, 260-268. <br />