Laserfiche WebLink
<br />target area). Long-term average precipitation for the May-to-August period is about 220 to 275 <br />mm (9 to 11 in). Figures 3.2 and 3.3 indicate the average May-to-August rainfall for the no-seed <br />(1948-74) and seed (1975-85) periods, respectively. Monthly rainfall averages in the project area <br />range between 50 and 100 mm (2 and 4 in) for May through August. Figures 3.4 and 3.5 present, <br />as examples, the average July rainfall for the no-seed "and seed years, respectively. <br /> <br />3.2 Organization of the Data <br /> <br />Two types of data are used in this study. The primary data set is rainfall observations from the <br />National Weather Service cooperative observer network (National Oimatic Data Center [6]). A <br />second data set is gridded rainfall fields interpolated from these observations. <br /> <br />3.2.1 National Weather Service Cooperative Observer Network Data <br /> <br />All observations in the study area (35.5 to 41.0oN and 96.0 to 104.00W) for the warm-season <br />months (April through September) for the years 1948 through 1985 were extracted from magnetic <br />tapes and stored on disk on the DARR computer. The data include records from many sites that <br />ceased observations after 1948 and others that started observations after 1948. A number of <br />observation sites were moved one or more times during the 38-year period. When the station <br />number changed as a result of the move, the site was considered to be new; when the number did <br />not change, it was assumed to be a continuous record. The rainfall record itself was not examined <br />for errors, nor was there any attempt to adjust a record to account for changes in elevation or <br />exposure that might have resulted from a minor move. <br /> <br />Daily rainfall amounts from the cooperative network were summed to give the monthly total for <br />each observation site. Frequently there were gaps in the monthly record. In these cases, the <br />missing days were assigned a value equal to the mean for the days with observations (including <br />zeroes). If five or more days were missing for any month,the month was skipped for that site. <br />Thus, the record for any given observation site may have gaps. The activation and deactivation of <br />observing sites and the effect of missing data result in variations in the number of observation sites <br />from month to month and from year to year. <br /> <br />3.2.2 Gridded Data <br /> <br />To facilitate analysis of rainfall patterns, the observed rainfall data (summed for the month) were <br />interpolated to create a rainfall grid. The grid cells are 15 minutes of latitude by 15 minutes of <br />longitude, and the rainfall, calculated at the center of the cell, is assumed to be representative of the <br />entire cell (the difference in cell area from north to south was considered to be insignificant). The <br />study area is 32 cells east-west by 22 cells north-south (704 cells in all). <br /> <br />The interpolation routine (SURFACE m uses the five observations nearest to the center of the grid <br />cell, each weighted by the reciprocal of the square of the distance between the observation site and <br />the center of the cell, to calculate the a rainfall estimate (Sampson [7]). Because of the <br />month-to-month variation in the number of observation sites, the specific sites used to estimate the <br />rainfall for a given cell were not constant over time. <br /> <br />. <br /> <br />'The interpolation routine was used to create gridded rainfall values for each month (April through <br />September) for each year (1948 through 1985). Values for May through August were summed to <br />give seasonal totals. <br /> <br />13 <br />