Selected Hydrographs and Statistical Analysis Characterizing the Water Resources of the Arkansas

Home Browse Search

Correlations of monthly and annual mean precipitation between every combination of pairs of the 30 stations were computed. The best interstation correlation coefficient for the monthly data was 0.85. About 3 percent of those correlations was greater than 0.80, 17 percent was greater than 0.70, and 30 percent was greater than 0.60. The highest interstation correlation for the annual values was 0.84. Less than 1 percent of those correlations was greater than 0.80; 11 percent was greater than 0.70; and 18 percent was greater than 0.60. Selected classes of interstation correlation of monthly precipitation for every pair of stations is presented in the tables in the "Supplemental Information" section at the end of the report. Correlation of monthly and annual mean precipitation between stations seemed to be related to geographical distance. In general, the closer the stations, the higher their interstation correlation. However, the stations are too far apart to develop useful distance - correlation relations. A statistical test (Kendall Tau) was computed for each of the 30 stations to identify possible time trends in the annual precipitation. The test indi- cated that two stations, 8064 (Sugarloaf Reservoir 1ESE) and 8931 (Westcliffe) had downward trends statistically different from zero at the 1- percent signif- icance level. The data for station 8931 (fig. 31) are suspect, although the station history gives no evidence to support such claim. The hydrograph for station 8064 (fig. 26) does seem to indicate a trend. Two factors can be cited to discount any real trend: (1) The record length is one of the short- est; and (2) the unfortunate timing of missing records around 1970, especially 1969, when nearby stations had above - average precipitation. Snowpack The moisture content in snowfall is included in recorded precipitation data. Snowpack, however, represents the accumulation of snowfall minus losses from melting and evaporation. Snowpack, as used in this report, measures the amount of water being held in storage as snow for potential runoff. Snow - survey courses, where snowpack data have been collected, were selected based on length of record and geographic location. The 18 snow - survey courses chosen are shown on plate 1 and listed in table 3. The period of record and mean April 1 snowpack at each site are shown in table 4. Hydrographs of the April 1 measurements of snowpack for these 18 stations are shown in figures 34 to 51. Tables of all the monthly measurements are presented in the "Supple- mental Information" section (tables 43 -60) at the end of the report. Similar to the precipitation data for the mountainous areas, Snowpack is dominated by orographic effects. A plot of the elevation and mean April 1 snowpack for the 18 snow- survey courses is shown in figure 52. Although considerable scatter occurs about the regression line (some of which might be explained by geographic location), and although different periods of record are used to compute the means, the regression analysis indicates that, in general, the mean April 1 snowpack increases one -half in. for every 100 -ft increase in elevation. Though not shown in figure 52, the intercept of the regression line occurs at about 8,140 ft. This would be the lowest elevation, on the average, at which one would expect snow accumulation on April 1. 24