Laserfiche WebLink
<br />12 <br /> <br />PAVG. = the average precipitation for the available record at the <br />J <br /> <br />station for which the specific monthly value is being <br />estimated. <br /> <br />For the purpose of makin9 these estimat.es, 2!i state cl imatic divisions <br /> <br />were used (Doesken et al., 1983). These divisions are shown in <br /> <br />Figure 3. <br /> <br />B. Rati 0 adj ustment JJrocedur:!:. <br /> <br />Priority 3 stations (only 15-;~4 complete yeal"S of data) had far too <br /> <br />much missing data to justify estimating values for each missing month. <br /> <br />For these stations, annual averages were calculated based on only the <br /> <br />available complete years of data. Then annual averages were adjusted to <br /> <br />the 1951-1980 period using the rat'io adjustment method defined below. <br /> <br />LTAVG. <br />J <br /> <br />STAVG. <br />= -)- x LTAVGk, <br />STAVGk <br /> <br />where <br /> <br />L TAVG. <br />J <br />STAVG. = <br />J <br /> <br />= adjusted 1951-80 annual mean pl'ecip'itation for station j. <br /> <br />short term annual mean precipitation calculated from <br />available complete years of data for station j. <br /> <br />STAVGk = annual mean precipitation for station k (priority 1 <br /> <br /> <br />station) computed for those years station j had complete <br /> <br />data. <br /> <br />LTAVGk = 1951-80 mean annual precipitation for station k. <br />In order to determine which "long-term" 3D-year p'riority 1 station might <br /> <br />provide the best comparison witll any part'icular short term priority 3 <br /> <br />station, the state was div'ided into 7 regions (Figure 4). Correl ation <br /> <br />coefficients were then computed for all possible combinations of short- <br /> <br />term and 3D-year stations in each l^egion baSE!d on precipitation total s <br />