Laserfiche WebLink
data in Statemod format, 2) create a Statemod format file for historic record prior to 1974, and 3) <br />add the two Statemod format files to create a file for the period of record (with a complete record <br />in 1974-1996). This three-step process used three TSTooI command files, labeled <br />Stationnamelcom, Stationname2com and Stationfillcom. Regressions were made on all data <br />values as one set rather than performing separate regressions for each month. Shown in Table 1 <br />are the goodness of fit coefficients (r~) as well as the independent station used for filling. The <br />resulting data file, named StationNamefzll, was used in subsequent regressions (see Table 3). <br />Stations other than the four primary independent data sets were also used on occasion as the <br />independent data sets as indicated in Table 3. <br />There were also a number of instances where one climate station was discontinued and <br />subsequently another one started at a similar location. To create one time series that is the sum <br />of two station traces, the operations described in Table 2 were performed using the TSTOOL add <br />command. <br />Following the completion of the processes indicated in Tables 1 and 2, TSTOOL was used to <br />perform regressions between unfilled temperature series and the four primary independent data <br />sets, or other stations. The best filling basis was determined as reflected by the largest goodness <br />of fit parameter (r~ -correlation coefficient squared). Common data for the entire historical <br />period of record was used to determine the relationships used to fill 1974 through 1996. Table 3 <br />shows the filling parameters used in the analysis. <br />All temperature station regression command files where combined to create one TSTOOL <br />command file named tempfillallcom. <br />culossapp Page A-3 of A-13 10/31/01 <br />