Edit / Fill Missing Data This function accessed from within the Data Module is used to replace missing values in data files with values estimated from nearby locations. When activated, a dialog appears showing a list of the type of data currently displayed in the Data Module:
The files for which missing data are to be filled and those files whose data is to be used for filling should be left selected. Upon pressing OK, the selected files are modified by having their missing data replaced with estimated data. If it might be necessary to preserve a record of missing data, make a copy of the selected files before using this function.
The latitude and longitude coordinates of the selected stations are used to determine the shortest route connecting the stations. The stations are then compared two at a time, first Station A and Station B, then Station B and Station C etc. After reaching the end of the list of stations, the comparisons are then made again in the reverse order to ensure that all missing data have the opportunity to be filled with estimated data.
To fill the data, each column of data is compared between the two files. In cases where both stations have valid data, a relationship is established relating the two stations. This is typically a ratio, but can be a shift relationship as with temperature in meteorology files. Then both files are scanned for missing data. In cases where one file has missing data at the same time and for the same data column for which the other station has valid data, the valid data is used to replace the missing data. The established relationship between the data columns of the two files is used to adjust the valid data from one station when filling in data at another. For example, if Station A averages 1.2 times as much precipitation as Station B, then precipitation data from Station B are multiplied by 1.2 before it replaces missing data in Station A. If there are dates and data columns where no selected data file has valid data, missing data will remain in all the files after this process has been completed.
The time required to complete this process depends on the number of data files and the number of lines in each file. It may take several minutes to process many large files.