TSTool_UserManual_10-21-00 - Laserfiche WebLink

TSTool Documentation Introduction - 12 H owever, this may lead to confusion and generally should b e avoided by using unique identifiers for each time series that is being processed. During processing, extra time series can accumulate and will be available for output. Use the F ree() command to free time series that are no longer needed. This removes the time series from memory. Output commands also may use the TSList parameter to indicate which time series are to be output. 4. After processing the time series, a list of available time series that are in memory are listed in the GUI. One or more of the se time series can be selected and viewed using the Results menu or analyzed using the Tools menu (also right click on time series listed in the results menu at the bottom of the main window). Time series also can be save d in some of recognized input type formats using the File...Save…Time Series As menus. If running in batch mode using the -commands option, all of the above steps occur in sequence and the GUI is not displayed. P rocessing the example shown above results in three time series in memory: 1. A time series identified by 08235350.USGS.Streamflow.Month , containing the sum of the two time series. 2. A time series identified by 08236000.DWR .Streamflow.Month , containing the input to the A dd() command. 3. A time series also identified by 08235350.USGS.Strea mflow.Month , containing the original data from the time series that is added to. This contains the original data because a time series identifier by itself in a command list will cause the time series to be read. These time series can be graphed or saved in an output file. 2 .7 Time Series Ensembles A time series ensemble is a group of related, typically overlapping, time series. Ensembles can be used to manage related scenarios (e.g., input and results of model scenarios ) or as a way of shifting a hist orical time series so that years overlap. Many commands operate on a list of time series by using the parameters TSList=EnsembleID and EnsembleID=”SomeID”. Statistics time series can be derived from ensembles, for example to calculate the a verage conditi on over time (although care must be taken in whether this can be interpreted as a time series of related values, for use as input to a process). Ensembles are assigned unique identifiers and are displayed at the bottom of the TSTool main window in the Ens embles results tab. Each time series in an ensemble is also available for individual processing and is listed in the Time Series r esults tab . Time series in ensembles often have a [Seq] sequence number in the TSID corresponding to the year of the data, a s discussed in Section 2.2 . The following fi gure illustrates an ensemble of annual time series created from a long historical time series . 24