|
Examples of Use TSTool Documentation
<br />3. Comparison of candidate time series (e.g., using the Results... Graph - XY- Scatter menu).
<br />4. Apply data filling commands.
<br />5. Review final results visually and review time series histories (by right - clicking on a time series in the
<br />Time Series Results list and selecting Time Series Properties).
<br />The data filling approach can be simple or complex. An example of a complex data filling technique is to
<br />use the Fi11MOVE2 ( ) command on daily streamflow data. In particular, consider the following case:
<br />Time series 1 (TS 1) has a long period of gaged unregulated data (e.g., a headwater): 1900 to 2000
<br />Time series 2 (TS2) has a shorter period of gaged data with 1920 to 1950 being unregulated and
<br />1950 to 2000 being regulated (e.g., due to the construction of a reservoir)
<br />The goal is to produce an estimate of unregulated flow for TS2 for the full period 1900 to 2000.
<br />This can be accomplished using the following commands:
<br /># Data filling example - assume daily DateValue time series as input
<br /># Generate some sample data for the example described above:
<br /># ts1 has observed values from 1900 -2000
<br /># ts2 has observed values from 1920 to 1950 and regulated thereafter
<br /># Although data are generated below, they could be read from files or a
<br /># database. In this case, the SetOutputPeriod() command might need to be used
<br /># to ensure that the final result is for a required period.
<br />TS ts1 = NewPatternTimeSeries (NewTSID= "ts1..Streamflow .Day ",Description = "Time series
<br />1, all unregulated",SetStart= "1900- 01- 01 ",SetEnd= "2000 -12-
<br />31 ", Units = "CFS ",PatternValues = "1,2,4,7,12,6,2,1.5 ")
<br />TS ts2 = NewPatternTimeSeries( NewTSID= "ts2..Streamflow .Day ",Description = "Time series
<br />1, unregulated from 1920 to 1950, regulated after ",SetStart = "1900 -01-
<br />01 ",SetEnd =" 2000- 12 -31 ", Units =" CFS", PatternValues = "2,3,6,10,15,3,2.5,2 ")
<br /># Clear out the period 1919- in ts2 because it was not recorded in our example.
<br />SetConstant (TSList = AllMatchingTSID, TSID=" ts2 ",ConstantValue=- 999,SetEnd= "1919- 12 -31 ")
<br /># Clear out the period 1951+ in ts2 because it is regulated and needs to
<br /># be filled with the result of unregulated MOVE2 analysis.
<br />SetConstant( TSList = AllMatchingTSID, TSID=" ts2 ",ConstantValue=- 999,SetStart= "1951 -01-
<br />01 ")
<br /># Analyze and fill the second time series. Transform the data to log10 and
<br /># use monthly equations...
<br />Fil1MOVE2( TSID=" ts2", IndependentTSID=" ts1", NumberOfEquations= MonthlyEquations,Depende
<br />ntAnalysisStart =" 1920- 01- 01 ",DependentAnalysisEnd= "1950 -12-
<br />31",IndependentAnalysisStart =" 1900- 01- 01",IndependentAnalysis End= "2000 -12-
<br />31",Fi1lFlag = "M ")
<br />The above example illustrates a somewhat complicated situation where data filling is facilitated by the
<br />features of the FillMOVE2 ( ) command. If the FillMOVE2 ( ) command is not appropriate, then the
<br />FillRegression ( ) or other commands can be applied. In some cases, it maybe appropriate to fill
<br />different parts of the period using different independent time series. A simpler approach may involve
<br />only a single filling step (e.g., fill the entire period using a single FillRegression ( ) command).
<br />Examples - 8 88
<br />
|