From the Hydrologic Sampling Editor, click the Historical Record tab (review the Hydrographs Tab section); this tab is where, for the Bootstrapping Historical/Synthetic Basin-wide Events sampling method, the user must at least one historical hydrograph DSS record for each hydrograph location must be defined. A record is needed for each parameter (variable) at each location.

All historical records (for all locations and all parameters of each) must span exactly the same time-window, meaning that the hydrographs have the same start and end dates and times. Within each year of the historical period of record, the relevant hydrograph time series for a study can have time-windows that extend from a single day to a full year. The benefit of defining shorter time-windows (less than a full year) is a reduction in the FRA compute time, but this should be balanced with ensuring that time-windows are long enough to capture full routing of the storm events of interest. If the Include Forecasts option (review Selecting Bootstrapping Historical/Synthetic Basin-wide Events method, Including Forecasts (Optional)) has been selected, forecast information about the selected historical locations will be entered on the Forecasts tab (see the Forecasts Tab section).

Hydrographs Tab

The Hydrographs tab contains a DSS time series panel, where a user will select the DSS records for the time series spanning the historical period of record, providing the events to be sampled. The selected DSS time series records need to span the historical period of record and contain data only for the relevant time periods within each year (e.g., the flood event(s) time-window(s)), with blanks for all other time periods. The length of an event within each year can vary. Furthermore, the flood event time-windows (start and end dates and times) must also be the same for all locations.

An example historic hydrographs plot is provided below which displays a historical flow time series for an entire period of record for the watershed.

The following figure displays an example annual rain event (left plot) and an example annual snowmelt runoff (right plot) for three different locations in the watershed.

However, if the FRA compute in the watershed model is actually intended to simulate a full year, rather than shorter events within each year, the time series records can be continuous. In this case, data would be present in all time periods within the entire period of record without blanks. The Start of Water Year (review Selecting Bootstrapping Historical/Synthetic Basin-wide Events method) defines the beginning of each year to divide the historical record into annual events for sampling, and it is important to specify this correctly when using continuous data.

To specify the DSS time series that defines the historical inflow hydrographs (or time-series for other defined parameters) throughout the watershed, from the Hydrologic Sampling Editor, click the Historical Record tab. From the Hydrographs tab, from the DSS time series panel:

1. For each Location (hydrograph) listed, and each parameter at that location, a DSS path must be identified (DSS File and pathname Parts A – F). To define the DSS record for a location of interest, click on a cell of the row of the hydrograph location.
2. Click Select DSS Path, the HEC-DSSVue dialog box will open. Refer to the HEC-DSSVue User's Manual (available online, here: HEC-DSSVue Documentation) for more information regarding DSS pathnames and HEC-DSSVue.
   
3. From the HEC-DSSVue dialog box, click ; the Open HEC-DSS File browser will open. Browse to the location of a DSS file (*.dss) that contains the time series of interest (e.g., hydrographs). Click on the appropriate *.dss file, the File name box now contains the name of the selected file (e.g., trimmed_Bluestone_In.dss).
   
   
4. Click Open, the Open HEC-DSS File browser will close. The HEC-DSSVue dialog box now displays the DSS records for the selected DSS file.
5. For the location (e.g., Bluestone Inflow Near Glen Lynn) that was selected in the DSS input table in the Hydrologic Sampling Editor, find and select the proper pathname (record) from the list of DSS records in the HEC-DSSVue dialog box, click Set Pathname. From the Hydrologic Sampling Editor, the DSS input table now contains the DSS File pathname (Parts A – F) for the selected Location (e.g., Bluestone Inflow Near Glenn Lynn).
   
6. Select the next location, from the HEC-DSSVue dialog box, find and select the proper pathname (record) from the list of DSS records in the HEC-DSSVue dialog box, click Set Pathname. Repeat this step until all locations in the DSS input table contain the correct DSS filename and the DSS pathnames.
7. Click Apply, this saves data and keeps the Hydrologic Sampling Editor open for further modifications. When finished, click OK to save the selected hydrologic sampling alternative and close the Hydrologic Sampling Editor. Data Check (review Hydrologic Sampling Editor Interface, Data Check section) can be used to search for data consistency error(s) prior to closing the editor.

Forecasts Tab

From the Forecasts tab, of the Historical Record tab, a user will select the DSS records for the observed values at forecast locations that were defined in the Settings tab (please review Ranking Events, Uncertainty Method, and Synthetic Event Probabilities (Settings tab)).

Select Historical Forecast Time Series

Forecast DSS time series data is required for the forecast sampling option and is used as the starting volume to which randomly sampled forecast error is added. Most commonly, these values will be the actual or "true" volumes that a forecast is targeting. For example, the actual April through August volume of a year is typically used as the starting value for a randomly generated snowmelt forecast of April through August volume. A generated error deflects the flow forecast from the actual value by an amount randomly sampled from the forecast error statistics and an Autoregressive lag 1, AR(1), sampling process. Each specified forecast DSS time series must have a volume value in KAF for every forecast date of every year.  The DSS time series of initial forecast volumes for the historical record (Forecasts tab) must match the dates of the historical flow record (Hydrographs tab).  To illustrate this period of record relationship, the period of record graph displays an example of a flow record (for snowmelt runoff hydrology, displayed in blue) at a location with forecasts, and the corresponding time series of forecast volumes (displayed in red) for the same period of record.

When including forecasts in the Bootstrapping Historical/Synthetic Basin-wide Events sampling method, the forecast time series must be specified in a particular manner so that models following a hydrologic sampling alternative in an HEC-WAT compute do not interpolate a forecast volume between forecast dates incorrectly. To ensure a correct interpolated value, the time series data must contain a value for both the beginning and the end of the period each forecast will be used. 

The forecast timeseries record identified here is used as the observed volume for the purpose of generating a forecast.  The Hydrologic Sampler does not compute the observed volume from the corresponding hydrograph timeseries mapped in the Hydrographs tab.  While in most cases the observed volume should match that computed from the hydrograph, it is not a requirement.

For example, the period of record graph displays an example historical historical hydrographs from continuous snowmelt runoff (blue lines), and forecast (red lines) time series for a period of record (1966 through 1981).

Another example in the abbreviated graph illustrates the forecast time series requirements for two types of forecasts.

The abbreviated graph includes both a close-up view of a representative three years of the forecast time series displayed in the period of record graph (red time series), and an example of a different type of forecast (green time series). In the abbreviated graph, the red time series shows monthly forecasts for March 1 through June 1 for which each forecast targets the April through July streamflow volume. Within the forecast period, the true value of every forecast is thus constant. Conversely, the green time series displayed in the abbreviated graph contains monthly forecasts for March 1 through July 1 for which each forecast targets the streamflow volume from the date of the forecast through the end of July (called "date – July"). Specifically, for the green time series (in the abbreviated graph), the true value of each forecast thus decreases by the volume of the previous month, within the forecast period.

To further illustrate the forecast time series requirements, the tabulated forecast time series provides the data graphically displayed in the period of record graph. Specifically, the tabulated time series for location BLNO7, which corresponds to the red time series (in the period of record graph), lists a constant forecast volume (12,070.7 KAF) for March 1 through July 31. In comparison, the tabulated times series for location SUEK7, which corresponds to the green time series displayed in the period of record graph, lists the March forecast volume (e.g., 12,567.5 KAF) for March 1 through March 31, then changes to a lower value on April 1. Specification of both the beginning and ending date is necessary for forecasts so that the sampled forecast value is accurately identified for any day within the forecast period of interest (e.g., March through July).

To specify the DSS time series that defines the forecast volume for specific locations, from the Hydrologic Sampling Editor, click the Historical Record tab. From the Historical Record tab click the Forecasts tab, from the DSS input table, users must specify the DSS time series that defines the historical forecast volume for each selected forecast location listed in the DSS input table. Note that each specified forecast time series must have a volume value for every forecast date of every year in the period of record graph. Review the Sampling Forecast Uncertainty page for instructions on adding forecast locations.

To complete the Forecasts tab of the Historical Record tab:

1. For each Location (forecast) listed, a DSS path must be identified (DSS File and pathname Parts A – F). To define the DSS record for a location of interest, click on a cell of the row of the forecast location.
2. Click Select DSS Path, the HEC-DSSVue dialog box will open. Refer to the HEC-DSSVue User's Manual (available online, here: HEC-DSSVue Documentation) for more information regarding DSS pathnames and HEC-DSSVue.
3. From the HEC-DSSVue dialog box, click ; the Open HEC-DSS File browser will open. Browse to the location of a DSS file (.dss) that contains the time series of interest (e.g., hydrographs). Click on the appropriate .dss file, the File name box now contains the name of the selected file (e.g., Bluestone_Forecasts-Historical.dss).
4. Click Open, the Open HEC-DSS File browser will close. The HEC-DSSVue dialog box now displays the DSS records for the selected DSS file.
5. For the location (e.g., Bluestone Inflow Near Glen Lynn) that was selected in the DSS input table in the Hydrologic Sampling Editor, find and select the proper pathname (record) from the list of DSS records in the HEC-DSSVue dialog box, click Set Pathname. From the Hydrologic Sampling Editor, the DSS input table now contains the DSS File pathname (Parts A – F) for the selected Location (e.g., Bluestone Inflow Near Glenn Lynn).
   
6. Select the next location, from the HEC-DSSVue dialog box, find and select the proper pathname (record) from the list of DSS records in the HEC-DSSVue dialog box, click Set Pathname. Repeat this step until all locations in the DSS input table contain the correct DSS filename and the DSS pathnames.
7. Click Apply, this saves data and keeps the Hydrologic Sampling Editor open for further modifications. When finished, click OK to save the selected hydrologic sampling alternative and close the Hydrologic Sampling Editor. Data Check (review Hydrologic Sampling Editor Interface, Data Check section) can be used to search for data consistency error(s) prior to closing the editor.