II.5.9 - RUN Button

• scenario - runs CRiSP.1 for a single year
• monte carlo - opens the tool for controlling Monte Carlo runs
• realtime - used in-house for production of in-season estimates of fish arrival distributions; this mode requires several additional files that will not be available to the user, so we do not recommend attempting to use it.
• settings - allows the user to select different functional relationships describing how predation rate changes with pool volume and fish migration rate.

The Scenario Mode is used for determining migration and survival for a single year at a time. The model can be run in the Scenario mode using its default settings. Normally however, the default settings are replaced by reading the base.dat file with the Input/Output Tool. Then release sites are selected and CRiSP.1 is run. The results can be viewed in the passage windows, and statistical results can be viewed in the message window or in the summary.dat file.

Scenario Mode procedure

2. Read parameters into CRiSP.1
   • Drag a base.dat file into the load/save window using the file manager.
   • Click the Read button to load data into CRiSP.1 then click
     • replace in most cases, or
     • append when adding new releases while still retaining sites from a previous run. Append does not affect other model parameters.
3. Select release sites

   New release sites can be identified by placing the mouse arrow on release sites on the River Map and opening a release window with the appropriate mouse button (right by default). See River Description file (columbia.desc) section II.2.3 for further information.

4. Select Passage Points

   CRiSP.1 computes daily fish passage on a release-specific basis through all river segments and dams. To view and retain information on survival and travel time, open passage windows (Fig. 63) by left-clicking on the desired locations on the River Map.

5. Select flow observation Point

   A representative river flow over the travel time of a release can be obtained by opening a flow output window at a selected location. the average flow will be reported in the message window information. Only one flow point can be computed at time.

6. Select Messages from the Message Tool
7. Left-click scenario to run the model.

   The progress of calculations will be displayed at the top of the map.

Fig. 63 Passage histogram at Bonneville Dam

Fish survival and migration time at specific points along the river is displayed in the passage histograms and in the message window after the run completes. Left-click desired river segments, or at a dam, to see passage histograms which illustrate the number of fish passing a river segment or dam each day, the cumulative number that eventually pass over the whole season, and percent survival on a species basis.

Scenario Mode results

Survival and travel time information for all selected points are written in the file summary.dat and in the message window which is opened from the I/O button menu. To display information, click on the Messages button. The user may click off other buttons. Messages (Fig. 64) include:

• first day of initial release first day of passage
• last day of passage
• number passed through this reach or dam
• mean passage day
• median passage day
• mode passage day
• standard deviation of passage
• average flow at a site while fish are migrating
• total number of fish released
• number of fish transported to a site and released
• number of fish that failed to migrate out of the system during the model run (censored fish)
• percentage of release survival through the site

Fig. 64 Scenario Mode run results written to the message window and the Summary.dat file

Monte Carlo Mode

The Monte Carlo Mode can run CRiSP.1 with one or more different alternatives. An alternative is a file containing flow file names, the number of games (runs), the number of years, and the name of the input file to use for each year in the alternative. The Monte Carlo window (Fig. 65) has four panels. From top to bottom, these are:

• Alternative control panel
• Flow Archive control panel
• Yearly Input control panel
• Status panel

Fig. 65 Monte Carlo window

Alternative control panel

This panel of the Monte Carlo window provides ways to select alternatives to run from the disk or to create a new alternative. It has the following controls:

• Working directory identifies where to work
• Alts on disk list shows which alternatives are available
• Alts to run list shows which alternatives will be run
• The Transfer Arrow moves a highlighted alternative between disk and run lists
• Remove button to remove a highlighted alternative from a list without affecting the other list
• New button - used to create an alternative. To store a new alternative, highlight and transfer it back to the Alts on disk list
• Rename button - used to change the name of a highlighted alternative in Alts to run list.

This panel controls the flow file for a highlighted alternative in the Alts to run list. It has the following controls:

• Flow from: - toggle switch selects if flows comes from flow archive file or the yearly input file. Left-click desired box to select.
• Flow Archive - line identifies the file containing period averaged flows and spills at dams as computed by a hydro model such as HYSSR or HYDROSIM¹.
• Planned spill from - toggle switch selects spill information from period averaged data in Flow Archive File or from hourly spill in Yearly Input File. Left-click desired box to select.
• Elevation from - toggle switch selects reservoir elevation information from period averaged elevation in Flow Archive File or from daily elevation in Yearly Input File. Left-click desired box to select.
• Customize games - opens a list of games in the Flow Archive File by their water year (Fig. 66). Identify number of times to run each game by keyboard input or clicking on the increment arrows.

Fig. 66 Sample Customize Games window

Yearly Input File control panel

This panel deals with yearly input files in a highlighted alternative. These files set all model information other than flow. It has the following controls:

• Number of yearly input files in the alternative is set with the increment boxes or by keyboard input on the line
• Input file (Year #) - name is entered for each year to be run in an alternative.
• Edit control file button - opens a window to construct an input file from individual subfiles by data category. Details of this process are given in the next section.

The Control File Editor (Fig. 67). allows the user to read or write data as subfiles of the base.dat file. In this manner, groups of data can be replaced while other elements are unchanged.

Fig. 67 Control File Editor is invoked through edit control file in the I/O Button or through Edit control file in the Monte Carlo frame

The editor has the following functionality:

• Buttons to read and write all data ¹
• Year: Each control file is referenced to an alternative and is stored along with other information for the alternative ²
• Control files ³ contain a list of subfile names (Table 6) which are identified in the lines below the Control File line ⁴. Control files are designated by the suffix .ctrl.
• Subfiles ⁴ are actual subgroups of data and are identical to grouping stored with the I/O Tool. Each subfile has a predefined suffix (Table 6) that is automatically attached if selecting default names. Variables within each data group are listed in the edit control file: Management of Files section II.5.6.
• Default control file names can be created by typing a file name in the Control File line. Then by clicking Defaults button ⁵ corresponding subfiles with suffixes appear on subfile lines.
• Subfiles are not required to have the same base name. The utility of using the control editor is to mix subfiles from different alternatives. Each specific combination of subfiles can be read or written as a control file using the Read and Write buttons ⁶.

Status panel

The status panel starts the Monte Carlo run and gives information on the status of the run. It has the following controls

• run monte carlo button - starts the Monte Carlo run using the information in the Alts to run list
• Alternative Games identifies the number of actual model runs to be completed
• Alternative Years identifies the number of years to be run
• Alternatives identifies the number of alternatives to be run
• write supplemental data check box - if checked, writes messages window information to disk.

The steps to set up a Monte Carlo run are outlined below using the four panels of the Monte Carlo Tool window described above.

Set up alternative

2. Right-hold the RUN button in the CRiSP.1 control panel, and drag the mouse down to choose monte carlo.
3. If creating a new alternative left-click the new button to create an alternative.
4. Rename the new alternative if desired with the rename button.
5. Type in the name of the Flow Archive file.
6. Customize games in the Flow Archive file.
7. Select source for spill and reservoir elevation information.
8. Determine the number of years in the alternative.
9. Type in a Yearly Input file for each year.
10. If using Control files as yearly input,
    • Type the name of the control files with a .ctrl suffix
    • Left-click the Edit Control File button.
    • Set up the names of the subfiles. Note that the subfiles can be created from an existing yearly file. This is done by loading existing yearly file into memory using the load/save tool under the I/O tool and store each subfile using the categories under the write function. For further information see edit control file: Management of Files section II.5.6.
11. Store new alternative by highlighting and moving back to the Alts on disk list.

2. To run one or more alternatives highlight each desired alternative in the Alts on disk list and move it to the Alts to run list.
3. Select the choice of spill schedules from the yearly input file or flow archive file.
4. Select if supplemental data is required. (This produces a considerable amount of information in the message window).
5. Open passage and flow output histograms on main map locations where passage and flow² information is desired. CRiSP.1 only saves information from sites identified by open histograms.
6. Left-click the run monte carlo button.
7. The user may close CRiSP.1 into an icon while it is running by closing the main panel. Do not quit while the model is running.

1. To run one or more alternatives in batch mode, you must first have created the alternatives, specifying the flow archive, data file(s), and so forth. These alternatives must then be written to disk.

2. A second (and invisible) file must also exist in the same directory where your CRiSP.1 binary, and all the alternatives, reside. This file must be named .crisp1-alts. It is a list of the names of the alternatives to be run, one on each line, thus:

   my.first.alternative

   my.second.alternative

   ...

   This file can be created using any text editing tool (e.g. vi or emacs).

3. Once you have a number of alternatives created, and a .crisp1-alts file in place, CRiSP.1 can be run in batch mode from a terminal which would otherwise not be able to run the model graphics (i.e. a "dumb" terminal). It is run from the command line prompt by issuing the command:

   crisp1 -b -r columbia.desc

   where the "-b" flag tells CRiSP.1 to run in batch mode, and the "-r" flag tells it what river description file to use; in this case, "columbia.desc". The model will run all the alternatives listed in the .crisp1-alts file before concluding.

Viewing results

2. To save survival and travel time information for any point in the river, the appropriate passage histogram windows must be open prior to running the Monte Carlo games.
3. The average flow experienced by fish between the mean release date and mean passage date to any point in the river can be saved by opening a flow window for the point where average flow is desired.

This mode of operation is used by CRiSP Project staff to generate predictions of passage timing and survival for use by system managers; the output from these model runs is available via the World Wide Web.

When the model runs in this fashion, it takes output from another model (PIT Forecaster, developed by J. Skalski) to generate a "release" of fish at Lower Granite Dam. These fish are then tracked down the river and this information is written to a special directory and output file for later processing.

Settings: Different Model Configurations

CRiSP.1 can be run with different functions that affect the movement and survival of fish through the river system. Settings are selected or deselected by left-clicking the respective check-box (Fig. 68).

Fig. 68 Run-time settings

The check boxes are described below.

Variance suppression

This setting turns off all stochastic parts of the model: when it is selected, each subsequent run of the model uses exactly the same value for dam mortality parameters, fge, etc. that would otherwise be drawn from a distribution. This is useful for doing quick tests of sensitivity. By default, it is turned ON.

Predator Density / Volume interaction

This setting increases the effective predator density as pools are lowered which in turn increases the predation rate. By default, it is turned OFF.

Age dependent FGE

This setting varies FGE at dams according to the settings in the FGE equation window. FGE varies with fish age, which is set in the release window. Deselecting this function causes FGE to be selected from FGE sliders. For a tutorial demonstrating fish guidance dynamics see Dam Individual Based Model (DamIBM) Tutorial. By default, this option is turned OFF.

² Only one flow point may be identified.

Columbia River Salmon Passage Model CRiSP.1.5 User Manual
Copyright © 1996, Columbia Basin Research. All rights reserved.