Following are special attention notes, indicating that:

If you have any questions, contact riverware-support@colorado.edu

The computational subbasin Prior Appropriation method (in the Water Rights Allocation category) was changed to no longer set Slot Inflows to zero at the beginning of run. The introduction of user specified execution times for Object Level Accounting Methods (OLAMs) makes this automatic setting unnecessary and undesired.

See “Prior Appropriation No Longer Sets Zeros For Unknown Slot Inflows” for more information on the model changes necessary.

The category Instream Flow Reference Level on the control point was incorrectly implemented as an Object Level Accounting Method. It has been converted to a normal method category. If you had user-defined methods in this category, they will be deleted when loading the model in 6.0. Please move them to another category or re-implement them as necessary in RiverWare 5.2.10 before moving to 6.0

You are now able to specify when Object Level Accounting Methods (OLAMs) execute. The choices are:

To facilitate the selection of the execution time, the Accounting Methods are now displayed in a separate “Accounting Methods” tab on the Open Object dialog.

A new compiled OLAM has been added to Reaches, Control Points, and Reservoirs. The Copy Slot to Slot Inflows method copies the object’s local inflow to the target account’s Slot Inflow and sets the other accounts’ Slot Inflow to zero. The target account is specified in a list slot called Target Account. See “Copy Slot to Slot Inflows” in Accounting for more information.

The multi-object method selector was enhanced to allow you to better select OLAMs. Now the OLAM categories are shaded grey. From the multi-object method selector, you can also set the OLAM execution time and for the Copy Slot to Slot Inflow method, you can set the Target Account for many objects at once. See “Selecting Object Level Accounting Methods (OLAMs)” in User Interface for more information.

The default category/policy group names for the object level accounting methods were changed as shown in the following table. If you customized the name of any of these, they will be changed back to the default.

A new category on the diversion object and OLAM policy group was added: “Diversion Object Account Slot Inflow”. Two compiled methods are available from the Diversion Object’s Accounting Methods tab: No Method and Zero Slot Inflows. See “Compiled Accounting Methods” in Accounting for more information

The category Instream Flow Reference Level on the control point was incorrectly implemented as an Object Level Accounting Method. It has been converted to a normal method category. If you had user-defined methods in this category, they will be deleted when loading the model in 6.0. Please move them to another category or re-implement them as necessary in RiverWare 5.2.10 before moving to 6.0

In the Supplies Manager, double-clicking a supply opens the Edit Supply dialog where you can rename and configure that supply. Previously, this operation opened the Edit Account dialog for the destination account.

In the Supplies Manager, the menu operation File Copy List to Clipboard copies a "tabular" string representation of the currently displayed supply list to the System Clipboard. Fields are delimited with tab characters and lines are delimited with new line characters -- suitable for importing into a spreadsheet, such as Microsoft Excel.

You can now specify the format for supply names in the Supply Name Format dialog. Once configured, the supply name format is used for all new supply names. Also, you can apply that format to one or more existing supply names in the Supplies Manager. See “Supply Name Format” in Accounting for more information.

In the accounting system, the SolveWaterRights and SolveWaterRightsWithLags RPL predefined functions were copying more values than necessary. Now, the function copies values in the range: [current timestep, current timestep + maximum subbasin lag] This change can have a large impact on performance. For example, it led to a 30% decrease in total run time for the 5 year daily model used to test this change.

The computational subbasin Prior Appropriation method (in the Water Rights Allocation category) was changed to no longer set Slot Inflows to zero at the beginning of run. The introduction of user specified execution times for Object Level Accounting Methods (OLAMs) makes this automatic setting unnecessary and undesired.

To reproduce previous results you need to select the Zero Slot Inflows method on each object in the subbasin and ensure that it is set to execute at Beginning of Run (default for this method). You can use the Multi-Object Method Selector to perform this selection on many objects of the same type.

You may also wish to consider using the new Copy Slot to Slot Inflow method (Reservoirs, Reaches, Control Points) which can be used to automatically copy local inflows into the allocatable flow’s slot inflow at the beginning of run or beginning of timestep. This prevents having to input or set via rules these values on the Slot Inflows.

One note, currently, most OLAMs, (including Zero Slot Inflows and Copy Slot to Slot Inflows) will fail (and stop the run) if it attempts to overwrite an INPUT slot inflow. In this case, you will need to either remove the input and let the method set it or use a different method to set Slot Inflows.

Accounting period can be changed so that it doesn't cover the run period for a single run, causing inconsistencies when the model is run more than once. A check was added and will issue an error if the accounting period does not encompass the run period.

In the Accounts Manager, the menu operation File Copy List to Clipboard copies a "tabular" string representation of the currently displayed account list to the System Clipboard. Fields are delimited with tab characters and lines are delimited with new line characters -- suitable for importing into a spreadsheet, such as Microsoft Excel.

The following categories are only applicable to accounting. Now, their visibility is dependent on having one of the accounting controllers selected in the Run Control dialog:

No model changes are necessary.

It is now possible to create two or more supplies between the same two accounts directly from the workspace. Previously, you could only do this from the downstream account. Now, when you create a supply using the accounting view's right-click menus, if there is already a supply between the two accounts you are now presented with a confirmation dialog asking whether you wish to continue. A unique name will be generated for the new supply.

At the beginning of the run, non-input values are cleared from accounts over the accounting period. Previously, if return flows are lagged from a Diversion Account, the timesteps prior to the accounting period are not then solved to populate the lagged return flow values that are at the beginning of the accounting period, effectively leaving a hole in the accounting data for return flows. Now, after the outputs have been cleared, the account resolves and fills in the missing lagged data at the start of the run.

The RiverWare Command Language (RCL) GetSlot and SetSlot commands were enhanced to support Scalar Slots. The commands have the following syntax:

The <date time> is only required for Series Slots. Also, GetValue now returns the value in user units (it previously returned it in internal units). SetValue expects the value in user units (as before). For more information, see “GetSlot” in Automation Tools.

RCL scripts can now set environment variables for use within the script and within the RiverWare run. See the Batch Mode section “Environment Variables” in Automation Tools for more details.

In Excel Database DMI datasets, environment variables are now allowed in paths.

Because Excel limits worksheet tab lengths to 31 characters, the interface for Excel Database DMI datasets was modified so that Excel worksheet tab names are limited to 31 characters. In addition, the Excel dataset execution was modified so that if slot names are in an orientation to be written to a tab, the name length is checked and if it exceeds 31 characters, the slot is not written and the user is given a message and advised to use a Name Map to map to a shorter name for writing the slot to Excel.

All input DMIs called from RPL can import values only to unlinked series slots on data objects.

Database DMIs set values and the output flag when executed from RPL. Note, input control file-executable DMIs executed from RPL still set the input flag on the imported values.

A separate installation package is available for 64 bit operating systems. With this version of the executable, you are able to take advantage of the increased memory of the 64 bit system. For more information on performance and why you would need 64 bit, see “Consider Running on a 64-bit Architecture” in Debugging and Analysis.

Additional information was added to the Help  About Show System Info dialog regarding your license including, Type, Node-locked or Floating, and days until expiration.

New RiverWare licenses will be created by the Reprise licensing system. Old licenses (FlexLM) will continue to work but will be replaced by CADSWES when they expire.

A new icon was added to describe model files. This icon is displayed where model files are listed such as the File Reopen menu. Further, files that end in the “.mdl” suffix will display this icon in windows explorer. Double-clicking the icon will open the latest installed version of RiverWare and load that model. Because of the general nature of the “.gz” extension, this file association has not been made. In windows, you can manually associate “.gz” files with RiverWare so that RiverWare will automatically open when you double-click a “.gz” file. Alternatively, you can right-click on the “.gz” file and use window’s “Open With” command and choose RiverWare.

On windows, the RiverWare installation setup file no longer sets the RIVERWARE_HOME_XXX environment variable. The relevant information is now automatically determined at run time. Other applications that explicitly access this variable should be modified.

On the Model Run Analysis grid, a separator was added to distinguish pre-simulation timesteps from simulation timesteps. A solid blue divider line separates the initial and start timestep. Also, timesteps before the start time are denoted with a blue dotted line in their timestep label cell. See the screenshot to the right for a sample.

Following are miscellaneous changes to this dialog:

For iterative MRM runs, the rules executed before the first run were renamed “Pre-MRM Run Rules” while the rules executed after each run have were renamed “Post-Run Rules”. This was done to avoid confusion between these rules and the new Initialization Rules “Initialization Rules”.

Editing of MRM configurations was modified to no longer have two tiered user confirmation. The apply (green check) and cancel (red x) buttons have been removed from the multiple run control dialog. When you click OK or Apply in the MRM configuration dialog, the edits are immediately applied.

RiverWare was upgraded to use CPLEX 12.2.

The Diagnostics Manager dialog was modified to give you control over optimization diagnostics. The previously disabled Optimization button was removed while an Enable All Optimization Diagnostics toggle was added. When checked, this gives you all optimization diagnostics.

All non-default settings of optimization parameters are now saved in the model file. As a result, the "Save non-default settings to parameter file" checkbox was removed from the "Set Optimization Solver Parameters" dialog. Further, RiverWare no longer makes use of the following files:

If you had been using one of the override files, cplex.init.par or goals.init.par in the optimization working directory, and want to continue to use these non-default values with a given model (we don't expect this to be the case), then you will need to load the model, interactively set parameters to non-default values as necessary, then save the model.

The following new Optimization “goal” parameters were added:

Following are miscellaneous changes to the optimization solution:

Performance improvements to RPL are described “Performance”. Performance improvements to water right accounting are described “Performance improvement”. Following are general performance improvements:

When you have many slot or account dialogs open, the run can slow down as these update. Now, the dialogs are only updated every 0.8 seconds while running and 0.1 seconds while not-running. This improves the performance of runs with slot dialogs open, especially accounting models. In one accounting model, having one Edit Account dialog open would run in 30 seconds while it ran in only 11 seconds when the dialog was closed. Now the run takes 12 seconds if you have the dialog open.

The representation of strings within RiverWare was re-implemented. In the course of this work, performance improvements were made. In one large test model that had a lot of list processing and lists of strings, there was a 24% decrease in run time. Another model showed about 10% reduction in run time.

A Copy Graph and Paste Graph feature was added to allow you to move graphs (i.e. individual plots) amongst layouts. The Graph  Copy Graph menu allows you to copy the selected graph to the plotting clipboard and then Graph Paste Graph menu allows you to paste that graph to a new location. These two menus allow you to quickly combine or separate multi-layout graphs. For example, you may have two great 1X1 plots configured as desired. Now you wish to combine those into a 2X1 graph so that you can see them on a linked x axis. Select a 1X1 graph and copy it. Then create a new 2X1 graph and paste the first graph onto one of the empty graphs. Repeat the copy for the second graph and then paste it into the empty graph in the 2X1 layout.

NaNs within a data series were plotted misleadingly. NaNs are now represented as a gap in the plot curve.

You can now open a slot directly from a plot. Select the curve legend and choose Open Slot(s)...

A new “Plot Template” device was added to the output manager. This utility allows you to:

The plot template utility is described in “Plotting Templates” in Output Utilities and Data Visualization.

Any type of plot-able slot can now be added to the Slot List in the Plot Page accessible from the Output Manager. Previously, only series slots were listed. Further you can now open a slot directly from this slot list by double-clicking the slot name.

For the output manager's single run RDF and CDF output, the timesteps field name in the header was changed from "timesteps" to "time_steps" to match the name in MRM output as well as the documentation.

In the plot page dialog, you can specify whether to include snapshot data in plots. Previously, this was an all or nothing operations, you either showed plots of all snapshots or none. Now, snapshot slots be shown in plots on a per-snapshot selection. This is controlled through checkable items in the snapshot selection area.

Support for inline comments within expressions was extended in several ways:

RPL editors have panels which are optionally displayed. The display of these optional dialog items are controlled via checkboxes in the View or Expression menu. For example, the Rule Editor dialog allows the user to optionally show or hide the text description associated with that rule.

An additional row of checkboxes was added to these dialogs to provide more convenient control over the display of the most commonly viewed optional items. In addition, frames around each of these checkboxes provides a visual indication of whether or not the hidden item contains non-default content.

See “Working With RPL Dialogs” in RiverWare Policy Language (RPL) for additional information.

Initialization rules were added to RiverWare to allow you to execute logic to “set up” a run, set default values, or set initial conditions. Initialization rules are saved in the model file in a separate set that can be accessed from the workspace Policy  Initialization Rules. These rules have the following properties:

For more information, see “Initialization Rules” in Solution Approaches.

The palette button labels were simplified and made more consistent. The palette is shown to the right. Full documentation of the palette is provided “RPL Palette” in RiverWare Policy Language (RPL).

In the RiverWare Policy Language, the equality and inequality operators (== and !=) were extended to all data types including BOOLEAN, SLOT, OBJECT, and LIST. Previously, only DATETIME, NUMERIC, and STRING were supported. Expressions can now be written more efficiently. For example, the following expression checks if the looping variable res (of type OBJECT) is the object named BigRes. Previously, you would have written:

Now, it should be re-written:

This change improves the performance and readability of the expression.

The RPL language provides seven operators for accessing list items, one for each type which might be returned, i.e GET NUMERIC @ INDEX <num> FROM <list>, GET BOOLEAN @ INDEX <num> FROM <list>, etc... These operators were replaced with a single operator for which the type of the item being accessed remains unspecified. The new operator has the following syntax:

An example is shown to the right. In this screenshot, triplet is a LIST containing three items: {slot, datetime, numeric}.

This new operator has the following advantages:

The runtime performance of the RPL Debugger was improved; in the sample model the time decreased by an order of magnitude. Although enabling the RPL Debugger will slow down most models, this improvement makes the RPL debugger much easier to use for many real applications.

General RPL performance was improved by modifying the processing of numeric values, list values and units.

New predefined functions called ElevationToAreaAtDate, StorageToAreaAtDate, and StorageToElevationAtDate were added for use with the changing Elevation Volume and Elevation Area methods described “Changing Elevation Volume and Elevation Area Methods”. These functions behave identically to the non “AtDate” versions but allow you to specify at which date the conversions should be performed. This datetime is required only when the Changing Elevation Volume / Area methods are selected.

The following predefined functions no longer require a fully specified datetime:

They evaluate correctly as long as they have the required information. For example, GetMonth will work if the month is specified. Thus, you can now successfully call GetMonth(@“Current Month”).

Additional checking and warning messages were added to the hypothetical simulation predefined functions which iterate over individual hypothetical simulations to achieve a desired result. These warnings make it easier to identify when algorithmic assumptions have been violated which might affect the integrity of the results. Documentation for the group of hypothetical simulation functions was also improved.

See “Hypothetical Simulations” in RiverWare Policy Language (RPL) for additional information.

Functions to compute the "PercentRank", “PercentRank” in RiverWare Policy Language (RPL), and "Percentile", “Percentile” in RiverWare Policy Language (RPL), were added to the RiverWare Policy Language (RPL) predefined functions.

Previous Storage is no longer registered as a dependency by getMaxOutflowGivenHW, getMaxOutflowGivenInflow, getMaxOutflowGivenStorage, getMaxReleaseGivenInflow, and getMinSpillGivenInflowRelease.

When a rule encounters an invalid value during rule execution, rule execution is halted, the rule finishes ineffectively, and the rule will execute again only after one of its dependencies changes. As an alternative, a property was added to rules which allows the user to change this behavior for rules that are not expected to encounter an invalid value. When the Stop On NaN property has been set for a rule (using the Rule  Stop on NaN menu), invalid values cause the simulation to halt with an error message, allowing the user to investigate the cause of the unexpected invalid value.

See “Stop On NaN” in RiverWare Policy Language (RPL) for additional information.

For RPL Set Editor dialogs, RiverWare now makes a distinction between closing the window (dialog) versus closing the set. For previous versions of RiverWare, closing a RPL set editor window also closed the set. Following is a description of the two behaviors:

Closing a RPL Set Editor window closes the editor for the RPL set, but does not close the set itself. Closing a set window is easily reversible: selecting the appropriate item in the Workspace Policy menu will reopen the editor.

From within RiverWare, this action is initiated from the main menu of a RPL set editor by selecting File  Close Window (Ctrl+W). This closes that window without affecting any other window.

In addition, the set editor can be closed using whatever mechanisms are provided by the operating system's window manager for closing windows. For example, Windows operating systems support closing a window when the user selects on the red X button in the window's upper right corner.

Closing a set removes that set from RiverWare's internal memory, and so is not easily reversible. This action is initiated from the set editor's menu by selecting File  Close (and Unload) Set. The user is then presented with a confirmation dialog pointing out that unsaved changes to the set will be lost and giving them the option to cancel the operation. This behavior has not changed, though the "Ctrl+W" shortcut for this action has been removed.

Sets that are saved within the model file can not be closed directly; to remove these sets from memory the user must clear the workspace or load a new model. Thus the only types of RPL sets which may be deleted via the RPL Set Editor's File menu are:

Note that clearing the workspace or loading a model implicitly closes any existing sets, and in this case the user is presented with a confirmation dialog analogous to the one described above.

Each application of RPL now has a unique color associated with it. For those sets always saved with the model file, the icons and color bar is always the same. For RBS, Optimization and Global RPL sets, the editors associated with a single set will have the same color, but that color is different for each set. The line color for loaded RBS sets remains red (to match the icons). For the loaded optimization set's editors, the line is purple, to match the goal set icons.

The RPL Layout Editor was renamed the RPL Display Settings dialog and was re-implemented to more intuitive and easier to use. The dialog can be accessed from the Set  Display Settings menu on any RPL dialog. See “Formatting—Display Settings” in RiverWare Policy Language (RPL) for more information.

Further, within expressions, calls to user-defined RPL functions can now be shown in a user-settable color. The color is configured in the display settings dialog shown to the right.

RPL sets that are saved in the model file now have a model file icon displayed next to the title bar.

When a RPL expression accesses a value on a slot, there are now several options for how the slot value is represented internally:

The RPL slot value representation scheme has two primary impacts:

A command line argument was added to control the RPL slot unit representation. the argument is "--rplslotvalunits" and it requires an argument as listed in the above table

Using the default “mixed” units will reproduce results in your model but you may consider using the “user” option for diagnostic and/or accuracy purposes.

The RPL unit grammar was also extended to allow parsing of the caret character to indicate "raised to the power of". This brings the input and display syntax into agreement. For example, in the past a value was displayed as"m^2" but that was not legal input; now it is.

On the run control dialog, the button label was changed from Step to Init. When you click the Init button, the run proceeds through the initialization phase and pauses before the first timestep. Once paused, the button the says Step again.

When there are initialization rules defined and you have either the simulation or rulebased simulation controller selected, the Run Control dialog now allows you to specify whether you wish to execute Initialization Rules. Check the box to execute the rules, uncheck the box to skip execution.

SCT

On the Groundwater Storage object, Wetted Sand Evaporation method, three slots were added to model Soil Limited Evaporation. These slots are optional inputs; without inputs, the previous functionality is used.

If the user inputs the "Soil Limited Evaporation Elevation", an alternative equation may be used if the wetted area excluded is zero (reach above flow is zero) and the previous elevation is less than the Soil Limited Evap Elevation. Then evaporation is the

Soil Limited Evap Rate * Soil Limited Evap Factor * Wetted Area.

Otherwise, the existing equation is used.

Functionality was added to allow the elevation vs volume and elevation vs area relationships on the storage and level power reservoir to have data that changes at various times within a run. When trying to match historical results, these methods can be used to simulate a re-survey of the reservoir bathymetry that occurs periodically. A new method, Time Varying Elevation Volume was added to the Sedimentation Calculation category. Also, a new category and method called Surface Area Modification and Time Varying Elevation Area method were added. In both of these methods, the user provides the survey date as a column header on a table. The column then represents the new data to use until the next column in the table becomes valid. For more information on using these methods, see “Time Varying Elevation Volume” in Objects and Methods and “Time Varying Elevation Area” in Objects and Methods.

The slot Convergence Percentage is now used correctly as a percentage. This tightens up convergence in many spill computations. Some models may now abort with an error when before they did not.

The following method and category names on the reservoir objects were changed to be more consistent:

In the PeakBasePowerCalc, if the outflow was less than the first value on the Base Flow Table, it would use a very small value. Now it issues an error if the lookup value is not on the table. This change could lead to a new error condition. If you want to match old results, you can insert 0 outflow and 0 baseflow as the first row of the table.

For certain inputs, the unregulated spill near the spillway crest was not correct. Note, this change can lead to slightly different numerical results. See “Unregulated” in Objects and Methods for more information on the change.

A new category, Modified Load Evaluation, was added to the Thermal object. Within this category, the Calculate Modified Load method subtracts the net hydropower generation from the System Load. Also, within optimization, it allows you to write constraints to level the load to the extent possible.

For scalar slots, table slots, table series slots and periodic slots, a Clear Values operation was added to the Edit menu.

Enhancements and bug fixes were made to editing of values in the Open Slot, SCT, and Edit Account dialogs. Following is a list of changes that may impact you:

Previously, the “i” flag was called the Computed Input flag. Because it is used by Iterative MRM and the new Initialization rules, “Initialization Rules”, which both perform model setup functionality, the flag is now referred to as the Initialization flag.

The List Slot, Scalar Slot, and Scalar Expression Slot were re-designed to remove the Ok/Apply/Cancel editing paradigm. Now, the values are applied immediately. In addition, on the two scalar slots, new buttons were added to configure the slot and edit the description (on data objects). Also, on newly created scalar slots on data objects, the default is now NaN instead of 0.0.

Import/Export now works correctly for scalar slots.

Previously, it was impossible to remove element objects from the subbasin and save/load the model. Now, when an element object is removed from a subbasin, it is no longer automatically added back when the model is saved and loaded. Also, when the object selector is used to add objects to a subbasin, aggregate elements are only added to the subbasin if they are explicitly selected.

It is now easier to add dividers to an SCT. Two menus were added to the Slots menu to allow you to Insert Slot Divider or Append Slots Divider. As before, the selector can also be used to add dividers. The slots menu wording was modified because of this change.

Previously, interpolation on the SCT was only possible for one slot at a time. Now, Interpolate Selection is possible for a rectangular selection of slots. The interpolation is still performed across timesteps and is separate for each slot. See “Set Multiple Values” in User Interface for more information.

An Edit  Adjust Slot Values operation was added to the SCT. This operation allows you to change multiple selected slot values by a percentage or an absolute value (if they all have the same units).

On the SCT, the keyboard accelerators were modified as follows:

A new Geospatial View canvas was added to the RiverWare workspace. This canvas allows you to add and configure a georeferenced image file and then specify georeferenced coordinates for each object. This canvas also has enhanced display settings for the size of objects and label location, font and color.

For more information, see “Geospatial View” in User Interface.

Two enhancements were made to the background image support for the Simulation View only. These enhancements are not relevant for the new Geospatial View, which implements distinct support for a background image.

For more information, see “Background Image” in User Interface.

Note that the Geospatial View will not support independent re-scaling of the image's horizontal and vertical dimensions, nor independent horizontal and vertical scales for real-world coordinates. Also the Geospatial View will not support map images rotated with respect to compass direction. Considering those limitations, some users may prefer to continue to use the Simulation View with a background map image.

The following bugs have been closed for this release. For more information on any bug, see the CADSWES website. The bugs are listed in order by bug number: