v.4.13.0 Release Notes

Initial beta release: TBD

Final version release: TBD

New Features

Grid to Point Additional Statistical Time Series

In previous versions of the Grid to Point utility, users could only extract a polygon average from a gridded dataset. Now users can extract the following statistical time series, in addition to the average: minimum, maximum, median, 25th percentile, 75th percentile, percentage of cells greater than 0, and percentage of cells greater than the 25th percentile. 

This feature was funded by the FEMA Mixed Populations work unit. Initial code implementation and documentation were completed by Mike Bartles and Avital Breverman. Testing was completed by Mike Bartles, Avital Breverman, and Tom Brauer.

Available Webinar

This feature was presented in a webinar. 

User-specified Solar Shading Radius

In previous versions of HEC-HMS, solar shading computations for snowmelt modeling were computed for every cell in the terrain. Now, users have the option to specify a radius over which solar shading computations are performed. When the distance between 2 grid cells exceeds the user-specified radius, cell 1 does not contribute to solar shading of cell 2 and solar shading computations are not computed (as shown in the image below). If the radius field is left blank, the program will consider all grid cells in the terrain when computing solar shading coefficients.

This feature was funded by USACE Arid Regions R&D work unit. Initial code implementation and documentation were completed by Mike Bartles and Avital Breverman. Testing was completed by Avital Breverman, Lauren Coe, and Tom Brauer.

Change ET method for meteorologic model from the Parameters menu

The meteorologic evapotranspiration method is administered globally for all subbasins in a meteorologic model and has been traditionally been set using the meteorologic model editor panel. An additional option has been added to the Parameters menu to provide similar program operation to the Change Method dialogs of the basin model methods.

This feature was funded by USACE Hydrology, Hydraulics, & Coastal Science Engineering and Technology. Initial code implementation and documentation were completed by Thomas Brauer. Testing was completed by David Ho.

Bugs Fixed

The following bugs were present in previous versions and have been fixed within version 4.13.

Multiple Compute option disabled when project contains only Uncertainty Analyses

When a project contained Uncertainty Analyses and no other simulation types, the Compute | Multiple Compute option was disabled. The Multiple Compute option is now enabled when a project contains only Uncertainty Analyses. This issue was introduced prior to v4.4.1 and fixed in v4.13-beta.1.

Menus not properly reactivated when closing a graph or a table

Appropriate menus would not re-activate when closing an internal frame (graph or table displayed inside the desktop), such as basin element result graph. This issue has been present since version 4.4 or earlier. The issue was fixed in 4.13-beta.1.

Dates in 1936 Increment by 1 hour when Saving

When a Control Specification, Forecast Alternative, etc. utilized a Start and/or End date between 01Mar1936 and 15Nov1936, the user's computer was set to the Central time zone, and the Save button was clicked, the Start and/or End date incremented by 1 hour.  A fix was instituted to prevent the Start and/or End date from unnecessarily incrementing. This issue was introduced prior to v4.0 and fixed in v4.13-beta.1.

Only the first Ensemble Analysis is computed in Multiple Compute option

When users selected multiple Ensemble Analyses in the Compute | Multiple Compute option, only the first Ensemble Analysis was computed. This issue was introduced in v4.11 and fixed in v4.13-beta.1.

Changes in Reservoir Volume Reduction Results

There are some differences in calculated deposition when the sediment only contains clay and silt, and the deposition shape is set to taper for the other grain types. There was an error in the previous code that attempted to calculate the change in surface area for tapered deposition even when there was no sediment deposited in the taper. The effect was that the surface area was less than it should have been and affected the trap coefficient calculated by Chen Sediment Trap.  This issue was in the original implementation of reservoir volume reduction and fixed in v4.13-beta.1.

Results Lost When Resuming Optimization

When an optimization is resumed after increasing the maximum number of iterations, in some cases, the prior parameter and objective function values are lost. This issue was fixed in v4.13-beta.1.

Error 10000 when Changing Meteorology Method

An error occurred when a run has been selected and a meteorology method is changed in the component editor. This issue was fixed in v4.13-beta.1.

Paired data table is removed upon rename if a duplicate name exists

When a paired data table was renamed to a name that already exists, the table was removed. Instead, the original table and name is restored and an error message is displayed. The paired data names are case-insensitive. For example, attempts to rename a table to "alpha" if another table named "Alpha" already exists will result in an error. Renaming the same table to change the case (for example from "alpha" to "Alpha") is allowed. This issue was present prior to v4.4 and was fixed in v4.13-beta.1.

Frequency Storm depths do not migrate properly from 4.10 to post 4.10

When migrating Frequency Storm depths from version 4.10 and below to a later version of HMS, the depths would not migrate properly if the duration was missing for the lowest depths.  The original code assumed the frequency storm depths needed to be included from the intensity duration to the storm duration and therefore the migration would not occur properly if there was missing depths for the intensity duration.  This issue was present from version 4.11 and fixed in v4.13-beta.1.

Precipitation time series were not outputted when using MetSim precipitation method with a lumped Basin Model

When applying the MetSim precipitation method to a lumped Basin Model, rainfall-runoff computations were performed but the precipitation time series (Precipitation, Cumulative Precipitation, Precipitation Standard Deviation, and Precipitation Loss) were not outputted to the results within the user interface or to the Meteorologic Model HEC-DSS file. Users can now view precipitation time series in the program and in the Meteorologic Model DSS file. This issue was introduced in v4.12 and fixed in v4.13-beta.1.

Error in uncertainty results when output time interval is less than simulation time interval

If uncertainty time interval was increased after setting output time interval for selected time series, the mean, minimum, maximum, etc time series were not correct if the output time interval was less than the simulation time interval. This issue was introduced in v4.10 and fixed in v4.13-beta.1.

Uncertainty Analysis stats tables do not follow precision settings

The uncertainty analysis results statistics tables displayed values with varying precision, often to 6+ decimal places. Values are now displayed according to the program settings. This issue was introduced prior to v4.4 and fixed in v4.13-beta.1.

Distribution parameters sometimes referenced previously selected parameter types

In prior versions of the Uncertainty Analysis, distribution-specific parameters could still reference the previously selected parameter type, if changed. Since allowable user input ranges are based on parameter type, bogus errors could appear when HEC-HMS checks allowable sampling ranges for a given parameter type during the beginning of a compute. Added observer logic now allows distribution parameters to track when the user changes the selected parameter type within an uncertainty analysis. This issue was introduced prior to v4.4 and fixed in v4.13-beta.1.

Handle all caps text parsing when deleting old time-series results

When computing a simulation run, HEC-HMS attempts to delete old time-series results from the simulation DSS file. In some cases, a pathname with a d-part string in all caps would cause the simulation to abort prematurely. Logic was added to handle all caps text strings and parsing exceptions. This issue was introduced in v4.11 and fixed in v4.13-beta.1.

Failure to compute daily time-series for Interpolated Air Temperature and Interpolated Shortwave Radiation

Failure to compute daily average/minimum/maximum time-series for Interpolated Air Temperature and Interpolated Shortwave Radiation. This impacts any process that relies on daily average/minimum/maximum time-series and is combined with with interpolated temperature or shortwave radiation, for example Hamon ET combined with Interpolated Air Temperature, Hargreaves ET combined with Interpolated Air Temperature, Penman Monteith ET combined with Interpolated Air Temperature, Hargreaves Shortwave Radiation combined with Interpolated Air Temperature, and RTI snowmelt combined with Interpolated Shortwave Radiation. This issue was introduced in 4.12 and fixed in 4.13-beta.1.

This correction will change computed results.

Terrain-projection combination duplicated when reprojecting basin

The Improved Terrain Data Management feature added in v4.11 manages terrain for unique terrain-projection combinations. When reprojecting a basin model there is a possibility that concurrent threads could create a duplicate subdirectory for a given terrain-projection combination. This issue was only observed when reprojecting a basin model and was not observed to impact program operation. This issue was introduced in v4.11 and fixed in v4.13-beta.1.

Check tile west coord against screen max X

For some basin location/projection combinations the hillshade draw layer would fail to render even though the hillshade draw option was enabled on the raster display properties dialog. This issue was attributed to a mismatch in handling logic for X and Y coordinates. This issue has been present since raster rendering was added to the basin map. This issue was introduced in 4.4 and fixed in 4.13-beta.1.

Error 10000 on some forecast runs in the CAVI

HEC-HMS Error 10000 was being generated by some forecast runs in the CAVI. This was being caused by a null pointer exception in the paired data table module. This issue was first reported in version 4.12 and has been fixed in version 4.13-beta 1. 

Relativized Map Layer dropped during rename

Non program-managed map layers such as shapefiles can be added to the basin map since v4.3. A bug was introduced in v4.7 where relativized, non program-managed map layers, such as a shapefile internal to the project directory, were dropped when a project was renamed. The program was failed to track and update the relative reference during the project rename. This issue was introduced in v4.7 and fixed in v4.13-beta.1.

Project Rename does not completely remove old project files

The Project Rename feature failed to completely remove old project files. This issue was introduced in 4.0 and fixed in v4.13-beta.1.

Project Rename drops terrain data

The Project Rename feature drops terrain data references such that terrain data disappears after rename. This issue was introduced in v4.12 and fixed in v4.13-beta.1.

Precision handling for Parameter Value Samples

Previously for Parameter Value Sample input, a large amount of precision was assigned to each input value. A systematic approach was implemented for handling Parameter Value Sample precision. Precision is approximated for each value based on the user-specified input value. This issue has been present since Parameter Value Samples Paired Data was introduced. This issue was introduced in v4.3 and fixed in v4.13-beta.1.

Vortex: date/time handling for non-temporal rasters

The Vortex Importer can be used to convert non-temporal rasters to HEC-DSS format. For example, convert an ASC or GeoTIFF raster with no temporal information can be converted to a gridded data record with no D or E part in HEC-DSS. An issue was fixed that prevented this conversion due to due to assumptions around the temporal qualities of the input data. This issue was introduced in Vortex v0.11.10, HEC-HMS v4.12-beta.6 and fixed in Vortex v0.11.15, HEC-HMS v4.13-beta.1. 

Vortex: failure to write DSS records when user locale is not English

Many of the Vortex utilities would fail to write DSS records if the user locale settings were not English. A change was made to always assume English locale when parsing HEC-DSS dates and times, regardless of the user locale setting. This has been an issue in all previous versions of Vortex and fixed in Vortex v0.11.13 and HEC-HMS v4.13-beta.1.

Vortex: Importer Target Cell Size Units

In the Vortex Importer utility, geoprocessing step, the UI formerly indicated that the target cell size was to be provided in units of meters. In actuality, the units of the target cell size corresponded with the horizontal units of the target projection, or the projection of the input data if no target projection was specified. A change was made that allows the user to specify the target cell size units as "Meters" or "Feet". This issue has been present for all prior vortex versions. A fix has been made in vortex v0.11.15. This issue was introduced in v4.7 and fixed in v4.13-beta.1.

Vortex: Add support for Stereographic Azimuthal projection

Add support for Stereographic Azimuthal projection in Vortex Importer. This has been an issue in all previous versions of Vortex and fixed in Vortex v0.11.15, HEC-HMS v4.13-beta.1.

Vortex: HEC-DSS grid processing failure when undefined time represented as empty string

When working with HEC-DSS grids in Vortex utilities, an undefined start time represented by an empty string could cause processing failures. Previously an undefined start time was represented by the string "31 December 1899, 00:00". In more recent HEC-DSS files, undefined time is represented by and empty string, i.e. "". Handling logic for both cases has been added to the Vortex code. This issue has been presented for all prior versions of Vortex corresponding to HEC-HMS v4.7 and fixed in Vortex v0.11.15, HEC-HMS v4.13-beta.1

Nodes close when a new paired data table is created or an existing table is copied

When a new paired data table is created or an existing table is copied, the currently opened nodes in the selection tree get closed. This issue was introduced in v4.12 and fixed in v4.13-beta.1.

Element result graph doesn't open when parameter is selected in optimization trial

If a parameter is selected in an optimization trial, and an element result time series is selected, the times series graph fails to open. This issue was introduced in 4.11 and fixed in v4.13-veta.1.

4-hour and 8-hour time intervals failing during compute

Computes failed if 4-hour or 8-hour time interval was specified in control settings This issue was introduced prior to 4.4 and fixed in v4.13-beta.1.

Project fails to open due to a corrupt .gage file

The presence of an incomplete .gage file prohibited a project from loading in HEC-HMS. The corrupt .gage file was likely introduced due to an interruption during save. Logic was added to handle this scenario by allowing a project to load. This issue was introduced prior to v4.10 and fixed in v4.13-beta.1.

Corrupt stream threshold units trigger generic ERROR 10000 when loading terrain file

In v4.12 terrain data management was upgraded to reduce redundant raster files stored to disk. As part of that upgrade, logic was modified for storing the stream threshold for a terrain dataset. The units for stream threshold are stored as length units squared with a superscript 2. Errors have been reported where the superscript 2 stored in the project *terrain file has been replaced with the replacement character "�" or a question mark "?" depending on the formatting of the file. This can happen if the *.terrain file is opened in a text editor (or some other process) and saved to a formatting other than UTF-8. Handling logic has been added to convert the replacement character or question mark to a superscript 2 so the the terrain file can load without error. This issue was introduced in 4.12 and fixed in 4.13.

Precipitation component editors randomly disappear

In limited cases precipitation component editors could disappear when the precipitation method does not require subbasin parameterization. This issue was observed when the precipitation component editor was open while simulations were being computed. After a few repeat simulations, the component editors disappeared. This issue was introduced in 4.9 and fixed in 4.13-beta.1.

Meteorologic sampling error when grid shift applied

When discretizations were added in v4.7, a grid cell re-indexing function had to be added to rectify georeferenced grid cells with the "principal grid info" to ensure that gridded surface and loss methods could still do index-based (non-georeferenced) sampling.

As part of the re-indexing, previously unique grid cells could take on identical identifiers if they fell within the same "principal grid cell". This has only been observed to have happened when the principal grid cell size is larger than the discretization cell size. This is a valid and more efficient approach when no spatial grid shift is applied.

In v4.11 a new feature was added that allowed a spatial shift of meteorologic grids. A condition was observed when a grid shift was applied that assuming the same identifier for all discretization cells nested within a principal grid cell is no longer valid and the geometry of the grid cell, in addition to it's indexed location, had to be added to the uniqueness of the grid cell to provide accurate sampling.

In the image on the left no grid shift is applied. It is appropriate to assume all discretization cells (magenta) have the same location because they will all sample the same meteorologic grid cell (yellow). The image on the right illustrates a spatial grid shift applied. It is no longer appropriate to assume all discretization cells (magenta) have the same location. The discretization grid cells marked with a yellow rectangle will sample a different meteorologic grid cell (pink) based on their centroids. 

This change makes gridded sampling slightly less efficient for the case that no shift is applied but ensures accurate sampling for the case that a spatial grid shift is applied. 

Note: Grid cell re-indexing is scheduled to be removed as part of an effort to make sampling for gridded surface and loss methods spatially aware, not relying on index-based sampling alone. Once this transpires, the added geometry check for uniqueness can potentially be removed.

This issue was introduced in v4.11 and fixed in v4.13-beta.1.

This correction will change computed results. Conditions: discretization grid cells are smaller than meteorologic grid cells and transpose option is enabled for gridded precipitation.

Unnecessary error when travel length grid cells cannot be computed

Unstructured discretizations do not always have travel lengths computed and they are not neccesary when using an unstructured discretization with the 2D Diffusion Wave transform. Previously a simulation using an unstructured discretization without grid cell travel lengths would fail even it was being used with the 2D Diffusion Wave transform. Now checks are made in the ModClark and 2D Diffusion Wave transform methods themselves which gives an appropriate level of granularity to the grid cell travel lengths requirements. This issue was introduced in 4.7 and fixed in 4.13-beta.1.

ERROR 42631 is thrown when using a Semi-Circular Culvert Outlet

When using Rule-Based Reservoir Operations and a Semi-Circular shaped Culvert Outlet, the program would throw ERROR 42631, indicating that the Diameter parameter was not defined. When using a Semi-Circular shaped Culvert, the vertical dimension is described as Rise, rather than Diameter.  This issue was introduced prior to 4.4.1 and fixed in v4.13-beta.1.

Optimization Trial Search resumes after change to Search Method, Parameter(s), Objective Function, Population Size, and/or Seed Value

When a user made a change to the Search Method, a parameter, or the Objective Function in an Optimization Trial after computing the trial, the search would continue from the previous iteration. In addition, if a user changed the Population Size or Seed Value in a Differential Evolution Optimization Trial, the search would continue from the previous iteration. Any of these actions now cause the search to restart. This issue was introduced in v4.10 and fixed in v4.13-beta.1. Additional information on changes that will allow a search to resume are documented in the User's Manual (see the Resuming an Optimization Trial section).

This correction will change computed results. 

Inverse Distance met model - 2 nodes assigned with weights of 1 double rainfall

When creating Meteorological Model using Inverse Distance method and assigning two basin nodes, assigning a weight of 1 and 1 to both nodes doubles the rainfall volume when the node weights should be normalized to 1. This bug was introduced as far back as v4.4 and fixed in v4.13-beta 1.

Precipitation gage table values significant digits set to zero

When linking the rainfall gage to DSS file and navigating to the Table, the number of decimals will correspond to the Program Settings depth digits. When toggling back to the table, the values will be changed to zero and display no digits.  This bug was introduced as far back at v4.4 and fixed in v4.13-beta 1.

Migration path for project with deleted terrain data

v4.12 added improved terrain data management. As part of the migration process a in-tact project was assumed. This fix addresses the condition that a user deleted the contents of the terrain directory effectively corrupting the project. For this condition HEC-HMS now upgrades the project assuming that the first set of basin rasters encountered is the "base" terrain data. This issue was introduced in 4.12 and fixed in 4.13.

Beta Releases

beta.1 release: TBD

final release: TBD

The following table presents features that were added and bugs that were fixed within each beta release.