The latest release of Flood Modeller provides some key updates that directly relate to enhancing user workflows in 1D and 2D modelling projects and hydrological analyses. In addition, numerous bugs identified recently by our users, have been addressed.
WINFAP v5.3 and new FEH methodology
Version 7.4.0 introduces access to the additional FEH catchment descriptors that were introduced in 2025 (catchment descriptor datasets v5.1.0). These can be integrated into the updated peak flow analyses available in WINFAP v5.3. The new variables include:
SAAR9120: An update to SAAR6190, which describes the average annual rainfall for the 30-year period 1991-2020, based on the Met Office's HadUK gridded datasets (OGL).
FARL2015: An update to FARL, which describes the attenuation of peak flow due to reservoirs and lakes. This makes use of OS data and the UKCEH Land Cover Map 2015.
URBEXT2015: An update to URBEXT2000, which describes the proportion of urban and suburban land in catchments, making use of UKCEH Land Cover Map 2015.
BFIHOST19SCALED: A minor update to adjust BFIHOST19 which removes water bodies from the calculation. Uses the same model coefficients at BFIHOST19. Based on HOST soil classification data.
CD version: The version number of a catchment descriptor dataset is now stored as a variable within the Hydrology+ database. This version number is then displayed as part of the General data for the catchment descriptor dataset.
Full details are available from UK CEH ukceh_feh_catchment_descriptors_2025_v3.pdf
All downloaded catchment descriptors are now editable within a calculation point, and these data can then be passed through to one or multiple WINFAP analyses for each calculation point location.
Note that to update points that already have catchment descriptor datasets associated to them within Hydrology+ you will need to re-import the catchment descriptor XML file (this may need to be re-downloaded if your original copy predates the addition of the new variables). Importing again will set up a new (incremental) version of the catchment descriptor dataset associated to point, which will include all new descriptors.
If using a gauge site for your calculation point analysis, then it should be noted that you will need to have downloaded a Peak Flows Dataset version that is v14 or later. Only these later versions of the database include the new catchment descriptor variables described here. The Peak Flows Dataset should then be connected to your WINFAP software (via WINFAP Options > Peak flow dataset setting). Flood Modeller adopts the same Peak Flows Dataset as WINFAP and uses these data to extract catchment descriptor datasets at gauge sites.
The version of catchment descriptor datasets is now also recorded with each revision created. When a WINFAP analysis is initiated in Flood Modeller the catchment descriptor dataset is passed to WINFAP. If this is a v5.1.0 dataset, then WINFAP will enable both 2008 and 2025 analysis methods to select from.
Compatibility with WINFAP v5.3
The 2025 analysis methodology available within WINFAP v5.3 introduces an additional distribution curve option for your Pooling Groups. This is the “GL non-flood years” curve. Calculation points in Flood Modeller v7.4.0 will accommodate this additional result and make it available to view within Flood Modeller and utilise when reconciling hydrographs to develop model boundary data.
Hydrology+ workflows extended to allow enhanced single site WINFAP analyses
A limitation of previous iterations of Flood Modeller was the capability to perform enhanced single site hydrological analyses within calculation points (only pooling group analyses could be performed). Now in v7.4.0 when a calculation point starts a WINFAP analysis, in addition to passing in the selected catchment descriptor dataset it also provides WINFAP with the associated “.am” and “.pt” files (annual maxima and peaks over threshold) for compatible sites. Flood Modeller automatically extracts these from the latest version of the peak flow dataset associated with your WINFAP software. Thus, it should be noted that this analysis is only possible for calculation points defined at NRFA gauge sites which have the required data associated to them within the peak flows database.
Single site WINFAP analyses are stored in your Hydrology+ database (against the associated calculation point) in the same way as pooled analyses and thus can then be used to develop boundary hydrographs for your models.
Revised 2D model build interface
The second significant change in v7.4.0 is an update to the 2D model build interface. This aims to deliver a more intuitive workflow.
The main simplification has been to restrict models to a single domain. The original use of multiple domains was to solve limitations in computing power at the time. It is now assumed these limitations are rarely a problem and so models can be single domains, maybe with multiple active area polygon features, that will all use the same computational grid resolution. Thus, the inputs and outputs for Flood Modeller 2D models are simplified. Note that existing multi-domain models can still be loaded into the interface and run – but editing of anything other than the first domain must be via a text editor (to edit the 2D XML file directly).
The move to a single domain mode definition has enabled the model build interface to be simplified to a single row of intuitive tabs as shown below:
To define a new 2D model, simply work through the tabs in sequence providing relevant data in each. The tabs are defined as follows:
General – specify global model attributes such as run times, timestep, log filename, 2D solver type and whether there is a linked 1D model (specifying this activates the 1D river links tab).
2D extent – specify all topographic features (DTMs and topographic adjustments) and model extent as defined by active and inactive areas (polygon shapefiles)
Roughness data – specify global values and spatially varying values (in grid or shapefile formats, sometimes accompanied by lookup files)
1D river links – specify 1D river linked model and link line shapefiles that defines where models interact.
Boundary conditions – specify model inflows and outflow boundaries. These are located within the model extent by shapefiles with associated flow (or depth) data files, e.g. csv text files, Hydrology+ output files (csv format) or 1D river event files. Time series can also be specified to be incorporated into the 2D model file.
Rainfall / infiltration - specify rainfall, evaporation and infiltration inputs. These are located within the model extent by shapefiles with associated data files, e.g. csv text files, Hydrology+ output files (csv format) or 1D river event files. Time series can also be specified to be incorporated into the 2D model file.
1D structures – specify any flow paths that can be represented by a 1D river embedded structure, i.e. culverts, weirs or orifices. Structure locations are defined by shapefiles with associated data files that specify 1D structure properties (embedded structure tool within shapefile layer editor is used to create this data file).
Outputs – specify outputs, e.g. time series variables, maxima grids, mass balance data, point outputs, flow lines and check grids / diagnostics. Note this tab has default settings so making changes here are optional.
Note the location of 2D model outputs, i.e. results and diagnostics, has been updated to utilise a more intuitive folder structure, defined as follows:
All results are written to a sub-folder located in the same folder as the 2D model definition file (2D XML file) that takes the name of the 2D model file plus “_results”. The results folder contains 4 sub-folders and the spatial diagnostics file (extension “.sdd”).
Time series results in xmdf (or SMS) format are written to sub-folder named 2D model file plus “_2dts”.
Diagnostic check grids are written to sub-folder named 2D model file plus “_check”
Maxima grids for each variable are written to sub-folder named 2D model file plus “_maxgrids”
Mass balance data file is written to sub-folder named 2D model file plus “_rt_diag”. The mass balance default name has been updated to use the 2D model file name plus “_MB.csv” as its default name (this can be edited).
Older models and multi-domain models re-run will use domain names instead of 2D model file name in each sub-folder name.
Options – modify advanced settings related to the selected solver and adjust selected settings from their default values.
Note all 2D model setup settings related to the ADI and TVD solvers within this revised interface were present in previous versions of Flood Modeller. They have just been rearranged into a more logical and intuitive order in v7.4.
Minor enhancements
v7.4.0 delivers a series of minor enhancements that are designed to improve user workflows. The majority of these have originated from user requests for functionality or users reporting encountered issues to our Support Team. These additions to the software are summarised below:
When multiple 1D river nodes are selected (highlighted) in the network table and in the map view users can now select additional nodes or de-select highlighted nodes by depressing the Ctrl key and clicking on the relevant node in the network table.
Note clicking on a node on the map with the Ctrl key depressed will add the node to the existing selection but will not de-select nodes already selected.
The 1D model summary tool from previous versions of Flood Modeller has been reinstated to the Toolbox. It is located in the Model Review Tools > 1D River Models section (it can also be located by typing “1D model” into the Toolbox search tool).
More information during a 1D river simulation is written to the diagnostics file. The following modifications have been to the data recorded in a 1D river log file:
Lateral inflow nodes added to network check – duplicates or invalid connections are now trapped in a better way. Previously messaging on this was confusing (caused iflag=-12 errors) - now clearly reports issue to make it easier to fix.
Path to TUFLOW tlf files using shortcut (“<<…>>”) notation now picked up for linked TUFLOW simulations and added to log file
Full path to network (.dat) file location now specified – enables user to know which folder path from the hierarchy was used, if full path not previously specified
More information and error trapping provided on duplicate units (i.e. with the same node labels) appearing in the same file (e.g. boundaries, blockages, etc.)
Time outputs for W2050 warning messages (where area ratio exceeds 2) are now always written in full in the1D river diagnostics file (zzd). Previously, if this time value exceeded 100hrs it would appear as “*****” (due to the definition in the solver of this numeric value). The limit for this time output has now been increased to 10,000hrs.
Presence of invalid remote water level nodes now trapped and error output to diagnostic file
Better information on invalid event file (ief) keywords written to log (and also references to valid “dummy” keywords, e.g. LaunchDoublePrecisionVersion, removed to avoid confusion)
Warning given in log if LaunchDoublePrecisionVersion flag set in the Event (.ief) File, if run with the Single Precision version (e.g. if run from the command line/batch file)
Removal of unhelpful/unclear messages from log file, e.g. “stopped in <procedure name>” (replaced with “program stopping”) and “error at node label: <blank>” (i.e. when there is no specific node label to reference)
Ensured that ‘interpolation out of range’ error messages always output the correct node label and are more appropriate, e.g. QH CONTROL, QHBDY and WIND units would previously display misleading references
Errors in Hydrology+/csv inflow files trapped explicitly, i.e. if missing or invalid data inputs present
1D simulations can now specify a restriction to the numbers of repeated warning types written to the log file
Full information of any QTBDY multiplicative factors used are now output to the log file
Failures due to Muskingum sections containing distance to next = 0 mid-reach trapped and output explicitly to log file.
More information during a 2D simulation is written to the diagnostics file. The following modifications have been made to the data recorded in a 2D log file:
CPU specifications are now reported, i.e. username, computer name, RAM, number of CPUs and cores.
Confusing warning 5016 (water reached domain boundary) suppressed for cells with elevation links, elevation/depth boundary conditions, normal-depth slope conditions.
Link-flow summary expanded to separate multi-links with the same 1D node. In addition to the log file a new check file (shapefile) is introduced for links with 1D River and 1D Urban.
Traps added to detect buried embedded structures – these are reported in the log file.
Traps added for missing Green-Ampt index – in a simulation, Green-Ampt infiltration becomes ineffective for any cells that have been associated with a GA index (from the loss shapefile) that was not found in the GA lookup table. Such cells are now marked with the negative GA index in the loss check file and 5020 error messages are written to the spatial diagnostics file (SDD). The log file also provides a summary of the issues identified.
Issues addressed with units of check grids in US unit system.
Diagnostic files (1D and 2D) now also log the hardware used – this information can be utilised for carbon footprint calculations within modelling projects.
The workflow has been improved for inspecting the conveyance properties of 1D river sections and using panel markers to remove sudden changes in conveyance vs depth (which in turn can lead to model instabilities). In v7.4, the following process is now possible:
Select a river cross-section unit and open its properties window
Click the Plot button – the cross-section data are displayed in a chart (in a new window) together with the conveyance curve and panel markers by default (they can be easily turned off in the Legend).
At a point in the section where a panel marker is required, right-click on the cross-section plot and select the “Add/remove panel marker” menu option. The effect of the panel marker is immediately seen in the conveyance curve. An alternative option for adding a panel marker is to depress the Alt key when clicking on a cross section point.
When the cross-section changes are complete, clicking Next or Previous on the node properties window will load the properties of the next or previous node and if this is another cross section then the section plot will be displayed.
If the plot window (for the next/previous section) appears behind the main Flood Modeller user interface then use the setting in the Chart Options dropdown menu (in the chart toolbar) named “Show Plot As Top Most Window” to force the plot window to the front on your screen. Note you will need to close the plot window and re-open it to apply this change (this action will also save the current plot window size as the future default size). Once this setting is applied, the chart window will appear above all applications, so you may only want to use this option while reviewing and editing cross sections.
Note: A current limitation of this process is that to work on both the node properties window and the cross section plot these windows must be on separate screens or separate parts of the same screen.
The Flood Modeller project export zip file format has been modified to be compatible with the Flood Platform system (it is now compatible with Linux operating systems)
The Bing Maps online mapping service has been retired by Microsoft (except for corporate contracts). Therefore, all references to Bing Maps in the base maps menu and the General settings window have been removed.
Bug fixes
This release of Flood Modeller also addresses some key bugs that were reported by our users in the recent releases of Flood Modeller. These are summarized below:
Issue with 2D post-processing tools crashing when trying to access 2D results files has been addressed. The tools affected by this that are now working are:
2D flood map tool – extracts individual timesteps from 2D XMDF or SMS time series grids and outputs these to raster grid files (in geoTIFF or ASCII format) or flood extent shapefiles.
2D time series tool – plots time series from a user specified point location for one or multiple variables from one or multiple 2D simulations
2D plot section tool – plots a cross-section through your terrain data that can be overlaid with one or multiple 2D results datasets (e.g. water levels from multiple simulations)
2D flow line tool – plots flows crossing a user specified line within the model extent from one or multiple simulations.
Detection of duplicate boundary units, i.e. using the same name, in the same file (e.g. both in the network [.dat] file) is now recognised as a model error in the initialisation of a 1D river simulation. Previously simulations would continue but would always use the first such unit, which may not be as intended.
Breach trigger level in 1D river simulations now works with lateral spills. Previously, for a lateral spill, as the node attached to the river spill label does not contain a representative water level, the level used for the trigger stayed was the dummy level (-9999.99; hence never triggered). Now the unit applies the mean of the upstream and downstream river section water levels to the trigger. The breach trigger can also work in reverse order, i.e. the water level on the downstream side of the spill can trigger a breach as well as the upstream.
Issue when writing pump data to 1D river results file (ZZX) when no pump in network resolved. Solver still outputs correct data when pumps exist and are on (and output option is on) and does not crash when no pumps exist.
Improved messaging and handling of issues arising from model simulations in which the network includes a blockage unit with an empty data table. Previously, this led to a confusing node count being reported and an assumption by the solver that the model contained disconnected 1D river units or reaches. In v7.4, models with this incorrect blockage unit setup will correctly be aborted, reporting in the log file a fatal error message that details the cause relates to missing data in a blockage unit.
Enhanced definition of Green-Ampt infiltration data in 2D model build interface – the date entry process has been made more robust after users reported issues.
2D Section plot now also works with GPU XMDF scalar outputs from Flood Modeller 2D. Previously only worked with selected xmdf files.
Revised method to detect edges associated with embedded weirs in 2D solver. Now only the relevant edges are included in calculations. Previously additional irrelevant edges were selected and included in calculations, thus reducing efficiency of solver.
Improvements to robustness of Direct Method implemented to allow more 1D river models to run successfully in this mode.
Messaging related to 1D river energy junction node has been improved. Warnings are now written to the log file by the solver if supercritical flow detected at energy junction (Fr > 1).
Known Limitations
With each new release of Flood Modeller it is not always possible to address all known bugs and issues. Thus, some of the more minor issues are not all fixed. The following provides a summary of the known minor bugs and issues identified within v7.4. None of these should prevent any aspect of the software from working and we will aim to fix some or all of these during the next development cycle
When viewing TUFLOW results in Flood Modeller, the following limitations should be noted:
- “Water Elevation” results show all cells as wet, even those with “0” depth. For “0” depth cells, returned value will equal cell bed.
- 2D Flood Map tool only displays 0 and 9999 as available timesteps for post-processing.
- 2D Flood Map tool limitation in processing results from models with rotated domains.
There is a known issue with 2D embedded orifice structures that are flapped, i.e. opening type = 'FLAPPED' which does not allow reverse flows. Using this setting will lead to inaccurate results. Note that opening type = 'OPEN' (which allows bidirectional flow) is fine for use in embedded structures.