V 24.11

Add a new option that allows GRAL to produce reproducible results (thanks to @JoshLovesFun for idea, development and testing)
Fix a bug that caused the top line of a *.gff file not to be read

V24.04 Service Release

• Fix a bug that has been causing concentrations in the building since version 23.11 for flat terrain and certain building configurations
• Fix a bug that leads to high initial wind speeds near the ground when transferring the GRAMM wind profile to the GRAL Grid
• Correct the plume rise calculation, as since version 23.11 the plume rise is too high for very small point sources
• Enable the optional usage of AVX512 processor extension for flow field calculations

V23.11

What's Changed

• Moved to .NET8.0
• The GRAL release is no longer published as an all-in-one file by default This means that the user must install the .NET8 Runtime and therefore benefits in several ways:
  -- Smaller published files
  -- Fewer false warnings from anti-virus programs
  -- Better performance
  -- Individual setting options for the Runtime for the respective computer in the *.runtimeconfig.json file
  -- Same compilation for Windows, Linux and macOS
  -- For Windows users, a published version as a single-file is still available in a separate download
• Revison of the Plume Rise algorithm
  The exhaust plume now rises a little more at the source and is released earlier. This fixes small programming flaw and better validation results are achieved.
• The reflection algorithm has been revised
  It could happen that individual particles remained in an infinite loop in the reflection algorithm. This effect led to a performance drop in large projects.
• The result files xxxxx.grz are written in a separate thread
  For very large projects, writing the zipped result files can take several minutes. With this change, writing takes place while the next wind field is being loaded or calculated, which increases overall performance.
• Store original meteopgt.all stabilityclass
  The displayed stability class of the original meteorology was incorrect in some cases.
• The used memory is now released at the end of calculation, even if the console window is not closed
• Particles that move very slowly over a longer period of time or are trapped in a cell are sorted out

Download pre-compiled binaries: https://gral.tugraz.at/

Full Changelog: V2209...V23.11

V22.09 Release

1. An error that occurred when the operating system reported 0 free processor cores, resulting in a division by 0, has been fixed
2. Fixed a breaking change of .NET6 when submitting the number of available processor cores
3. Added additional status messages for the user
4. Improve the reading and writing performance for binary files

The compiled binaries are available on GitHub:

V22.09 binaries for GRAL, GRAMM and the GUI

V22.03 Release

Bug fixes

• Fixed an error which occurred when the operating system reported 0 free processor cores, resulting in a division by 0
• An incorrect concentration evaluation above buildings for calculations with flat terrain was fixed
• In the transient calculation mode with flat terrain, building heights were included as terrain in the 3D concentration file

V21.09 Release

GRAL Improvements

• Optimized computational accuracy for the prognostic flow field calculation
• Reduced memory requirement
• Optional deactivation of the online output (reduction of file accesses during the calculation)
• Deposition within vegetation areas: by default, GRAL increases the deposition velocity within vegetation areas by a factor of “1.5 * Coverage” for gases and the particle classes PM2.5 and PM10 and by a factor of “3 * Coverage” for particles PM30 and larger. Coverage is a value that can be entered for each vegetation area. Now it is possible to replace these fixed factors 1.5 and 3 with your own factors. Thus, in combination with the coverage value, a spatial varying deposition velocity can be set. If you want to specify your own factors, you must create a file named “VegetationDepoFactor.txt” in the Computation folder. The first line specifies the factor for gases, PM2.5 and PM10, the second line the factor for PM30 and larger. The user defined deposition velocity for each source is subsequently increased by the value “Factor * Coverage” within vegetation zones. If GRAL can read these user settings, you will find the imported values in the terminal output and in the file "Logfile_GRALCore.txt".
• Output of a file “PrognosticSubDomainAreas.txt” containing the prognostic sub-domains areas. This file is created when building_heights.txt is also written.
• Optional reduction of prognostic sub-domains depending on the distance to sources: by default the prognostic sub-domains are created around buildings using the building height multiplied with the sub-domain factor. You can check the prognostic sub domain size using the new output file “PrognosticSubDomainAreas.txt”. If the new option is used, prognostic subdomains are created around sources at a defined distance only. With reduced sub-domains, memory requirements and computation time are reduced. If large contiguous prognostic wind fields are required, as in urban areas, this option should not be used. This option can be set in the GUI under Special Options (Wrench Button).
• Several performance improvements by using the latest .NET5 high performance instruction sets
• Linux: compiled for .NET5 as single file applications.
• Start GRAL or GRAMM using ./GRAL or ./GRAMM or using the Linux GUI. The .NetCore or .Net5 Framework no longer needs to be installed for Linux when using the self-extracting executables. It is still possible to use the operating system independent *.dll files if the .NetCore Framework is installed

Bug fixes

• Fixed the rasterization of the vegetation mesh
• A slight overestimation (about 20 %) of the deposition rate has been fixed

Full Changelog: V20.09...V21.09

V20.09 release

Final release of V20.09