This month in AvaFrame - January 2026 edition

Posted by Felix on Sunday, February 1, 2026

Welcome to the January 2026 update:

January brought changes to simulation naming, improved time dependent release handling for debris flow simulations, and added validation checks for AIMEC coordinate system transformations.

Module Name in Simulation Names

PR #1226 (https://github.com/OpenNHM/AvaFrame/pull/1226 ) added the computational module name as a component of the simulation name. The format changed from relNameSim_simHash_defID_frictIndi_simType_modelType to relNameSim_simHash_modName_defID_frictIndi_simType_modelType. For example, a simulation previously named release1_a1b2c3_C_S_ent_dfa now becomes release1_a1b2c3_com1DFA_C_S_ent_dfa. This makes it easier to track and distinguish simulations from different modules. A centralized parser handles both old and new formats for backward compatibility.

Time Dependent Release for Timestep 0

PR #1215 (https://github.com/OpenNHM/AvaFrame/pull/1215 ) improved the time dependent release option (formerly called “hydrograph”) in com1DFA. The feature now works at timestep 0 and no longer requires a separate release area in the REL folder. The terminology was renamed from “hydrograph” to “time dependent release” throughout the codebase. Input polygon and CSV files are now located in the REL folder, and time dependent values are saved in the configurationFiles folder. A new example data setup and standard test were added. This PR closes several DebrisFrame issues related to debris flow input handling.

AIMEC Coordinate System Validation

PR #1225 (https://github.com/OpenNHM/AvaFrame/pull/1225 ) added validation checks for the AIMEC coordinate system transformation in ana3AIMEC. When defining a new coordinate system along the thalweg, excessive curvature can cause overlapping regions in the transformed coordinates. The new checks detect self-intersecting domain boundaries and overlapping coordinate lines, issuing warnings when problems are found. Standard tests show no differences with this change.

Build System and Dependency Updates

PR #1229 (https://github.com/OpenNHM/AvaFrame/pull/1229 ) updated version pinning in pyproject.toml. Cython was updated to >=3.0.10 (needed for Python 3.12/3.13 support), pyshp was updated to >=3.0.3, and the salib version pin was removed. The pixi clean task was rewritten to avoid platform-dependent shell commands, and the documentation was updated to reflect that pixi handles the Python installation, so a separate Python install is no longer required.

MacOS Multiprocessing Fix

PR #1231 (https://github.com/OpenNHM/AvaFrame/pull/1231 ) removed the MacOS-specific workaround that forced ThreadPool usage instead of the standard multiprocessing Pool. Both com1DFA and com4FlowPy now use the same multiprocessing approach on MacOS as on Linux, based on feedback from MacOS users. The PR also includes minor code formatting changes in com4FlowPy.

In other news we took part in the “BFW-Praxistag Innsbruck” with a “Marktstand” and our ISeeSnow Paper finally found an editor and can be viewed here: https://egusphere.copernicus.org/preprints/2026/egusphere-2025-6053/

See you in February!

Felix