v5.14.0 release notes¶
Clawpack 5.14.0 was released on January 26, 2026. See Installing Clawpack.
Permanent DOI: https://doi.org/10.5281/zenodo.18382457
Changes relative to Clawpack 5.13.1 (Sept. 12, 2025) are shown below.
To see more recent changes that are in the the master branch but not yet released, see Changes to master since v5.14.0.
Changes that are not backward compatible¶
A couple bugs were discovered in the GeoClaw code from v5.13.0 and v5.13.1, so GeoClaw users should be careful using results obtained with those.
The most serious bug fix is a change to the main GeoClaw Riemann solver, discussed further in the riemann section below. The pre-v5.13.0 version has been restored.
The ranking algorithm for ordering the topofiles provided by the user has also changed slightly, and some changes might be observed if the user specifies two topofiles with essentially the same resolution but that differ slightly due to rounding errors in the dx and/or dy values. See geoclaw section below for more details. Also, the new feature introduced in v5.13.0 to allow more control over topo file ordering did not work properly at all and has now been fixed.
General changes¶
None
Changes to classic¶
None
See classic diffs
Changes to clawutil¶
Makefile.common was updated to include compiler flags for the intel ifx compiler.
minor bug fixes and other changes.
See clawutil diffs
Changes to visclaw¶
A few enhancements.
See visclaw diffs
Changes to riemann¶
The Riemann solver in geoclaw_riemann_utils.f was modified to revert one change back to the version from v5.12.0 and earlier. This change was made in commit 4340757 and released in v5.13.0 to improve the order of accuracy on some test problems, but then it was found to give streaks in velocity on some practical problems. Reviewing the git history, this was seen earlier, and the code that we are reverting to in this commit was added in 2016 (in commit 9c02ede) to fix this problem.
Other minor changes.
See riemann diffs
Changes to amrclaw¶
Fix example $CLAW/amrclaw/examples/advection_2d_inflow.
Introduce binary output option for gauges in 1D.
See amrclaw diffs
Changes to geoclaw¶
The modification of the Riemann solver discussed above in the riemann repository section may give different results.
The new feature introduced in v5.13.0 to allow more control over topo file ordering, described at Topography file ordering, which is used by setting the new setrun.py parameter to:
rundata.topo_data.override_order = True
did not work as advertised and would have ordered topo files in the opposite order than what is indicated in the documentation. Moreover, if one or more dtopo file is specified, a more serious problem arose, since special topo_for_dtopo topography arrays are introduced at the resolution of each dtopo file. These were not necessarily being properly ordered with topo_data.override_order = True.
A refactoring of the ordering algorithms has addressed this problem. The topo_data.override_order now only controls the ordering of the original topo files specified in setrun.py, and then the topo_for_dtopo files are inserted in the ordered list. For more discussion see
As part of the refactor mentioned above, the ordering of topofiles may also change in the case where topo_data.override_order = False but two files at essentially the resolution are specified. If the resolutions are exactly equal then the one listed last in setrun.py is given preference, but if their resolutions (as judged by dx*dy) is slightly different due to rounding errors in dx and/or dy, then the one listed first might have been used if it had slightly smaller dx*dy. The new code introduces a fudge factor so that the resolution is considered better only if the area if the difference in dx*dy is greater than some relative tolerance (set to 0.001) times the smaller area.
This change could lead to different results from GeoClaw than past versions (even when not using the topo_data.override_order option), if it means a different topo file at a nearly equal resolution is now given priority. This should only happen if the user has nearly-equal resolution topo files that cover an overlapping region and would only be a problem if their topo values Z differ signficiantly in this region.
As part of the refactor, the information about the ordering of topo and topo_for_dtopo files that is written to _output/fort.geo has been clarified, and users are encouraged to check this file to insure that topo files are being given the desired order of priority.
The module center_points.py was improved and a new utility $CLAW/geoclaw/src/python/geoclaw/center_gauges.py was added, to center gauges or other specific points in grids cells at a specified resolution.
Arguments buffer and align were added to the topotools functions crop and read_netcdf, to allow user to insure a buffer of points surrounds the desired extent, and to subsample (when coarsen > 1) with the starting index chosen so the resulting points are aligned with specified values if possible. This is useful if the original DEM has points aligned with integer longitudes and latitudes, for example, and this properly should be preserved for the coarser grid (i.e. the distance between points and an integer longitude or latitude is an integer number of grid cells on the coarser grid).
kmltools.make_input_data_kmls improved so it doesn’t die if a topofile or dtopofile is not found (it just does not make a kml file showing the extent), useful if doing make data on a laptop for a problem where large topofiles only exist on a supercomputer, for example.
Add make_function methods to both topotools.Topography and dtopotools.DTopography to easily compute an interpolating function that can be used to evaluate along a transect, for example.
Several other more minor bug fixes and enhancements.
See geoclaw diffs
Changes to PyClaw¶
Better command line interface for gauge comparisons.
See pyclaw diffs
Other Clawpack Repositories¶
The repositories below are not included in the Clawpack tarfile or pip install, but changes to these repositories may also be of interest.
