mpas_tools.mesh.mask.compute_mpas_transect_masks

mpas_tools.mesh.mask.compute_mpas_transect_masks(dsMesh, fcMask, earthRadius, maskTypes=('cell', 'edge', 'vertex'), logger=None, pool=None, chunkSize=1000, showProgress=False, subdivisionResolution=10000.0, addEdgeSign=False)[source]

Use shapely and processes to create a set of masks from a feature collection made up of transects (line strings)

Parameters:

  • dsMesh (xarray.Dataset) – An MPAS mesh on which the masks should be created

  • fcMask (geometric_features.FeatureCollection) – A feature collection containing features to use to create the mask

  • earthRadius (float) – The radius of the earth in meters

  • maskTypes (tuple of {'cell', 'edge', 'vertex'}, optional) – Which type(s) of masks to make. Masks are created based on whether the latitude and longitude associated with each of these locations (e.g. dsMesh.latCell and dsMesh.lonCell for 'cells') are inside or outside of the transects in fcMask.

  • logger (logging.Logger, optional) – A logger for the output if not stdout

  • pool (multiprocessing.Pool, optional) – A pool for performing multiprocessing

  • chunkSize (int, optional) – The number of cells, vertices or edges that are processed in one operation. Experimentation has shown that 1000 is a reasonable compromise between dividing the work into sufficient subtasks to distribute the load and having sufficient work for each thread.

  • showProgress (bool, optional) – Whether to show a progress bar

  • subdivisionResolution (float, optional) – The maximum resolution (in meters) of segments in a transect. If a transect is too coarse, it will be subdivided. Pass None for no subdivision.

  • addEdgeSign (bool, optional) – Whether to add the edgeSign variable, which requires significant extra computation

Returns:

dsMask (xarray.Dataset) – The masks