Skip to content

Latest commit

 

History

History
100 lines (67 loc) · 4.1 KB

README.md

File metadata and controls

100 lines (67 loc) · 4.1 KB

This subdirectory contains the preproccing steps file for the project.

This list all single time preprocessing steps required for a particular domain

Table of Contents

Step1

The first step is to generate an overall mask using a set of different masks and the following procedure

It implements the following 5 step masking logic

  1. "mask_schism_boundary_atlantic.shp","exterior"
  2. "mask_state_boundaries_conus.shp","exterior"
  3. "mask_levee_protected_area_conus.shp","interior"
  4. "mask_nwm_lakes_conus.shp","interior"
  5. "mask_water_polygon_conus.shp","interior"

A sample of folder path containing different masks:

output

This can be achieved by executing zonal_fim.py and activating generate_mask flag while other flags are set to False also one can alter the name of masks shapefile names in case they have changed by these flags:

--water_table_name
--nwm_table_name
--levee_table_name
--state_table_name
--schisim_table_name

Several paths must be specified including:

-k '/path/shape_file_folder' that is the folder path where the schisim element shape file lives
-l '/path/output_elements_folder' that is the folder path where the schisim element parquet and geopackage will be saved to
-a '/path/masks.duckdb' path to the mask database that will store the generated overall mask
-v '/path/atl_orig_id.csv' path to the crosswalk table between gr3 and netcdf node ids

and there is an optional item to dissolve all geometries into a single multi-polygon feature that can be set but recommend the default True value

--dissolve

python zonal_fim.py --generate_mask True --preprocess False --generate_wse False --generate_depth False --zarr_format False  --execute False  --dissolve True -k '/path/shape_file_folder' -l '/path/output_elements_folder' -a '/path/masks.duckdb' -v '/path/atl_orig_id.csv' -p 'exterior_mask_schism_boundary_atlantic_buffer_atlgulf' -n 'exterior_mask_state_boundaries_conus_atlgulf' -x 'interior_mask_levee_protected_area_conus_atlgulf' -t 'interior_mask_nwm_lakes_conus_atlgulf' -f 'interior_mask_water_polygon_conus_atlgulf'

Step2

Next step is to generate coverage fractions from zonal in R

This step requires:

  1. the ElementPolygons.gpkg file generated from preprocessing step 1
  2. A TB DEM grid

The following two steps needed to generate cell coverage fractions:

  1. Write the mask file to .gpkg By executing the generate_zonal_req.ipynb we add the mask layer to the ElementPolygons.gpkg

  2. Generate coverage-fraction and elevation values for all cells The weights_mask_elev_table.R file then does the following

    1. Takes the triangles and TB dataset to compute the relative coverage fraction of each cell and polygon ID.
    2. For each cell, the cell the area is determined as mask or not, and an elevation is assigned

    The result is a table that contains the following columns:

    > head(fin)
    pg_id      cell coverage_fraction elevation  mask
    <num>     <num>             <num>     <num> <num>
1: 122881 720766867       0.023299048  5.497267    NA
2: 122881 720801591       0.552611947  5.367274    NA
3: 122881 720801592       0.040903941  5.823075    NA
4: 122881 720836314       0.006548808  4.383194    NA
5: 122881 720836315       0.748685956  4.613204    NA
6: 122881 720836316       0.334290981  4.148118    NA

Step3

Final step in preprocessing is to generate barycentric weights for all none-masked Schisim node. This can be done by calling the script with the following configuration.

Setting all flags to False except for
--preprocess True 

python zonal_fim.py --generate_mask False --preprocess True --generate_wse False --generate_depth False --zarr_format False  --execute False -u '/path/ElementPolygons.parquet' -o '/path/TBDEM_AtlanticGulf_Mosaic_NWM_3_Revised_v4_COG.tif' -i '/path/agGridfile.gr3' -c '/path/schisim_database.duckdb' -v '/path/atl_orig_id.csv' -w '/path/coverage_fraction.parquet'