CA MPA counterfactuals

[brunj7]

Define areas with similar characteristics to MPAs to test hypotheses about the attractivity of MPAs.

• Need to compile the layers needed
• Compute distance from those assets (need to choose raster or vector approach)
• Need to combine those layers and extract Characteristics for each MPA
• Need to compute areas that are similar to MPAs characteristics but are not MPA

[brunj7]

Potential metrics:

Metric	Source
Total area (sqkm)
Average distance from shore (km)	CDFW coastline layer (distance from centroid)
Average depth (m)	CDFW bathymetry layer
Total length of sandy beach (km)	CA ROMS model layer
Park area (sqkm) within 500 m	CA Parks parks shapefile
Number of people within 50 km	US Census data
Number of public beaches	Coastal Commission public beach layer
Proximity to Public Beach access points	Beach access points

Other datasets State Parks:

• Campgrounds
• Parking
• Entry

[brunj7]

sf::st_distance() and raster::DistanceFromPoints() could be options to compute the distances metrics

[brunj7]

The following metrics (raster layers) were processed:

Metric	Source
Total area (sqkm)	MPA shapefile
Average distance from shore (m)	Global Watch Fisheries
Average depth (m)	CDFW bathymetry layer
Total length of sandy beach (m)	CA ROMS model layer <--- not processed yet waiting for the new dataset
Distance from closest park entry point	CA Parks parks entry points
Number of park entry points within 600m	CA Parks parks entry points
Number of people within 50 km	US Census data
Distance from closest public beaches	Coastal Commission public beach layer
Number of public beaches	Beach access points
Distance from closest Public Beach access points	Beach access points

The Script processing those metrics is here: https://github.com/NCEAS/ca-mpa/blob/main/data/counterfactuals_rasters/counterfactuals_raster_processing.R

[brunj7]

Statistics processing script: https://github.com/NCEAS/ca-mpa/blob/main/analyses/3performance_human/code/counterfactuals_mpa_stats.R

This script creates a raster cube with all the tiff layers processed above and computes statistics for the MPA and transforms this raster cube into a large data frame.

Outputs:

• mpas_counterfactuals_stats_epsg3309.geojson data frame with the stats for the MPAs
• counterfactual_layers_epsg3309.Rds raster cube transformed into a data frame
• counterfactual_layers_shallow_epsg3309.Rds is the same but filtered for depth above the deepest pixel found in an MPA (aka smaller file)

All of those data files have been saved on the team's Google Drive: data/sync-data/counterfactuals

[brunj7]

Also added rasterization rasters of the MPAs matching the bathy grid:

• binary file (1 == MPA): mpa_200m_epsg3309.tif
• ID file (grid value matches MPA ID): mpa_200m_id_epsg3309.tif

updated the RDS files accordingly: f70949

[brunj7]

Chris and Cori have conducted the counterfactual analysis. Closing the issue