1 water

water

Read global surface water from https://global-surface-water.appspot.com/

https://doi.org/10.1038/nature20584

1.1 Example:

gsw = GSW(agg=8) Create water mask mask = gsw.sel(latitude=lat, longitude=lon, method=‘nearest’) > 50

1.2 Classes

Name	Description
GSW	Global surface water reader
GSWLandMask	BlockProcessor that adds GSW water occurrence data and raises a flag

1.2.1 GSW

water.GSW(l1=None, lat=None, lon=None, directory=None, agg=8)

Global surface water reader

1.2.1.1 Args:

lat: tuple of (min, max) latitude bounds lon: tuple of (min, max) longitude bounds directory: directory for tile storage agg: Factor of aggregation

1.2.1.2 Returns:

A xarray.DataArray of the water occurrence between 0 and 100

1.2.2 GSWLandMask

water.GSWLandMask(
    l1=None,
    lat=None,
    lon=None,
    directory=None,
    agg=8,
    threshold=50,
    flag_var='flags',
    flag_name='LAND',
    flag_value=1,
)

BlockProcessor that adds GSW water occurrence data and raises a flag where water occurrence is below threshold.

Arguments: l1: Level-1 dataset for cropping GSW data (passed to GSW) lat: latitude bounds tuple (passed to GSW) lon: longitude bounds tuple (passed to GSW) directory: directory for GSW tile storage (passed to GSW) agg: GSW aggregation factor (passed to GSW) threshold: water occurrence threshold in % (default: 50) flag_var: name of the flags variable to modify (default: “flags”) flag_name: name of the flag to raise (default: “LAND”) flag_value: bitmask value for the flag (default: 1)

1.3 Functions

Name	Description
generate_full_gsw_mosaic	Build a lazy dask array representing the full global surface water mosaic
list_tiles	Return list of latlon acronym
read_tile	Read a single tile as a dask array

1.3.1 generate_full_gsw_mosaic

water.generate_full_gsw_mosaic(directory, agg)

Build a lazy dask array representing the full global surface water mosaic

1.3.2 list_tiles

water.list_tiles()

Return list of latlon acronym

1.3.3 read_tile

water.read_tile(tile_name, agg, directory)

Read a single tile as a dask array

Data is accessed on demand