Binned Wind Climate (BWC)#

The binned wind climate contains a histogram representation of the wind for different wind direction sectors. In WAsP, this is called an “Observed Wind Climate” and is stored in .tab and .owc files.

Data Structure#

BWC objects are xarray.Dataset with the following schema:

Dimensions:  (sector, wsbin, ...)
Variables:
    wsfreq  (sector, wsbin, ...)  - wind speed frequency per sector/bin
    wdfreq  (sector, ...)         - wind direction frequency per sector

In addition to the core spatial coordinates, BWC objects contain:

wsbin - wind speed bin centers
wsfloor, wsceil - bin boundaries
sector - sector centers
sector_floor, sector_ceil - sector boundaries

I/O#

Format	Read	Write	Description
`.tab`	yes	yes	WAsP TAB format
`.owc`	yes	no	WAsP Observed Wind Climate
`.omwc`	yes	no	WAsP Observed Mesoscale Wind Climate
`.nc`	yes	via `.to_netcdf()`	NetCDF format
`.yaml`, `.yml`	yes	yes	windIO (experimental)

Reading:

windkit.read_bwc() reads binned wind climates:

bwc = wk.read_bwc("bwc.omwc", crs="EPSG:4326")

Writing:

BWC’s can be written to NetCDF using xarray’s built-in method:

bwc.to_netcdf("output.nc")

windkit.bwc_to_file() can write BWC’s to .tab and windIO YAML files:

wk.bwc_to_file(bwc, "output.tab")
wk.bwc_to_file(bwc, "output.yaml")

Creating from Arrays#

Creating from TSWC#

windkit.bwc_from_tswc() creates BWC from time series data:

# First create a TSWC
tswc = xr.Dataset(
    data_vars=dict(
        wind_speed = xr.DataArray(np.random.rand(1000)*15, dims=['time']),
        wind_direction = xr.DataArray(np.random.rand(1000)*360, dims=['time']),
    ),
    coords=dict(
        time = pd.date_range('2020-01-01', periods=1000, freq='h'),
        height = 80.0,
    )
)
tswc = tswc.expand_dims("stacked_point").assign_coords(
    west_east = (("stacked_point",), [10.0]),
    south_north = (("stacked_point",), [56.0]),
)
tswc = tswc.expand_dims("height").assign_coords(
    height = (("height",), [80.0]),
)
tswc = wk.spatial.set_crs(tswc, crs=4326)

# Convert to BWC
bwc = wk.bwc_from_tswc(tswc)

Wind Direction Frame#

The sectors of a BWC are normally relative to true north. A model that works on a projected grid may instead need them relative to grid north, which differs from true north by the grid convergence. The wind_dir_crs argument states which frame the directions are in.

windkit.bwc_from_tswc() with wind_dir_crs reprojects the time series to that CRS and rotates each wind direction from true north to its grid north before binning. The CRS is recorded in the wind_dir_crs attribute:

bwc_grid = wk.bwc_from_tswc(tswc, wind_dir_crs="EPSG:32632")
print(wk.spatial.get_crs(bwc_grid).name)
print(f"Rotation: {wk.spatial.grid_convergence(bwc_grid).item():.2f} degrees")
print("wind_dir_crs" in bwc_grid.attrs)

WGS 84 / UTM zone 32N
Rotation: 0.83 degrees
True

If the time series already has a wind_dir_crs attribute, its directions are taken to be in that frame, so they are not rotated twice.

windkit.create_bwc(), windkit.read_bwc() and windIO import accept the same argument but do not rotate anything: use it there to declare that the frequencies you have are already relative to the grid north of that CRS. A geographic CRS denotes true north.

Combining BWCs#

windkit.combine_bwcs() combines multiple BWC datasets into one:

# Create two BWCs at different locations
bwc1 = bwc.expand_dims("point").assign_coords(
    west_east=(("point",), [10.0]),
    south_north=(("point",), [56.0]),
    height=(("point",), [80.0]),
)
bwc1 = wk.spatial.set_crs(bwc1, crs=4326)

bwc2 = bwc.expand_dims("point").assign_coords(
    west_east=(("point",), [11.0]),
    south_north=(("point",), [57.0]),
    height=(("point",), [80.0]),
)
bwc2 = wk.spatial.set_crs(bwc2, crs=4326)

# Combine into one dataset
combined = wk.combine_bwcs([bwc1, bwc2])
print(f"Combined BWC points: {wk.spatial.count_spatial_points(combined)}")

Combined BWC points: 1

Validation#

windkit.validate_bwc() - Validate BWC dataset structure
windkit.is_bwc() - Check if dataset is valid BWC

wk.is_bwc(bwc)

True