xradar is in an early stage of development, please report any issues here!

CfRadial1#

Imports#

[1]:

import os
import tempfile
import cmweather
import numpy as np
import xarray as xr
import xradar as xd
import datatree as xt
from open_radar_data import DATASETS

Plot Azimuth vs. Range#

[4]:

radar.sweep_0.DBZ.plot(cmap="ChaseSpectral", vmin=-10, vmax=70)

[4]:

<matplotlib.collections.QuadMesh at 0x7f0b7474a1d0>

../_images/notebooks_CfRadial1_Export_7_1.png

Plot Time vs. Range#

[5]:

radar.sweep_0.DBZ.swap_dims({"azimuth": "time"}).sortby("time").plot(
    cmap="ChaseSpectral", vmin=-10, vmax=70
)

[5]:

<matplotlib.collections.QuadMesh at 0x7f0b745f93d0>

../_images/notebooks_CfRadial1_Export_9_1.png

Plot PPI#

[7]:

radar["sweep_0"]["DBZ"].plot(x="x", y="y", cmap="ChaseSpectral", vmin=-10, vmax=70)

[7]:

<matplotlib.collections.QuadMesh at 0x7f0b747693d0>

../_images/notebooks_CfRadial1_Export_13_1.png

Filter#

Apply basic reflectivity filter. This is just a demonstration.

[8]:

def ref_filter(dtree, sweep="sweep_0", field="DBZ"):
    ds = dtree[sweep].where((dtree[sweep][field] >= -10) & (dtree[sweep][field] <= 70))
    red_patch = ds.where(
        (
            (ds[field] >= ds[field].max().values - 0.5)
            & (ds[field] <= ds[field].max().values + 0.5)
        ),
        drop=True,
    )
    rmin, rmax = int(red_patch.range.min().values - 150), int(
        red_patch.range.max().values + 150
    )
    out_of_range_mask = (ds.range < rmin) | (ds.range > rmax)
    ds[field] = ds[field].where(out_of_range_mask)
    # Interpolate missing values using the slinear method along the 'range' dimension
    ds[field] = ds[field].interpolate_na(dim="range", method="slinear")
    dtree[sweep][f"corr_{field}"] = ds[field].copy()
    return dtree[sweep]

[9]:

swp0 = ref_filter(radar, sweep="sweep_0", field="DBZ")

[10]:

swp0.corr_DBZ.plot(x="x", y="y", cmap="ChaseSpectral", vmin=-10, vmax=70)

[10]:

<matplotlib.collections.QuadMesh at 0x7f0b745a93d0>

../_images/notebooks_CfRadial1_Export_17_1.png

Filter full volume#

[11]:

# Initialize an empty DataTree
result_tree = xt.DataTree()

for sweep in radar.sweep_group_name.values:
    corrected_data = ref_filter(radar, sweep, field="DBZ")

    # Convert the xarray Dataset to a DataTree and add it to the result_tree
    data_tree = xt.DataTree.from_dict(corrected_data.to_dict())

    # Copy the contents of data_tree into result_tree
    for key, value in data_tree.items():
        result_tree[key] = value

[12]:

radar.sweep_6.corr_DBZ.plot(x="x", y="y", cmap="ChaseSpectral", vmin=-10, vmax=70)

[12]:

<matplotlib.collections.QuadMesh at 0x7f0b744c1e90>

../_images/notebooks_CfRadial1_Export_20_1.png

Export#

Export to CfRadial1

[13]:

xd.io.to_cfradial1(dtree=radar, filename="cfradial1_qced.nc", calibs=True)

[14]:

?xd.io.to_cfradial1

Note#

If filename is None in the xd.io.to_cfradial1 function, it will automatically generate a filename using the instrument name and the first available timestamp from the data.