Work with AWS#

This example shows how to access radar data from the Colombian national radar network public on Amazon Web Services. We will look at the bucket structure and plot a PPI using the Xradar library. Radar reflectivity is filtered using some polarimetric values and xarray functionality.

Imports#

[1]:

from datetime import datetime

import boto3
import botocore
import cmweather  # noqa
import fsspec
import matplotlib.pyplot as plt
import xarray as xr
from botocore.client import Config
from pandas import to_datetime

import xradar as xd

Filtering data#

The blue background color indicates that the radar reflectivity is less than -10 dBZ. we can filter radar data using xarray.where module

[11]:

fig, (ax, ax1) = plt.subplots(1, 2, figsize=(10, 4))
ds.DBZH.where(ds.DBZH >= -10).plot(
    x="x", y="y", cmap="ChaseSpectral", vmin=-10, vmax=50, ax=ax
)
ds.RHOHV.where(ds.DBZH >= -10).plot(
    x="x", y="y", cmap="ChaseSpectral", vmin=0, vmax=1, ax=ax1
)
ax.set_title("")
ax1.set_title("")
fig.tight_layout()

../_images/notebooks_Read-plot-Sigmet-data-from-AWS_20_0.png

Polarimetric variables can also be used as indicators to remove different noises from different sources. For example, the \(\rho_{HV}\) measures the consistency of the shapes and sizes of targets within the radar beam. Thus, the greater the \(\rho_{HV}\), the more consistent the measurement. For this example we can use \(\rho_{HV} > 0.80\) as an acceptable threshold

[12]:

fig, (ax, ax1) = plt.subplots(1, 2, figsize=(10, 4))
ds.DBZH.where(ds.DBZH >= -10).where(ds.RHOHV >= 0.80).plot(
    x="x", y="y", cmap="ChaseSpectral", vmin=-10, vmax=50, ax=ax
)

ds.DBZH.where(ds.DBZH >= -10).where(ds.RHOHV >= 0.85).plot(
    x="x", y="y", cmap="ChaseSpectral", vmin=0, vmax=1, ax=ax1
)
ax.set_title("")
ax1.set_title("")
fig.tight_layout()

../_images/notebooks_Read-plot-Sigmet-data-from-AWS_22_0.png

Axis labels and titles#

We can change some axis labels as well as the colorbar label

[13]:

fig, (ax, ax1) = plt.subplots(1, 2, figsize=(10, 4))
ds.DBZH.where(ds.DBZH >= -10).where(ds.RHOHV >= 0.85).plot(
    x="x",
    y="y",
    cmap="ChaseSpectral",
    vmin=-10,
    vmax=50,
    ax=ax,
    cbar_kwargs={"label": r"$Reflectivity \ [dBZ]$"},
)

ds.RHOHV.where(ds.DBZH >= -10).where(ds.RHOHV >= 0.85).plot(
    x="x",
    y="y",
    cmap="ChaseSpectral",
    vmin=0,
    vmax=1,
    ax=ax1,
    cbar_kwargs={"label": r"$Corr. \ Coef. \  [unitless]$"},
)

# lambda fucntion for unit trasformation m->km
m2km = lambda x, _: f"{x/1000:g}"
# set new ticks
ax.xaxis.set_major_formatter(m2km)
ax.yaxis.set_major_formatter(m2km)
ax1.xaxis.set_major_formatter(m2km)
ax1.yaxis.set_major_formatter(m2km)
# removing the title in both plots
ax.set_title("")
ax1.set_title("")

# renaming the axis
ax.set_ylabel(r"$North - South \ distance \ [km]$")
ax.set_xlabel(r"$East - West \ distance \ [km]$")
ax1.set_ylabel(r"$North - South \ distance \ [km]$")
ax1.set_xlabel(r"$East - West \ distance \ [km]$")

# setting up the title
ax.set_title(
    r"$Guaviare \ radar$"
    + "\n"
    + f"${to_datetime(ds.time.values[0]): %Y-%m-%d - %X}$"
    + "$ UTC$"
)
ax1.set_title(
    r"$Guaviare \ radar$"
    + "\n"
    + f"${to_datetime(ds.time.values[0]): %Y-%m-%d - %X}$"
    + "$ UTC$"
)
fig.tight_layout()

../_images/notebooks_Read-plot-Sigmet-data-from-AWS_24_0.png

Work with AWS#

Imports#

IDEAM AWS Bucket#

Connecting to the AWS bucket#

Reflectivity and Correlation coefficient plot#

Filtering data#

Axis labels and titles#

This Page