IMD Radar Data

The India Meteorological Department (IMD) provides radar data in a format similar to IRIS/Sigmet, but with some differences in structure and formatting. radarx offers specialized support for processing and analyzing this IMD radar data, ensuring compatibility and ease of use despite the variations from standard IRIS/Sigmet formats.

import os
import subprocess
import radarx as rx

Load IMD Radar Data

# Base URL for the radar data files on GitHub
base_url = (
    "https://raw.githubusercontent.com/syedhamidali/pyscancf_examples/main/data/goa16/"
)

# List of radar data files to download
files = [
    "GOA210516024101-IMD-B.nc",
    "GOA210516024101-IMD-B.nc.1",
    "GOA210516024101-IMD-B.nc.2",
    "GOA210516024101-IMD-B.nc.3",
    "GOA210516024101-IMD-B.nc.4",
    "GOA210516024101-IMD-B.nc.5",
    "GOA210516024101-IMD-B.nc.6",
    "GOA210516024101-IMD-B.nc.7",
    "GOA210516024101-IMD-B.nc.8",
    "GOA210516024101-IMD-B.nc.9",
]

# Target directory to save downloaded files
target_dir = os.path.join(os.getenv("GITHUB_WORKSPACE", "."), "radarx_data")
os.makedirs(target_dir, exist_ok=True)


# Function to download files using curl
def download_with_curl(file_name):
    file_url = base_url + file_name
    file_path = os.path.join(target_dir, file_name)

    if not os.path.exists(file_path):
        print(f"Downloading {file_name}...")
        subprocess.run(["curl", "-o", file_path, file_url], check=True)
        print(f"Downloaded {file_name}")
    else:
        print(f"{file_name} already exists.")

    return file_path


# Download the files
downloaded_files = []
for file_name in files:
    file_path = download_with_curl(file_name)
    downloaded_files.append(file_path)

print(f"Number of files: {len(downloaded_files)}")

Read a Sweep

# Load the first radar file using radarx
swp = rx.io.read_sweep(downloaded_files[0])
swp

We have successfully read one sweep in the above cell. Now, let’s create a volume scan.

Create a Volume

We have a total of 10 files, each representing a different sweep. Our goal is to combine these individual sweeps into a single CF-Radial volume scan dataset.

for file in downloaded_files:
    print(os.path.basename(file))

vol = rx.io.read_volume(downloaded_files)

vol

We have successfully generated a volume, which can now be explored further. The volume is stored in a DataTree, a high-level structure that organizes and manages the data within the xarray framework.

vol.groups

ds = vol["volume_0"].ds

ds

We have successfully created a volume, and we can export it to CF-Radial formatted NetCDF using xarray.

# We can save all these cfradial1 files by iterating over the vol object

outdir = "IMD_OUT"
os.makedirs(outdir, exist_ok=True)
for key in vol.children:
    ds = vol[key].ds
    time_str = ds.time.min().dt.strftime("%Y%m%d_%H%M%S").item()
    ds.to_netcdf(os.path.join(outdir, f"GOA_{time_str}_cfrad1.nc"))

os.listdir("IMD_OUT/")

Convert to CF/Radial2

We can also convert it to cfradial2 data using to_cfradial2 function

dtree = rx.io.to_cfradial2_volumes(vol)

dtree.groups

ds = dtree["volume_0"].ds

# Export all the data in a dtree

# for volume in dtree:
#     dtree[volume].to_netcdf()