Define or convert projection when netcdf saved as tif raster?

Question

A netcdf data structure showed as follows:

Dimensions:  (lat: 360, lon: 720)
Coordinates:
  * lon      (lon) float32 -179.75 -179.25 -178.75 -178.25 -177.75 -177.25 ...
  * lat      (lat) float32 -89.75 -89.25 -88.75 -88.25 -87.75 -87.25 -86.75 ...
    time     datetime64[ns] 1901-01-16
 Data variables:
    cld      (lat, lon) float64 nan nan nan nan nan nan nan nan nan nan nan ...
    stn      (lat, lon) int32 -999 -999 -999 -999 -999 -999 -999 -999 -999 ...

I want to save this layer as .tif raster with wgs84 proj. How to convert the projection?

Answer 1

You have a least a few ways to do it:

1. GDAL command line utilities:

Since version 2.1 GDAL working fine (read and write) with raster and vector netCDF files. You can use both GDAL command line utilities (gdal_translate and gdalwarp) to reproject and output to GeoTiff format file.

gdal_translate -of GTiff -ot Float32 -a_srs "+proj=longlat +ellps=WGS84" input.nc output.tif

2. Execute GDAL utility from Python scripts:

Here is some example:

import subprocess

def convert_netcdf_to_geotiff(infile, outfile, band_number=1, verbose=False):

    with open(os.devnull, 'w') as devnull:
        return subprocess.check_call([
            "gdal_translate", infile, outfile,
            "-of", "GTiff",
            "-ot", "Float32",
            "-a_srs", "+proj=longlat +ellps=WGS84",
            "-b", str(band_number)

        ],
            stdout=None if verbose else devnull
        )

3. Also, you can use GDAL utility in Docker container

If you don't want or can't install/build GDAL on your machine you can use prepared Docker images with GDAL 2.1 version or other versions with support netCDF driver.

There is one of the solutions:

cd /path/to/dir/with/netCDF/file
docker run -v $(pwd):/data geodata/gdal gdal_translate -of GTiff -ot Float32 -a_srs "+proj=longlat +ellps=WGS84" input.nc output.tif