I have been stuck attempting to convert Sentinel-3 LST netCDF data to geotiffs for a while now as part of my master's thesis.

Firstly, an introduction to the data. It is packaged as follow: image

The data is packaged as several .nc files, as well as a xml file containing metadata. The metadata suggests that the data should be in EPSG:4346:

     <metadataObject ID="measurementFrameSet" classification="DESCRIPTION" category="DMD">
        <metadataWrap mimeType="text/xml" vocabularyName="Sentinel-SAFE" textInfo="Frame Set">
                 <sentinel-safe:footPrint srsName="http://www.opengis.net/def/crs/EPSG/0/4326">
                    <gml:posList>-29.1992 8.81452 -29.1684 9.00931 -29.0806 9.52497 -28.9909 10.035 -28.896 10.5474 -28.801 11.0545 -28.7054 11.5734 -28.605 12.0778 -28.5082 12.5866 -28.4088 13.0983 -28.2994 13.6035 -28.197 14.1087 -28.0927 14.6147 -27.9761 15.1199 -27.8675 15.6236 -27.7564 16.1214 -27.6423 16.627 -27.525 17.1239 -27.4036 17.6145 -27.2806 18.1169 -27.1611 18.6177 -27.0344 19.1076 -26.9124 19.605 -26.781 20.0911 -26.6548 20.5894 -26.5286 21.0864 -26.3885 21.5669 -26.2565 22.0561 -26.1192 22.5398 -25.9879 23.0344 -25.8475 23.5168 -28.389 24.4405 -30.9637 25.4388 -33.5273 26.5043 -36.0695 27.6424 -36.078 27.6463 -36.2329 27.1108 -36.4066 26.5843 -36.5511 26.0457 -36.7063 25.5069 -36.8573 24.9628 -37.0105 24.4346 -37.1568 23.8819 -37.2966 23.3343 -37.4393 22.7945 -37.5805 22.24 -37.7181 21.6969 -37.8487 21.1329 -37.9835 20.584 -38.1138 20.0259 -38.2315 19.464 -38.3634 18.9072 -38.4821 18.3398 -38.5973 17.7783 -38.7109 17.2074 -38.8288 16.6385 -38.9366 16.0638 -39.0387 15.4947 -39.148 14.9205 -39.2443 14.3444 -39.3456 13.7696 -39.4421 13.1931 -39.5308 12.6058 -39.6274 12.0296 -39.7142 11.4424 -39.7416 11.2224 -39.7345 11.2088 -37.1002 10.5838 -34.4547 9.9742 -31.8069 9.37884 -29.1992 8.81452</gml:posList>

Now I attempted to load the data into xarray using this:

import rioxarray
import xarray
from pyproj import CRS
root = r'G:\MSc\Code\JupyterLab\temp\S3B_SL_2_LST____20201214T214826_20201214T215126_20201216T025309_0179_046_385_5400_LN2_O_NT_004.SEN3\*.nc'
xds = xarray.open_mfdataset(root,combine = 'by_coords', decode_times=False)

This did not work, and output the following error:

ValueError: arguments without labels along dimension 'columns' cannot be aligned because they have different dimension sizes: {130, 1500}

Now I believe the problem with this is that the _in files have a 1200 x 1500 grid, whereas the _tx files have a 500 x 1500 grid. Working around this problem, we can open up only the _in files (or the _tx files, but I am using _in for the example): image

If I am correct in interpreting the information, it appears that the netCDF does not have any spatial dimensions defined, despite that I have attempted to set the CRS using xds.rio.set_crs(4326) and exporting that to a geotiff using xds.to_netcdf(r'G:\MSc\TestImage\testout\testout.tiff'), but that did not work and seemed to only map the row and column values as coordinates. This did not work.

Now the Sentinel-3 LST product does have some form of spatial referencing in the data that I assume I can work with - the longitude_in/_tx and latitude_in/_tx bands. These are gridded values containing the geographic longitude/latitude coordinates for every cell in the netCDF. I have used this to extract point values from the netCDF grid in the past and have attempted to solve this problem by turning every cell into a point feature, but this proved far too slow to be of much use.

Anyway, I have a feeling that this information could be used somehow to georeference the netCDFs, either by creating a new set of spatial dimensions, or failing that they should be able to act as GCPs, but I fear that this second option may also be too slow. I have also tried using SNAP, but unfortunately it proved too slow and unstable to work for this problem. Speed is a bit of an issue, as I am aiming to process a year's worth of information over an AOI.

  • Are you able to upload example files?
    – snowman2
    May 4, 2021 at 0:20
  • I just spent a few hours with one of these datasets. I tried GDAL and ESA's SNAP software. The problem is that geodetic.nc has the lat long data, and the other files have the data. I guess they need to be merged. The best idea may be to combine them with gdal libraries in Python, etc. Perhaps it's possible with some NetCDF library.
    – wingnut
    May 4, 2021 at 5:54
  • I got it working with a small script to join the lat/long data to the rasters, create point files and regrid.
    – wingnut
    May 4, 2021 at 12:21

1 Answer 1


Short answer:

  1. use GDAL tools to extract lat and long data from an .nc file.
  2. Merge it with the data from any band you want to see to make an x,y,z dataset.
  3. Convert the irregular x,y,z data to a shapefile and apply a CRS.
  4. Grid the shapefile at desired resolution.

The lat/long data are not regularly sampled, so the data can't be loaded as an XYZ raster directly, and need to be gridded.

This flow works on the one dataset that I tried.

   gdal_translate -of XYZ NETCDF:"S4_radiance_an.nc":S4_radiance_an S4_radiance.csv
   gdal_translate -of XYZ NETCDF:"geodetic_an.nc":latitude_an latitude.csv
   gdal_translate -of XYZ NETCDF:"geodetic_an.nc":longitude_an longitude.csv

   echo X,Y,Z >merged.csv
   paste longitude.csv latitude.csv S4_radiance.csv|gawk 'BEGIN{OFS=","}{print($3/1000000,$6/1000000,$9)}' >>merged.csv
   ogr2ogr -a_srs "EPSG:4326"  -mapFieldType All=Integer -oo X_POSSIBLE_NAMES=X* -oo Y_POSSIBLE_NAMES=Y* -oo KEEP_GEOM_COLUMNS=NO newS4_radiance.shp merged.csv

enter image description here

Long Answer

Use GDAL tools from the command line. For example, file S4_radiance_an.nc

gdalinfo S4_radiance_an.nc

At the bottom of the output, there are lines like:

  SUBDATASET_1_DESC=[2400x3000] S4_exception_an (8-bit unsigned integer)
  SUBDATASET_2_DESC=[2400x374] toa_radiance_status_flag (8-bit unsigned integer)
  SUBDATASET_3_DESC=[2400x3000] toa_radiance (16-bit integer)
  SUBDATASET_4_DESC=[2400x374] toa_radiance (16-bit integer)

So resolution is 2400 x 3000 for bands that contain 'an' (nadir c.f. oblique).

The same information from geodetic_an.nc yields:

  SUBDATASET_1_DESC=[2400x3000] surface_altitude (16-bit integer)
  SUBDATASET_2_DESC=[2400x374] surface_altitude (16-bit integer)
  SUBDATASET_3_DESC=[2400x3000] latitude (32-bit integer)
  SUBDATASET_4_DESC=[2400x374] latitude (32-bit integer)
  SUBDATASET_5_DESC=[2400x3000] longitude (32-bit integer)
  SUBDATASET_6_DESC=[2400x374] longitude (32-bit integer)

Resolution is again 2400 x 3000, so we assume they are related.

The NAME and DESCRIPTION of set 3 are:

  SUBDATASET_3_DESC=[2400x3000] latitude (32-bit integer)

We can use the NAME to load using GDAL tools. First, translate to XYZ format with:

gdal_translate -of XYZ NETCDF:"geodetic_an.nc":latitude_an latitude.csv

Similarly for the longitude file. In each case, records are like:

2990.5 2399.5 7575671
2991.5 2399.5 7574636
2992.5 2399.5 7573601

So the format is I, J, Latitude, and latitude is in (degrees * 1000000).

We can merge the XYZ files representing Latitude, Longitude and, say, Radiance (from S4_radiance_an.nc, for example, which is also 2400 x 3000 in the band "S4_radiance_an.nc":S4_radiance_an"). The first and last records of these three files are:

0.5 0.5 -32768
1.5 0.5 -32768
2.5 0.5 -32768
2997.5 2399.5 -32768
2998.5 2399.5 -32768
2999.5 2399.5 -32768
0.5 0.5 -73571
1.5 0.5 -74567
2.5 0.5 -75563
2997.5 2399.5 7568425
2998.5 2399.5 7567390
2999.5 2399.5 7566355
0.5 0.5 95229130
1.5 0.5 95233511
2.5 0.5 95237892
2997.5 2399.5 110768320
2998.5 2399.5 110772730
2999.5 2399.5 110777141

So they have the same limits and are therefore related. Merge them using any method you like. They have the same number of rows (wc <S4_radiance.csv =>> 7200000 21600000 127768857), so on Unix paste and AWK will work:

paste longitude.csv latitude.csv S4_radiance.csv|gawk '{print($3/1000000,$6/1000000,$9)}' >merged.csv

This gives a file like:


The Z values are -32768, which means nodata, and is normal at the edge of a raster. X and Y values are in degrees These can be loaded as point data and gridded as required.

  • Thank you so much for the help! I do however have some follow up questions. Firstly, how can I refer to different bands within a subdataset? Secondly, is there a way to have gdal make use of the defined nodata, scale factor and offset data defined for each subdataset. Thirdly, Is there a way to pass gdal translate directly to pandas for further analysis?
    – TheOGMalan
    May 19, 2021 at 8:53
  • See the NAME above. That is the band. The nodata value is used by GDAL by default. The scale factor and offset should be applied automatically by GDAL tools. gdal_traslate creates a new dataset - I generally create TIFF files. The input is a bad from the .nc file, so the outputTIFF will have a singleband only. Load it with Geopandas. That will give you a Dataframe with a geometry column. The main thing is to experiment with all this. It becomes clearer when you do. There is also a GDAL Python library that does similar stuff, but frankly, command line is easier and scriptable.
    – wingnut
    May 20, 2021 at 4:02
  • thanks again, but I still however have a follow up, take the following example variable in the netCDF - 'T015812_0179_067_014_5400_LN2_O_NT_004.SEN3\met_tx.nc":sensible_heat_tx SUBDATASET_8_DESC=[5x1200x130] integral_of_surface_downward_sensible_heat_flux_wrt_time (32-bit floating-point)'. This subdataset has an additional time dimension that I need to be able to specify in the gdal translate call, but I am unsure how to do so.
    – TheOGMalan
    Jun 1, 2021 at 7:38
  • As I understand it, GDAL will translate these to individual bands in an output GeoTiff. If you want to manipulate the data in the .nc file, then the Python neetCDF4 library is helpful. Did you try the ESA tools (SNAP, Sentinel toolboxes, etc)? They certainly perform georeferencing for Sentinel 1 (I used it last week). They are designed to work with Sentinel data. The RUS Webinars on YouTube will get you started.
    – wingnut
    Jun 2, 2021 at 14:02
  • @wingnut I have nearly the same problem as OP, with the slight difference, that I'm dealing with Sentinel-3-Syn-2-datasets (a lot of them, thousands). I'm trying to implement a processing chain which georeferences all spectral bands of the S3 datasets. I followed your steps but I'm not sure what happens e.g. in paste longitude.csv latitude.csv S4_radiance.csv|gawk 'BEGIN{OFS=","}{print($3/1000000,$6/1000000,$9)}' >>merged.csv. Furthermore if I want to save as netcdf with gdal_translate I get: ERROR 1: Ungridded dataset: At line 4, too many stepY values. Do you know how to solve this?
    – Corbjn
    Nov 2, 2021 at 10:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.