0

I want a global file chopped into multiple netcdf files corresponding to the modis tiles.

I currently have a directory of GLOBAL netcdf files over time.

NOTE: ERA interim data is on the regular lat/lon grid.

Longitudes range from 0 to 360, which is equivalent to -180 to +180 in Geographic coordinate systems. LINK

This is how my directory currently looks:

ERAIN_SFC00_6H_10mWind.nc  ERAIN_SFC00_6H_2mtemperature.nc  ERAIN_SFC00_6H_TCW-O3.nc

What I want is a set of files on the sinusoidal projection, one file for each MODIS tile. Like this:

ERAIN_SFC00_6H_TCW-O3_h07v06.nc  ERAIN_SFC00_6H_TCW-O3_h24v03.nc         ERAIN_SFC00_6H_10mWind_H15V05.nc  ERAIN_SFC00_6H_10mWind_H31V06.nc
ERAIN_SFC00_6H_TCW-O3_h08v03.nc  ERAIN_SFC00_6H_TCW-O3_h24v04.nc         ERAIN_SFC00_6H_10mWind_H15V07.nc  ERAIN_SFC00_6H_10mWind_H31V07.nc
ERAIN_SFC00_6H_TCW-O3_h08v04.nc  ERAIN_SFC00_6H_TCW-O3_h24v05.nc         ERAIN_SFC00_6H_10mWind_H16V05.nc  ERAIN_SFC00_6H_10mWind_H31V08.nc
ERAIN_SFC00_6H_TCW-O3_h08v05.nc  ERAIN_SFC00_6H_TCW-O3_h24v06.nc         ERAIN_SFC00_6H_10mWind_H16V06.nc  ERAIN_SFC00_6H_10mWind_H31V09.nc
ERAIN_SFC00_6H_TCW-O3_h08v06.nc  ERAIN_SFC00_6H_TCW-O3_h24v07.nc         ERAIN_SFC00_6H_10mWind_H16V07.nc  ERAIN_SFC00_6H_10mWind_H31V10.nc
ERAIN_SFC00_6H_TCW-O3_h08v07.nc  ERAIN_SFC00_6H_TCW-O3_h25v02.nc         ERAIN_SFC00_6H_10mWind_H16V08.nc  ERAIN_SFC00_6H_10mWind_H31V11.nc
ERAIN_SFC00_6H_TCW-O3_h09v02.nc  ERAIN_SFC00_6H_TCW-O3_h25v03.nc         ERAIN_SFC00_6H_10mWind_H17V02.nc  ERAIN_SFC00_6H_10mWind_H31V12.nc

Pseudo code

1. Convert the files to sinusoidal projection
2. determine the extent of the MODIS tiles
3. chop the MODIS tile extents and save as new netCDF

I have absolutely no idea of a) how to convert to sinusoidal projection b) how to chop the modis tiles.

I made these tile extents from already existing modis netcdf files. I don't know if they're accurate (is there an official source defining MODIS tiles?) but they have values.

The data

EXAMPLE DATA HERE

import xarray as xr

ds = xr.open_dataset('ERA_test.nc')
ds

Out[]:
<xarray.Dataset>
Dimensions:        (lat: 256, lon: 512, time: 5)
Coordinates:
  * lon            (lon) float64 0.0 0.7031 1.406 ... -2.109 -1.406 -0.7031
  * lat            (lat) float64 89.46 88.77 88.07 ... -88.07 -88.77 -89.46
  * time           (time) datetime64[ns] 2008-01-02 ... 2008-01-03
Data variables:
    temperature2m  (time, lat, lon) float32 ...
Attributes:
    content:  ERA INTERIM data of var167; converted from grb to netCDF with C...
    history:

 The tile extents

import pickle

tiles = pickle.load(open('tile_extents.pkl','rb'))
tiles['h16v06']

Out[]:
{'lon_min': -23.09401076758503,
 'lon_max': -10.641777724759121,
 'lat_min': 20.0,
 'lat_max': 30.0}

Converting to sinusoidal projection

This is the one that I really don't know how to do. I think I have found the MODIS projection from pyproj.

import pyproj as Proj

p_modis_grid = Proj('+proj=sinu +R=6371007.181 +nadgrids=@null +wktext')

But how do I go about using this to reproject the netCDF data?

Happy to know if I'm using the wrong tool.

Will rasterio be better suited?

Or NCO/CDO?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.