BRDF atlas geographic coordinates

[Andrew Prata]

Hi James

I’m trying to plot the RTTOV BRDF atlas data from this file: CMSsolarbrdf_V1.0_0.1deg_200701_mask.nc

However, when I read the data (with python xarray), I don’t see any coordinates for the arrays:

xarray.Dataset
Dimensions: nwave: 7, Lon: 3600, Lat: 1800, mask: 2250931
Coordinates: (0)
Data variables:
wavelength (nwave) float32 dask.array<chunksize=(7,), meta=np.ndarray>
brdf_flag (Lon, Lat) int8 dask.array<chunksize=(2896, 1448), meta=np.ndarray>
fiso1 (mask) float32 dask.array<chunksize=(2097152,), meta=np.ndarray>
etc
…

I understand how to unpack and reshape the kernels; however, I am currently making an assumption about the latitude, longitude arrays based on the Vidot et al. paper stating that the grid resolution is 0.1 degrees.

Currently, I assume the geographic coordinates are as follows (where np is numpy):

lons_1d = np.arange(-180, 180, 0.1)
lats_1d = np.arange(-90, 90, 0.1)

Can you confirm that this is the correct way to re-construct the lat/lon grids?

Thanks
Andrew

[James Hocking]

Hi Andrew,

The longitudes go from -180 to +180 (as you already have), but the latitudes go from +90 to -90 (so opposite to what you have).

You can check this by plotting the ‘brdf_flag’ dataset:

import matplotlib.pyplot as plt
import netCDF4 as nc

rootgrp = nc.Dataset(‘CMSsolarbrdf_V1.0_0.1deg_200701_mask.nc’, ‘r’, format=’NETCDF4′)
data = rootgrp.variables[‘brdf_flag’][:]
plt.imshow(data.transpose())
plt.show()

Best wishes,
James

[Andrew Prata]

Ah yes I have already realised the issue with the latitudes.

So, to be very clear: longitude coords will go from -180, -179.9, …, 179.9 ? i.e. not include +180 at the end?
Same question for latitudes: it will be 90, 89.9, …, -89.9 ? (not inclusive of -90)?

Thanks
Andrew

[James Hocking]

Hi Andrew,

The atlas grid starts at 89.95 degrees latitude and -179.95 degrees longitude, so that the grid is symmetrical in lat and lon.

latitude: +89.95 -> -89.95 in steps of 0.1 degrees
longitude: -179.95 -> +179.95 in steps of 0.1 degrees

Sorry I missed this detail off my previous reply!

Best wishes,
James

[Author]

Posts