Does CRTM have limit on the satellite zenith angle?

Hi, I’m processing microwave data that may contains extremely high satellite zenith angle, around 85 degree or so. Is CRTM capable of processing such data? Thanks!

Jun, yes, there’s a couple of limits to be concerned with:

(1) Zenith angle maximum is 80 degrees, but the transmittance coefficient is only really good out to about 65 degrees.

(2) CRTM is a 1D model, so from a geometry perspective, it doesn’t make too much sense to go beyond 65 degrees anyway. because 3-D effects start to be potentially problematic in partially filled cloudy / precipitating scenes.

(3) The surface model over ocean, FASTEM-6, is limited to 60 degrees when surface scattering is turned on, so your emissivity bias will start to increase beyond 60 degrees. The error associated with that is a function of wind speed / surface roughness.

I generally recommend that people go no higher than 60 degrees with CRTM for those reasons. You may be able to get acceptable errors out to 70 degrees, but it starts to get especially tricky with the curvature of the Earth, Bragg scattering effects from coherent wave structures, etc.

If you have a need for accurate RT at angles larger than 60 degrees, I recommend a limb-specific RT model, that takes into account the Earth’s curvature, 3-D geometry, etc.

Thanks a lot! That’s very helpful!


Hi, Benjamin,
I didn’t see the limit of zenith angle mentioned in the CRTM user’s guide. Do you have an article I can cite?


Yes, you are right – that’s something that should be added to the User Guide.

The information I obtained from the source code itself.

See also:

RTTOV claims to support clear sky simulations up to 84 degrees, but they also note that the error is substantial:

For a more updated set of information and future capabilities, I recommend contacting Ming Chen (, he’s the primary point of contact on surface emissivity modeling right now.