Down-welling sky radiance in TIR , fixed_angle

[Marius Dahuron]

Hi,

I am using rttov to compute atmospheric terms for thermal ir radiative transfer. In order to retrieve the downwelling radiance from the atmosphere i am using lambertian option, with fixed_angle option.

I can’t figure out why the radiance retrieved from 55° fixed angle (according to the description of the option) gives different results from the 55° specular one, knowing that there is a power p that may be involved for a different viewing path. But here taking similar geometry i was not expecting this …

Maybe you could help me to understand this, I guess it may be a mistake of mine in my use of rttov ?

All the best,

Marius

[James Hocking]

Hi Marius,

This is due to the fact that the Lambertian option calculates the downwelling radiances using layer-average Planck radiances instead of applying the linear-in-tau approach to calculating the layer Planck radiance, the latter being the default for non-Lambertian simulations.

The linear-in-tau option is intended primarily for situations with cloud (or otherwise optically thick layers) where the Planck radiance should be from the cloud top (the temperature at the upper level bounding a layer) rather than the average of the upper and lower bounding level temperatures/radiances (here I am talking about upwelling radiation, but similar reasoning applies to downwelling radiances).

The linear-in-tau approximation only has a significant impact when layers are optically thick, and in these situations the downwelling radiance does not reach the top-of-atmosphere due to absorption. Therefore the linear-in-tau option is not really relevant for downwelling radiances and has no significant impact on TOA radiances. This is why you will observe that the TOA radiances/BTs are extremely similar for your Lambertian/non-Lambertian cases even though the downwelling radiances differ.

If you switch to layer-average radiances for the downwelling radiance calculations (opts % rt_all % rad_down_lin_tau = .false.) then RTTOV gives identical downwelling radiances for Lambertian/non-Lambertian simulations with a zenith angle of 55 degrees.

Best wishes,
James

[Author]

Posts