Atmospheric Spectroscopy Lab UMBC

NOAA-20 CrIS Longwave BT Diffs vs Scene Temperature

Fri, Nov 30, 2018 L. Larrabee Strow

1 Location

  • Code location: /home/strow/Work/Cris/Calval/Empirical_fov_corr
  • Document location: /home/strow/Git/aslhugo/content/strowpages/cris_lw_fov_diffs_by_scene_bt

2 Approach

All NOAA-20 radiances were collected over a 16-day period (all scan angles) for a selection of six channels in the far longwave part of the CrIS longwave band. Some of these channels show significant mean B(T) differences among FOVs. The question being addressed here is: Are these differences relatively scene independent and can therefore be mimimized by a constant B(T) or radiance offset for certain FOVs. For these channels, FOV 5 is the outlier, so it is highlighted in the plots shown below.

The quantiles of these radiances for the cumulative probabilities of [0:0.01:1] were collected. The mean (over FOV) of the quantile values were converted to brightness temperature, giving the x-axis for the plots shown below. The y-axis is the difference between the mean quantiles (mean over FOVs) and the quantiles for each FOV. This difference is converted to B(T) differences using dB(T)/dr derivatives.

These plots therefore show the statistical mean B(T) difference between FOVs for this data set, which is extremely large. We also show the PDF for each channel.

3 Results

lw_651.875_cm_dbt_vs_bt.png
Figure 1: FOV B(T) differences versus scene temperature for 651.875 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.
lw_654.375_cm_dbt_vs_bt.png
Figure 2: FOV B(T) differences versus scene temperature for 654.375 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.
lw_659.375_cm_dbt_vs_bt.png
Figure 3: FOV B(T) differences versus scene temperature for 659.375 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.
lw_663.75_cm_dbt_vs_bt.png
Figure 4: FOV B(T) differences versus scene temperature for 663.75 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.
lw_666.25_cm_dbt_vs_bt.png
Figure 5: FOV B(T) differences versus scene temperature for 666.25 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.
lw_668.125_cm_dbt_vs_bt.png
Figure 6: FOV B(T) differences versus scene temperature for 668.175 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.
bt_ifov_minus_mean_bt_but5.png
Figure 7: Mean B(T) (all scenes, 16 day length) difference among FOV's.

3.1 Radiance Sapce

Here is channel 666.25 cm-1 FOV differences, but here plotted in radiance space.

lw_666.25_cm_dradiance_vs_bt.png
Figure 8: FOV radiance differences versus scene temperature for 666.25 cm-1 channel. The channel B(T) PDF is also shown, offset by -0.5.

4 Conclusions

FOV B(T) differences for these channels, when they exist, are largely scene independent and therefore a constant bias adjustment to these channels (esp. for FOV 5) will improve agreement and lower NWP analysis standard deviations for these channels.


comments powered by Disqus