This is the NIST 2008_01 LuMOS matrices created using the NIST laser Jun 2008. The data was collected on LuMOS for the Odd bouy before the M241 deployment. I then fixed the problem data and interpolated a matrix. There was colored source data collected for the fibered collectors but not LuMOS making it harder to check the SLC matrices and make sure they were working well. We dont have any colored sourced data for comparison so what I thought I would do is compare the SLCed LuMOS to the SLC fibered data. If they compare well then that would be a very good sign. Comparison with the fibered data will give me an idea if the spectral shape of the SLCed data is correct. The main are which needs checking is the UV end where there have been problems with values which were too high.
So the graphs below try to tell a story. Starting with the basic changes in the spectral shape correct to uncorrected data. Then the LuMOS data is compared to the fibered Lus. Then there are comparisons of Kls and Lw. And finally the changes in Lw relative to the uncorrect LuMOS data. The header above each figure will discuss what the figure is suppose to be showing.
| Wavelength (nm) | |||||
|---|---|---|---|---|---|
| 412.9 | 444.5 | 481.2 | 556.3 | 674.6 | 741.9 |
| 1.0632 | 1.0182 | 1.007 | 0.99278 | 0.99346 | 0.78926 |
| 1.0634 | 1.0181 | 1.0067 | 0.99258 | 0.9936 | 0.70988 |
| 1.0637 | 1.0182 | 1.0068 | 0.99195 | 0.99399 | 0.78184 |
| 1.0641 | 1.0183 | 1.0068 | 0.99383 | 0.99456 | 0.79071 |
| 1.0645 | 1.0182 | 1.0065 | 0.9938 | 0.99367 | 0.75796 |
| 1.0638 | 1.0182 | 1.0067 | 0.99282 | 0.99416 | 0.80283 |
| 1.0638 | 1.018 | 1.0069 | 0.99298 | 0.99453 | 0.7988 |
| 1.0642 | 1.018 | 1.0064 | 0.9919 | 0.99394 | 0.7272 |
| 1.0647 | 1.0181 | 1.0066 | 0.99253 | 0.99499 | 0.80338 |
| 1.0638 | 1.0188 | 1.0077 | 0.99431 | 0.99417 | 0.81919 |
| 1.064 | 1.0188 | 1.0078 | 0.99075 | 0.9943 | 0.74638 |
| 1.0634 | 1.0187 | 1.0078 | 0.99443 | 0.99454 | 0.80518 |
| 1.0634 | 1.019 | 1.008 | 0.99361 | 0.99418 | 0.80054 |
One MOBY file plotted so you can see the spectral changes relative to the uncorrected data for LuMOS data. It is zoomed in to the BSG since that is where the largest changes are that effect satellite data.
Figure 1
THis shows the fibered Lu data with the LuMOS data. Spectrally the shape looks pretty good. The UV section does not have an upturn and the overlap region looks good. And the follows the shape of the other Lus pretty well.
Figure 2
This shows a number of MOBY files (good or questionable) LuMOS data SLCed verses the uncorrected data. You can see most of the change is at the edges of the BSG. The biggest change in the RSG is past 700 nm.
Figure 3
This is a comparison of the Kls calculated using the fibered Lu (1,2 and 3) and the LuMOS (Lu4). The fibered and LuMOS Kls (Kl4) normally looked wrong when compared to Kls calculated from fibered data only. This got even worse when the fibered data was SLCed and LuMOS was not. The Kl4 below looks much better. There are large wiggles in the UV area, probably from slight variations in the wavelength changes between fibered and unfibered data. And the blue is a little high but it looks way better than it did before.
Figure 4
Next was an Lw comparison, problem is we dont calculate an Lw using LuMOS. We have a Kl using LuMOS but that Kl extrapolated LuTop to the surface. So I calculated an Lw with LuMOS and using Kl4 (Lu1-Lu4). And then plotted The three Lws (1,2 and 7) calculated using a fibered Lu and Kls from fibered data, one Lw (5) calculated using a fibered Lu (LuTop) and Kl from LuTop and LuMOS and a Lw using LuMOS and Kl4.
Figure 5
This is a zoomed in view of the UV area to see how the spectral shape matches up, looks pretty good. No upturns that I see.
Figure 6
This is a zoomed in view of the overlap to see how the spectral shape matches up, looks pretty good. The overlap looks really nice.
Figure 7
This is an ratio of the SLCed Lwx and the uncorrected Lwx for a number of good (or questionable) files. The table of satellite wieght values for VIIRS-NPP are at the top of this website.
Figure 8
