SCA outlier

[yangxiaolei-prog]

Hi,

In the SCA outlier detection of 010groopsinstruments, I used GRACE-FO data from December 2024 for gross error detection in the SCA data, but a large number of outliers were detected and removed. I tried changing the threshold from 1.5 to 2, but the residual result graph shows significant up-and-down fluctuations. Are there any improvement methods?

Sincerely!

[felingeroehlix]

Hi,
I suppose you use the GRACE example scenario from our cookbook? These settings for the instrument preprocessing stem from the original relative pointing mode that GRACE and GRACE-FO were using. Since July 2023 (and tests in January/February) the GRACE-FO satellites were switched to relaxed pointing mode, with larger line-of-sight offsets. You might have also noticed that there is no more LRI data available since that transition. So, this simple threshold outlier detection derived from a simulated star camera is not meaningful anymore. You could play around and see what outlier detection might work, or you just skip this preprocessing step.
Best,
Felix

[yangxiaolei-prog]

Hi,
I see, thank you very much!
Sincerely!