Earthdata Forum

I am currently working on SWOT_L2_NALT_GDR_2.0 data. My goal is to retrieve SSH above geoid from 20Hz ku band products. To retrieve SSH above geoid (with instrumental corrections), I am following the following approach (actual variable names mentioned) -

SSH above geoid = "/data_20/altitude" - "/data_20/ku/range_ocean" - "/data_01/geoid"

However, the SSH values I am getting are orders of magnitude different than actual values. For example, the values under "mean_dynamic_topography" above geoid (which is the type of product I am looking for, I just need the actual SSH instead of the mean) are in range +-0.0xxxx range but my retrieved values are in the +-xx.xxxx range. I have the following concerns -

1. Am I missing any step somewhere in the process?
2. Do I have to account for the add_offset or the scale_factor present in the altitude and range_ocean variables metadata somewhere in the process? I have tried with adding add_offset and multiplied scale_factor with the variable values, however, the results are still orders of magnitude different from what it should be.

Thank you for your inquiry.

Please see the following reply from our SWOT data engineer:
--------------------------------------------------------------------------------
First, as you mentioned, I recommend that you check on the scale_factor. How are you reading in the data? (With Python xarray, or something else?) If you read the mean_dynamic_topography, and get a similar ranges of values as the attached image, then your current method is probably already handling the scale_factor. MDT should be in the tens of cm to <200 cm range.
At this link, there is an informative image for mean dynamic topography: https://www.esa.int/ESA_Multimedia/Images/2014/11/Mean_dynamic_topography

Second, I’m wondering how you handle the subtraction of the 1-Hz geoid data from 20-Hz altitude. Could some error result from that?

Third, it may be reasonable for your calculation of
SSH above geoid = "/data_20/altitude" - "/data_20/ku/range_ocean" - "/data_01/geoid"
to result in somewhat higher magnitudes because you are looking at 1 Hz, even 20 Hz, data. There would be a lot of small spatial/time scale and tidal variability in there. Mean Dynamic topography is much smoother, being the mean sea surface relative to the geoid. One can think of the portion as analogous to the "/data_20/altitude" - "/data_20/ku/range_ocean" [un-averaged mean sea surface]. But with the orders of magnitude difference that you are concerned about, I would first check the scale factor.
--------------------------------------------------------------------------------

Respectfully,
PO.DAAC User Services