Tropospheric correction for PSInSAR time series

Use this Forum to find information on, or ask a question about, NASA Earth Science data.
Post Reply
shadi7577
Posts: 3
Joined: Tue Dec 06, 2022 8:25 am America/New_York
Answers: 0

Tropospheric correction for PSInSAR time series

by shadi7577 » Tue Dec 06, 2022 9:02 am America/New_York

I am using Sentinel-1 A/B images to create a deformation time series using Persistent scatterer interferometry concept through Stanford Method for Persistent scatterers (StaMPS), and I have tried the linear atmospheric phase screen removal through matlab/StaMPS processing using Toolbox for Reducing Atmospheric InSAR Noise (TRAIN) through which a script (“oscar-client-python/get_modis.py”) is called, which I have cloned from https://github.com/jgomezdans/get_modis/.
however I was informed that this script is obsolete and is no longer a viable option to remove the tropospheric phase contribution to the interferometric deformational phase, Applying this script did remove some of the phase contribution but the noise still persists for each ps points and thus they shows random displacement pattern over time. Which recipe can I use in order to provide a dependable atmospheric correction for PSInSAR time series through TRAIN knowing that my study area is a natural landscape with minimal urban presence with variable relief beside the gulf of suez.
Attachments
velocity map with deformational phase only (after linear_APS removal)
velocity map with deformational phase only (after linear_APS removal)
velocity maps with deformational phase only.png (409.11 KiB) Not viewed yet
velocity maps with deformational phase and additional ramp, orbital error, DEM, atmospheric phase contributions
velocity maps with deformational phase and additional ramp, orbital error, DEM, atmospheric phase contributions
velocity map with all non-deformational phase contributions.png (318.15 KiB) Not viewed yet
the obsolete script get_modis.py
the obsolete script get_modis.py
aps_config.png (155.03 KiB) Not viewed yet

Tags:

dbekaert
Posts: 1
Joined: Tue Dec 06, 2022 2:28 pm America/New_York
Answers: 0

Re: Tropospheric correction for PSInSAR time series

by dbekaert » Tue Dec 06, 2022 2:30 pm America/New_York

The MODIS correction provided through OSCAR is no longer operational (server is offline) hence does not work with TRAIN anymore. Best option would be to change to a weather model-based correction. However TRAIN is currently not actively maintained and not sure if models such as ERA5 or MERRA2 would still be working.

shadi7577
Posts: 3
Joined: Tue Dec 06, 2022 8:25 am America/New_York
Answers: 0

Re: Tropospheric correction for PSInSAR time series

by shadi7577 » Wed Dec 07, 2022 4:34 am America/New_York

I appreciate letting me realize this issue with TRAIN. Would it be possible if you could direct me into a method of using weather model for corrections during StaMPS time series processing? I couldn't find guiding tutorials so far for such purpose except for applying TRAIN toolbox in combination with Matlab
Last edited by shadi7577 on Wed Dec 07, 2022 5:53 am America/New_York, edited 2 times in total.

shadi7577
Posts: 3
Joined: Tue Dec 06, 2022 8:25 am America/New_York
Answers: 0

Re: Tropospheric correction for PSInSAR time series

by shadi7577 » Sat Jan 21, 2023 6:15 am America/New_York

dbekaert wrote:
> The MODIS correction provided through OSCAR is no longer operational
> (server is offline) hence does not work with TRAIN anymore. Best option
> would be to change to a weather model-based correction. However TRAIN is
> currently not actively maintained and not sure if models such as ERA5 or
> MERRA2 would still be working.
I would like to use TRAIN For applying GACOS Model. Please I look forward to your expertise on how to adjust the directories to the suitable scripts for the APS_CONFIG.sh since the get_modis.py script won't work.

Post Reply