Questions & Answers
No, our PSI workflow does not require integrating any 3rd party library
No, unfortunately, SAR is not the right technology for under water applications
Yes, our current workflow requires that all SAR scenes in a stack are of the same beam mode, band, sensor, orbital track, polarization
Yes, we have a few options. If you just wish to obtain the measurements, you can subscribe to CATALYST Insights. If you wish to run the process yourself, but via a blackbox, you can use these scripts to simplify the setup.
Please see our subscription prices here: https://catalyst.earth/products/catalyst-pro/professional-bundles/
This technology is available in our SAR bundle or Complete bundle
Please check out the tutorial at this link: https://catalyst.earth/2021/03/31/psi-workflow/
They can be accessed here: https://catalyst.earth/2021/03/31/psi-workflow/
Yes.
Gabriel ran a test on his desktop computer and found that it took about 10 min to unwrap 14 887 730 points, roughly 33% of the image
The user has control over the maximum spatial and/or temporal baselines. Interferometric combinations with a temporal or spatial baseline exceeding the user specified limits are not processed. However they can be added later if the user chooses. The temporal baseline limit is dependent upon the local vegetation and weather conditions. As a general rule of thumb, the maximum spatial baseline should not exceed 250 m.
The ability to generate and remove the atmospheric contributions has been developed and is currently undergoing testing. We use an iterative combination of high and low pass filters.
Yes. You can download the CATALYST Professional Trial now and contact our sales and support team for licensing
The results shown for Vancouver (and also demonstrated for Mexico City) agree with the trends shown in various publications. The selection of the reference point is critical since all subsidence measurements are relative to this point. For the demo, a single random reference point was selected. We plan to improve the adjustment for random offset for each interferogram within the stack by incorporating GNSS information.
Our goal is to make the workflow as flexible and intuitive as possible. Optimized default values are often available and in many instances we have applied novel approaches to the process. For example, we use a locally adaptive alpha value for the Goldstein filter. This ensures stronger spectral filtering in noisy phase areas and less filtering in areas of high coherence. Other packages require the user to indicate an alpha value (between 0 and 1) which is applied to the entire interferogram. This may result in underfiltering noisy areas and overfiltering highly coherent regions.
We apply an optimized least squares approach to unwrap the spatially connected network of PS points followed by a least squares optimization in the temporal direction.
We will be publishing some studies comparing our PSI results to ground truth shortly
We apply an optimized least squares approach to unwrap the spatially connected network of PS points followed by a least squares optimization in the temporal direction. Subsequent releases will use an iterative technique to remove atmospheric effects and to minimize the effects of outliers.
For now, amplitude dispersion is used to identity the PS points. Subsequent passes using atmospherically adjusted PS points is currently under consideration.
Our objective is to make the workflow as intuitive as possible for InSAR users with limited experience. However, we also strive to provide as much capability and flexibility as possible to expert users.
The ability to generate and remove the atmospheric contributions has been developed and is currently undergoing testing. We use an iterative combination of high and low pass filters.
For now, amplitude dispersion is used to identity the PS points. Subsequent passes using atmospherically adjusted PS points is currently under consideration.
