Problem with Collocation of ENVISAT ASAR-Data

Dear all,

I want to do a Collocation of two ENVISAT ASA WSM scenes and then continue with Band Math. Once i click Run in the Collocation window I get an error message. I don’t know what I’m doing wrong. I have already deleted all masks and checked the pixels in Properties.

Does anyone have an idea how to solve the issue?

Thanks!
Christina

“Subset_subset_0_of_delta_total_ASA…” indicates you have already applied a couple of operations prior to this. Please list what you did and in which order. Did you switch the output format by chance?

I did the pre-processing of the two scenes in the order as followed:
Apply Orbit File
Thermal Noise Removal
Calibration
Speckle Filtering (Median 5x5)
Terrain Correction (SRTM 3-arc second; ~75 m resolution)|
Lineartofromdb

After that, I created the subsets and changed the brightness in the histograms. No, I didn’t touch the output format.

Thank you, looks alright.
Maybe you can then use the Create Stack operator (in Radar > Corregistration > Stack Tools)
Use “Product Geolocation” instead of “Orbit” because your data is already projected.

This should allow you to bring both images into one product.

1 Like

The Stacking worked. Thanks!
But now the subsets do not match the scenes anymore. Do you also have an idea for that problem?

Maybe it is better to import the vector after the stacking.
Can you please create an RGB image of both bands to see if they are aligned?


This is the RGB. I think only one of the scenes is wrong because the subset seems to fit at least with the green one. The two scenes are only 4 months apart, pink is the younger one.
Could it be anything with the coordinate systems? But I think I haven’t change anything since I downloaded the scenes from ESA.

You can try to coregister the files after calibration. Use the image which is more correct as reference.

Is it correct to do so via Radar > Coregistration > Coregistration? Because I cannot find how to set one scene as the master in the window that opens.

yes, that one works for this kind of data.

You can first apply orbit files, thermal noise removal and calibration, then coregister both and check if they align well. If so, you can filter and terrain correct them and import the vector as a last step.

Thanks for your help! I found out why your advice didn’t solve my problems. I had problems collocation and stacking the two scenes because they don’t have the same track information. I solved my problem by choosing a scene from another day where the track is the same :slight_smile:

thank you for sharing. Maybe ENVISAT data from different tracks is too difficult over these large areas.