can i bypass sunglint correction and proceed with band calculation? how do i determine these glint samples? my images seems to have none of those observable lined features on images
Hi, this is mainly for shallow underwater object mapping/monitoring and it is not a procedural nor a mandatory step. Sunglints are white stuffs you can see on the water surface as a result of sun light being reflected to the sensor direction. For sampling, you can select areas affected by sunglint and areas free of sunglint. If you didn't find any sunglint on your images then you can proceed to other image processing procedure.
Initially, thanks a lot for the tutorial video! I tried to do the same process but my result is awful. The big difference between my data and your data, is that my data is L2A and your data is L1C. Is there any chance that the L2A data include the Sunglint Correction and i do not have to process them?
Hi, thanks for watching. The L2A data does include atmospheric correction but not sunglint correction. If you read the method paper, they used the simple image-based atmospheric correction (e.g. Dark Object Subtraction) rather than the one applied to produce L2A product from L1C (which involve atmospheric models). But I don't see this would be problematic and other factors might play role (e.g. sample selection). Could you please elaborate how awful is your result? In some cases, applying deglint does make things worse. Keep in mind that sunglint correction is not procedural nor mandatory, meaning you can skip it if you didn't find any sunglint in your AOI. When you encounter heavy sunglints, you can also opt for clearer images from other acquisition date. If you are into programming and automation, kindly watch the other video using GEE: th-cam.com/video/_V0X1LdSta4/w-d-xo.htmlsi=tlUjlDXOMlHT62b-
@@geografif i have tried the older version.. i found sen2coral, but i couldn't add the sentinel product, always show that "unable to read metadata from MTD bla...bla...
Thanks for this! I have been trying to deglint an image that only extends 1km from shore. I am worried that deglinting may do more harm then good as there is not any deep (dark regions) what do you think?
@@geografif I wouldn't say the image I am deglinting has much glint at all. I'm just trying to develop a systematic approach as I will be monitoring bathymetric change in the region over a 5yr period.
Perhaps you could try estimating the depth with input of deglinted and undeglinted images and compare the accuracy to justify wether you need it or not
@@geografif Unfortunately I haven't been able to find a decent tutorial anywhere. I tried following the official documentation but ran into a dead end. Thanks for responding.
@@geografif Last time I was was looking into this I did get some help but still got stuck! It's something I need to look at again tbh. I randomly came across your video today and it made me think about finding a solution again.
Hi its so helfull! but I keep running into the same error with multiple files where upon pressing run it instantly fails with the error “already contains a tie-point grid with the name ‘tco3’”. I would be very grateful for some feedback on how to debug this, thanks
Hi, for Sen2Coral please check around 0:30. For Sen2Cor, I think there are other channels who already posted tutorial for that. So I probably wouldn't do it.
Check out how to perform sunglint correction in Google Earth Engine: th-cam.com/video/_V0X1LdSta4/w-d-xo.html
English and Indonesian subtitle is available, so please turn it on. Thanks.
how to plugin sen2coral in SNAP ?
@@kevinjean5444 check 0:10
@@geografif how to compare a channel grub for glind sample? i can't fixed ??
the colour band is broken after this
You can display a true color composite and draw samples where glints are present and not present
Thanks for the great video, very helpful! :)
Glad it can be helpful :)
can i bypass sunglint correction and proceed with band calculation? how do i determine these glint samples? my images seems to have none of those observable lined features on images
Hi, this is mainly for shallow underwater object mapping/monitoring and it is not a procedural nor a mandatory step.
Sunglints are white stuffs you can see on the water surface as a result of sun light being reflected to the sensor direction. For sampling, you can select areas affected by sunglint and areas free of sunglint.
If you didn't find any sunglint on your images then you can proceed to other image processing procedure.
Initially, thanks a lot for the tutorial video! I tried to do the same process but my result is awful. The big difference between my data and your data, is that my data is L2A and your data is L1C. Is there any chance that the L2A data include the Sunglint Correction and i do not have to process them?
Hi, thanks for watching.
The L2A data does include atmospheric correction but not sunglint correction. If you read the method paper, they used the simple image-based atmospheric correction (e.g. Dark Object Subtraction) rather than the one applied to produce L2A product from L1C (which involve atmospheric models). But I don't see this would be problematic and other factors might play role (e.g. sample selection). Could you please elaborate how awful is your result?
In some cases, applying deglint does make things worse. Keep in mind that sunglint correction is not procedural nor mandatory, meaning you can skip it if you didn't find any sunglint in your AOI. When you encounter heavy sunglints, you can also opt for clearer images from other acquisition date.
If you are into programming and automation, kindly watch the other video using GEE: th-cam.com/video/_V0X1LdSta4/w-d-xo.htmlsi=tlUjlDXOMlHT62b-
I didn't find sen2coral in SNAP version 11. could you guide me please?
You can either try the older version (e.g. 8) or do it by scripting: th-cam.com/video/_V0X1LdSta4/w-d-xo.htmlsi=o6LnCzwZ3YhMuuee
@@geografif i have tried the older version.. i found sen2coral, but i couldn't add the sentinel product, always show that "unable to read metadata from MTD bla...bla...
Try to open the individual .jp2 file if you downloaded the data from Copernicus hub
Thanks for this!
I have been trying to deglint an image that only extends 1km from shore. I am worried that deglinting may do more harm then good as there is not any deep (dark regions) what do you think?
Hard to say. Depends on what image you're using. Would you have acess to image from another acquisition where glints are not there?
@@geografif I wouldn't say the image I am deglinting has much glint at all. I'm just trying to develop a systematic approach as I will be monitoring bathymetric change in the region over a 5yr period.
Perhaps you could try estimating the depth with input of deglinted and undeglinted images and compare the accuracy to justify wether you need it or not
@@geografif thanks I will
i just downloaded a stacked planet data, is there any way you could demonstrate to us th sun glint and water column correction in QGIS or ArcGis?
Why dont you try SNAP that has built-in plugin to do that semiautomatically? It is free.
@@geografif it is not working in my laptop
Have you had any sucess in running the Sen2Coral Shallow Water Analytical Model (SWAM) in SNAP?
Hi, I have not tried it yet unfortunately
@@geografif Unfortunately I haven't been able to find a decent tutorial anywhere. I tried following the official documentation but ran into a dead end. Thanks for responding.
@@markbrown3321 Have you post your problem to the ESA STEP Forum and see if anyone has solution?
@@geografif Last time I was was looking into this I did get some help but still got stuck! It's something I need to look at again tbh. I randomly came across your video today and it made me think about finding a solution again.
@@markbrown3321 good luck, then
Hi its so helfull! but I keep running into the same error with multiple files where upon pressing run it instantly fails with the error “already contains a tie-point grid with the name ‘tco3’”.
I would be very grateful for some feedback on how to debug this, thanks
I've never encountered such problem. Probably have something to do with the software version.
what version are you using on this video? @@geografif
Been a while. I think it was SNAP 5.
Hello mister how to install Sen2cor on snap? can you to do a movie sure it
Hi, for Sen2Coral please check around 0:30.
For Sen2Cor, I think there are other channels who already posted tutorial for that. So I probably wouldn't do it.
Hi @Ibrahima barry, if you still need it, this tutorial seems on point: th-cam.com/video/BV46V_roYts/w-d-xo.html
cara menambahkan sen2coral kit bagaimana?
Halo, silakan mampir ke detik ke 0:08. Pada toolbar atas klik Tools > Plugins > pada tab Available Plugins pilih Sen2Coral Toolbox, dan install