I have downloaded Landsat images of date 22 May, 2016 of Bangladesh. But as Cyclone (ROANU) occurred few days back of that day, the image contains hude cloud coverage. I have asked USGS for help but they replied that the cloud could not be removed??? Is it actually not possible to remove the clouds anyway??
I have observed that you asked a similar question earlier. Did you not find the links helpful that were suggested in the earlier discussion? Removing cloudy pixels from one Landsat image and replacing them with a cloud free satellite imagery from a different time period is one of the ways but if you are trying to study a phenomena that happened exactly on the same date when the Landsat image was acquired with the presence of clouds, this technique is useless. The other option is to look for alternative imagery such as Sentinel or ASTER or some other similar optical kind. If your aim is to study the pre and post disaster situation prevailing over Bangladesh before and after Cyclone Roanu then I would suggest you to look at SAR Images which can 'see through' clouds. Look at the standard practice at Disaster Charter (link attached). They have used Landsat 8 for pre-disaster and Sentinel-1 for post disaster analysis. If your aim is not to study the destruction caused by Cyclone Roanu then I am sorry, I cannot help much. I am afraid I have to agree with USGS in this regard. It would be better if you specify the aim of your study.
https://www.researchgate.net/post/How_to_remove_clouds_from_Landsat_Images#view=572e20b23d7f4b27c07613f1
https://www.disasterscharter.org/web/guest/activations/-/article/cyclone-in-bangladesh
As Mr.JAmes rightly said please check for Sentinal 1-A, C-Band SAR Datasets for extracting flooded area. You see one of my Conference publication related to Chennai Disaster.
Pl follow the link: you can find the detailed methodology as well for the area which I have extracted Chennai Floods. You can merge SAR and LAndsat 8 datasets. But that would just emphasize the display quality of images!
https://www.researchgate.net/publication/286447298_Floods_in_Chennai_and_Kanchipuram_Districts_of_Tamil_Nadu_Nov_2015_Is_it_a_Disaster_A_Challenge_towards_Sustainability_in_the_Indian_Coastal_Zone_Or_a_Verdict
Conference Paper Floods in Chennai and Kanchipuram Districts of Tamil Nadu, N...
Mr. Amin. Completely remove the clouds may not be possible. But you can try to make a Radiometric Correction and also try Pure Pixel Index. Both can help eliminate noise in images, as excess clouds.
Best regards.
Hi
you can download another image that its time near to your image and in Envi software you can remove clouds.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators