Hi. I’am a freshmen, who would like to use Sentinel-1 in agricultural applications. But, after downloading some data on Stavropol from esa_hub (period from 01.june.2015 to 01.august.2015), i got some problems with many scenes. Intersection of two images have a significant offset (for example, river on one image dismatch with the same object on other image). How can I solve this shortcoming ? I tried to use Worksync Station in the Erdas, but it hadn’t worked. All of products is “S1A GRD VH”. Maybe, i should design my preprocessing workflow in SNAP and how i should do it ?
Are you using SRTM 1 Sec HGT for terrain correction? It is missing [at least on the server defined in SNAP] over the area you are interested in, so s1tbx will use 0 height, which leads to significant location offsets (order of >100 m). See also "Empty region" bug
Try using SRTM 3 Sec. Or wait until SRTM 1 Sec is copied.
Thank you, i’m trying now. I have another question: if i properly understood, there are not some classification techniques for SAR in SNAP? What format shall I choose into “Save as” checkbox for using geocorrected images in Erdas 9.1 and for Envi 4.7 ?
You could export as GeoTIFF or BigTiff but you may loose metadata.
SNAP supports some general clustering and for polarimetric data Whishart classification.
Erdas 9.1, trying to open BigTIFF from SNAP, failed in the next: “File is larger than 2G. Size will be reported as 2G” or “unsupported raster format”. =(
I guess you will need to subset first before exporting to something within the GeoTIFF limit.