agrouaze
commented
1 year ago So far, Doppler centroid annotations have been mapped like this:
├── DataTree('doppler_estimate')
│ Dimensions: (azimuthTime: 10, nb_fine_dce: 20)
│ Coordinates:
│ * azimuthTime (azimuthTime) datetime64[ns] 2018-01-01T03:...
│ * nb_fine_dce (nb_fine_dce) int64 0 1 2 3 4 ... 16 17 18 19
│ Data variables:
│ t0 (azimuthTime) float64 0.005339 ... 0.005338
│ geometryDcPolynomial (azimuthTime) object -0.612125 - 461.1503·x...
│ dataDcPolynomial (azimuthTime) object 15.83866 - 32002.58·x¹...
│ fineDceAzimuthStartTime (azimuthTime) datetime64[ns] 2018-01-01T03:...
│ fineDceAzimuthStopTime (azimuthTime) datetime64[ns] 2018-01-01T03:...
│ dataDcRmsError (azimuthTime) float64 7.03 5.025 ... 3.487
│ slantRangeTime (azimuthTime, nb_fine_dce) float64 0.005345...
│ frequency (azimuthTime, nb_fine_dce) float64 15.66 .....
│ dataDcRmsErrorAboveThreshold (azimuthTime) bool False False ... False False
│ Attributes:
│ history: annotationsNo feedback to change this xarray.Dataset -> I close this issue.
As discussed between @oarcher and @agrouaze in March 2022, the link between the doppler centroid estimates and the image dimensions (atrack/xtrack) is absent for now. this issue is not addressed in the #51 by purpose. This link is not needed for now but for some reasons it could be interesting to get the doppler estimates for a specific piece of a SAR image. to do this link 2 approaches will be possible:
s1meta.geoloc['azimuth_time'](20x10) and then returning the the atrack,xtrack coordinates associated.or
s1meta.burst_azimtime())vector and finding the closest azimuth_time DCE to return the atrack,xtrack index.For now the link between the DCE given only along azimuth_time coordinates and the fine DCE given along the slant_range_time dimension is not clear with respect to the
s1meta.geoloc['azimuth_time']ands1meta.geoloc['slant_range_time'].This issue should be addressed only if necessary and once sufficient knowledge on those information will be available.