I have retrieved CO2 concentrations from EnviSat sciamachy's data sets. But for particular months (June, July, August and September) the data quality is poor. In the sense there were very few points in the area of my interest (whole India). But for the remaining months the more or less the data covering is satisfactory.
I feel that Interpolation is not a good idea to find out the missing information. Is there any other way or idea to find out the missing data points (contains CO2 concentrations) in this region.
I suggest looking at this very interesting paper:
"A three-dimensional gap filling method for large geophysical datasets: Application to global satellite soil moisture observations", freely available here:
http://www.biomecardio.com/pageshtm/publi/envirmodellsoftw12.pdf
I think that simple interpolation is not the right way to face the problem.
More significant results could be applying the Variogram and kriging method using Geostatistics interpolations, They account for the statistics of "distances" between value observed and their spatial distances
Another way is to impose some behavior to the shape of values measured with time. The seasonal cicle of CO2 is well Known and you can use it to estimate, together with your values, an estimation of values in the gap.
I suggest looking at this very interesting paper:
"A three-dimensional gap filling method for large geophysical datasets: Application to global satellite soil moisture observations", freely available here:
http://www.biomecardio.com/pageshtm/publi/envirmodellsoftw12.pdf
You should use geostatistical methods for the interpolation of missing data of your satellite images.
Dear S.Masieri, Luca Brocca and Elif Sertel, Thank you for your valuable suggestions. I have gone through the geostatistical interpolation methods and I am feeling that those are good enough though I didn't make any check. And I will definitely look into the paper suggested. Once again thanks for your answers.
You could actually use the new envi 5.0 to correct gaps on landsat imagery. Its an extension, landsat_destripe.idlbinaryfile
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing