While course resolution derived SM products like SMOS and SMAP can well describe temporal soil moisture dynamics, they lack spatial detail and we can’t use them in analysis of local hydrological patterns. Contrary, SM product from sentinel-1 (A/B) sensor can resolve dynamics at this spatial level. However, its temporal resolution is less frequent and thus not suitable for acquisition of short-term variations. To overcome these spatial and temporal scale gaps, we can use data fusion approach that fuses multi sensors data. Is there any method which could help me do this?
I worked with Clark Labs' Idrisi for 5 years and found a couple of approaches that might help you.
With multiple bands there is a lot of multicollinearity in the data. Idrisi has the facility to do principal components analysis (PCA), which generates orthogonal components.
If you have bands of different resolution, you can transform by a variety of methods. One I used to fuse 30 meter and 60 meter was to expand the 60 meter by 2 times to generate a 30 meter band.
You do not get any extra information from the 60 meter band by doing this. However, because of multicollinearity with other bands, the analysis may allow the aggregate amount of information to be increased. Using the PCA cooeficients you can then generate an improved version of the 30m band you created from the 60 m band.
If you have a lot of noise in the data, PCA can be used to reduce noise. Clark Lab's manuals are very clear on the procedures and the academic discount is very generous.
Dear Frederick
Thanks a lot. I used PCA on Landsat and IRS data in ENVI; but now, I want to transform Sentinel-1 (is an active Radar data in C-Band) to Daily scale. nowadays it's accessible in temporal resolution of 24 days and also I have not access to it's resolution of 3-6 days.
Ayoob, I don't believe that your plan is feasible.
But it really depends what you intend to do with the result.
I am assuming that you are either prohibited from or do not have the budget for the 3-6 day dataset.
we need to have soil moisture in high spatiotemporal resolution. totally, Sentinel is a free data and don't have any cost but in some product it is not downloadable yet. also it will be accessible as soon as possible.
Yes, I checked out the terms and found it is still free.
What do you mean, that it will be accessible? Do you mean that in a month or so, you will have 3-6 day data?
Finally i found that sentine-1 could be combine with Radiometer data like SMOS.
I think there is no way to fused sentinel with optic data right now.
- Badges
- Science topic
- Similar topics
- Geoscience
- Hydrology