Urgent technical support needed to overcome an albatross - fusing landsat and spot 1-4

Hello dear Forum Members,

I must say that I am cutting my teeth on geo spatial software and manipulation of remotely sensed data and so I urgently need tecnical support.

I am carrying out a research to assess the impact of climate change on vegetation dynamics and I am measuring VIs ( NDVI)
Unfortunately, over my AOI, there is a huge challenge of data inconsistency and gaps.

Given that I am carrying out an assessment of rural communities of about 140km, I desire data with moderate spatial resolution and landsat and SPOT 1-4 falls into this category.

I elected to use Landsat but the data coverage under the timeline of review is poor due to the absence of sensor and the interruption caused by the commercialization of Landsat during the period. To this extent I have only been able to retrieve 3 clear images in a period of 10 years. After the 2003 Landsat SLC off issue, there are about 7 more appreciable Landsat images.

To this end , I am procuring data from SPOT which unfortunately also has only 9 cloud clear images over the AOI.

Consequently, I am strongly considering the data fusion of SPOT and Landsat and I would like to know whether SNAP could be useful in this regard.

For a temporal dense investigation in a timescale of 25 -30 years, I would be comfortable with at least 20 images over this time line to support the optimal trajectory investigation of changes in NDVI using a R statistical software.

Can SNAP accurately do this? or are there any suggestions? I would therefore be very grateful if anyone in this forum can offer useful ideas and advice / technical support on what data fusion software I can use and maybe a video tutorial / documentation on how to use the software for fusing LANDSAT and SPOT.

With kind regard


I can only say that there is no special tool for a data fusion.
There are tools like collocation and mosaicing or you can manually do band shifting using arithmetic expressions on bands. But there is no data fusion support.
Maybe someone can suggest how Kansy can proceed.

I agree with marpet. I’d personally first geocode both data sets, then collocate the rasters and then maybe perform a principal component analysis or unsupervised clustering in order to fuse information from both sensors.