Can I merge bands from separate files into a single file?

One product has hh,hv, vh,vv bands. Another product has entropy,anisotropy,alpha bands. I want to combine these bands into one single product and then export it to csv. How can I do that?

1 Like

You can follow these instructions:

Or you use the collocation tool (Raster / Geometric Operations / Collocation)

1 Like

Hi, sorry to disturb you. I am a graduate student from China. I recently used SNAP to extract the leaf area index of sentinel-2, but there were some problems that I could not solve.
I first downloaded the 2020 sentinel-2 surface reflectance data using GEE. But I know that using SNAP to calculate leaf area index requires sentinel-2 metadata. Specifically including B3, 84, 85, 86, 87, B8A, B11, B12 and sun_zenith etc. The problem I am facing is that the data downloaded by GEE does not have metadata information, that is, four angle data such as sun_zenith are missing. So I came up with the following solution.
1. Use the USGS website to download the corresponding sentinel-2 data I need and resample to a uniform resolution because it contains metadata.
2. Then, use the SNAP bands extractor tool to extract the four angle data, and then use the Raster-Geometric Operations-Collection tool to merge the four angle data with the data downloaded by GEE.
3. Finally, use the combined data to calculate LAI. But it prompted me that an error occurred: missing band at 560.0nm, etc… So I need to manually open the properties of each band to correct this error. In the end, although the leaf area index map is obtained, the value range inside is between 23-33, which is obviously wrong.
During the calculation, I also found some other problems, such as the inconsistency of the size of the two data; inconsistent projection coordinates, but I have not taken these factors into account, so I don’t know if these will affect the final calculation results. . I don’t know if my method is feasible or not. Do you have any better suggestions? Looking forward to your reply.

As you have noticed the GEE data is not supported because it lacks essential information.
But you can calculate the LAI directly on the USGS data. There is no need to merge USGS data with GEE data.
Also, the data from the Copernicus Data Hub works.

Thank you very much for your reply. I chose GEE data for the following reasons:

  1. I am currently doing time series research. GEE is very convenient and simple to download data.
  2. Because some images are of poor quality, I need to eliminate cloud pollution and repair the data. In addition, I also perform harmonic processing on the data. And this requires tif format. So I used to download data from GEE.
    If USGS data is used to calculate LAI, I need to spend a lot of time reprocessing a large amount of data. So if I insist on using GEE data to calculate LAI, is this method correct? Or do you have any better suggestions?

If you still want to follow your approach you need to resample the data first to be able to merge them.
It is strange that the data has different dimensions.
If this is done, you can merge the data. You can create graph to do it from the command line ( Bulk Processing with GPT).
The graph would contain the following steps:

  1. Merge tiff bands by band math. Add wavelength and maybe scale values
  2. Resample the result to the size of the other (or the other way around)
  3. Merge this result with the result of step 1 by using Merge operation.
  4. Use the result of step 4 for the LAI computation.

Regarding the strange result of the LAI calculation, the cause might be that the GEE Tiffs contain raw values, and they are note scaled to reflectances. Maybe the raw values need to be divided by 10000.

I don’t understand why you would need to reprocess a large amount of data. I thought you just start with the processing, and it didn’t work yet.
Before going through the full process of making the GEE data usable, it might be easier to process the USGS data, because you have it at hand anyway.

Thank you very much for your reply. But I recently discovered that the SNAP version has been updated to 8.0. Some changes have taken place in the interface for calculating LAI.


When I resample the sentinel-2 data (from usgs) to a resolution of 10 meters or 20 meters. An error occurred in calculating LAI. But this can be achieved in the previous version. may I know what is the reason?
In addition, when using USGS data to calculate LAI, in addition to resampling, what preprocessing is required? Are atmospheric correction and normalization steps that must be processed? The process of calculating LAI is very confusing to me. I hope I can get your reply.

tingge123456 and I are the same person