I have tried the l2bin GUI tool, and put all of the images into the input list - the tool runs but I do not get a meaningful output file. I have tried doing each image individually, but get an error on the second image. Clearly I am not using this l2bin tool properly. Any suggestions for how to accomplish this goal? I appreciate the help.
The output of the l2bin program is a bin file, the format of which is not readily displayed in SeaDAS as it is not a raster. Use l3mapgen to create a map to visualize the data.
Thanks for the reply. I used l3mapgen to create the raster of my binned file, but the resultant displayed image has almost no valid pixels. A few images have some NaN values, but they were all chosen because they were cloud-free. I tried to attach some images to show you what I see, but that doesn't seem to work on my browser. What I was looking for was to be able to get an average of all the lake Erie pixels in the 10 input images, and have a single, averaged, image file to then analyze with our processing.
I have also geo-rectified each image, and can collocate them in succession to get a single file with all the bands from all the images, on the same grid, so that I can then average each band myself to get the end product I am looking for. But this seems like a really long process & I was hoping there was a better tool that I am not seeing.
Thank you for all the help!
If your binned files were created running l2bin with default flag settings you might find it helpful to provide your own flag settings. The flag masks in the SeaDAS 7 GUI are helpful in understanding which flags are masking pixels you want to keep. You can also use l3bindump to get ascii files to do your own central tendency calculations using robust estimators. Using the coordinates of each bin you can create a mapped image of the central tendency results.
Mapping each level-2 file to a common grid is widely used, but you need to be aware that level-2 pixels near a pass edge will generally be mapped to several pixels in the resulting images. Since edge pixels tend to be lower quality, this approach tends to overweight lower quality data. This approach has the advantage that you can examine each level-2 image on the same raster to understand general patterns and it allows you to apply robust methods.