Lets take a look at the ‘Sierra de la Demanda’ (northern Spain) and see if there is significant snow over there this days. We are going to get the images from the Copernicus SciHub service.
Disclaimer: Only Open Source Software was used during this process
Search for the images in our area of interest using the Open Search SciHub API:
This query search for Sentinel 2 images, with cloud percentage lower than 10, date 2017/11/28 to 2017/12/01 on our area of interest:
https://scihub.copernicus.eu/dhus/search?q=( footprint:”Intersects(POLYGON((-3.3299599372780615 42.115249075987435,-2.6160662295430885 42.115249075987435,-2.6160662295430885 42.51117735999185,-3.3299599372780615 42.51117735999185,-3.3299599372780615 42.115249075987435)))” ) AND ( beginPosition:[2017-11-28T00:00:00.000Z TO 2017-12-01T23:59:59.999Z] AND endPosition:[2017-11-28T00:00:00.000Z TO 2017-12-01T23:59:59.999Z] ) AND (platformname:Sentinel-2 AND cloudcoverpercentage:[0 TO 10])
Download Level 2 Images:
Sentinel 2 has the MultiSpectral Instrument (MSI), all data acquired by the MSI instrument is systematically processed to Level-1C. Level-2A products are generated on the user side through the Sentinel-2 Toolbox.
The Level-2A processing includes a scene classification and an atmospheric correction applied to Top-Of-Atmosphere (TOA) Level-1C orthoimage products. Level-2A main output is an orthoimage Bottom-Of-Atmosphere (BOA) corrected reflectance product.
We select the two Level-2A images to download and process. The download is made with the Scihub OData API:
If we open the two images on SNAP, we see that they overlap a little:
We are going to make a mosaic with the two images into a single composite product. Mosaicking is achieved based on the geocoding of the input products, so they need to be terrain and radiometric corrected first. For this purpose level 2 images are good enough.
Snap is a free software created by the ESA to facilitate the exploitation of Earth Observation data. Let’s see the steps on SNAP:
To do the mosaic over all bands we need to do a re-sampling so all have the same resolution. Go to raster / geometric operations / resampling / select B2 as reference band.
Then use the mosaic operator: raster / geometric operations / mosaicing and select:
map projection UTM/WGS84
Mosaic bounds pixel size: 10m
variables&conditions: B2, B3, B4
Finally we do a subset center on the ‘Sierra de la Demanda’ area: raster /subset /
Open a RGB window: Window / Open RGB image window / red B4, green B3, Blue B2.
Finally, export the image to a .bmp file: File / export / other / view as image (select full scene)
Sensor on Satellites have a huge range of sensibility because they need to cover great areas, from deserts to mountains. The information that we get from the sensor is reflectance (amount of light reflected), when we put this information in a RGB image, the range of values found in the zone are smaller than the full range of values. This make low contrast images, they are darker than we expect. So, we need to increase the contrast and brightness of the image to make it looks more realistic.
We’ll use imagemagick to do some color enhancement :
$ convert sierra-demanda_RGB.bmp -monitor -modulate 100,130 -channel G -gamma 0.95 -channel B -gamma 0.875 -channel RGB -sigmoidal-contrast 8x20% -compress lzw output.jpg