# How can I filter my image collection based on incidence angle in SAR?

Im completely new to SAR, and Im trying to understand the basics.

Im studying a small area in Google Earth Engine, and Im making timeseries of water pixels with SAR imagery. However, I observed significant differences between some days, and when I looked the angle from those images, there were like 31 and 40.

Therefore, how could I filter my image collection based on angle? Something like to stack images with angle 30-35, and 35-40, etc. Which function can I use in the filtering part?

``````Map.addLayer(vuursteentocht, {}, 'ROI')
Map.centerObject(vuursteentocht, 8)

var collectionVV_desc = ee.ImageCollection('COPERNICUS/S1_GRD')
.filter(ee.Filter.eq('instrumentMode', 'IW'))
.filter(ee.Filter.listContains('transmitterReceiverPolarisation', 'VV'))
.filter(ee.Filter.eq('orbitProperties_pass', 'DESCENDING'))
.filterBounds(vuursteentocht)
.filterDate("2017-10-30", "2018-02-21")
print(collectionVV_desc);
``````

## 1 Answer

You can mask every image based on the `angle` band like this:

``````var maskedAngles = collectionVV_desc
.map(function (image) {
var angle = image.select('angle')
return image.updateMask(
angle.gte(30).and(angle.lt(35))
)
})
``````

An alternative would be to only use imagery from a specific orbit:

``````var relativeOrbitNumbers = ee.List(collectionVV_desc
.distinct('relativeOrbitNumber_start')
.aggregate_array('relativeOrbitNumber_start'))

print('relativeOrbitNumbers', relativeOrbitNumbers)

var fromASingleOrbit = collectionVV_desc
.filter(ee.Filter.eq('relativeOrbitNumber_start', relativeOrbitNumbers.get(0)))
``````
• Thanks!! So, for the first way: after i make the mask, should i apply it to collection with the .map function or something like this? Because if i print the collection again, im getting the same amount of images (48). What i need is to keep only the images with angle between 30-35, in my roi. Commented Apr 21, 2020 at 10:35
• The above code is not creating a mask, it is mapping over the collection, masking all images. The result is a new collection, with the same amount of imagery, but where pixels where the angle is outside of the range is masked. Commented Apr 21, 2020 at 10:38
• The second approach is simply filtering out complete images from the collection. Commented Apr 21, 2020 at 11:02