ESA Echoes in Space - Hazard: Volcanic eruption mapping with Sentinel-1

In this video we're going to apply the technique of differential interferometry to map deformation following a volcanic eruption And we'll do that with a Sentinel-1 Single Look Complex imagery The imagery includes two single complex Sentinel-1 scenes and if we expand the bands folder we will see that we have complex bands For each of the swaths of the interferometric wide swath product So here we have the complex bands, and we also have a virtual intensity band, which is created on the fly Now we have two products one acquired on the 3rd of November 2014 the second on the 27th of November And if we look at a map of the world we can see Where these images were acquired over an volcano of the coast of Africa If we look at one of the virtual intensity bands We can see that there are a number of bursts within each of these swaths Okay, so the volcano island Is situated on two of these bursts So the first thing we will do is only take one sub swath and the two burst within this sub swath And we'll discard the rest of the image to increase the speed of the processing We go to Radar Sentinel-1 TOPS Sentinel-1 TOPS split and here in processing parameters We can zoom in using the scroll wheel of the mouse and the right mouse button to pan We can see the various difference swaths We will take IW3 which covers the island volcano And will take the burst 1 to 2 so you can select the burst by By left clicking on this button here and dragging to include only bursts 1 to 2 And then we will select in the io parameters tab a suitable output file name We will remove some Parts of the file name and keep only the image acquisition date and time of the start of image acquisition And we will select run And then we will do the same for the second image Okay, then we can close this window And here we have our two split products The interferometric processing chain involves a number of steps First we will apply orbit files in order to obtain very precise orbits for each of the images Then we will coregister the two images in such a way that the pixels exactly correspond to each other After that we can compute the interferogram by multiplying one image with the complex conjugate of the other We will also calculate the interferometric coherence After that We will Subtract the phase due to the topography in order to have only the phase due to the displacement We will then filter that phase And then we will geometrically correct a product and then overlay that onto an imaging Google Earth so the first step is to apply orbit files we go to radar apply orbit file And here in the processing parameters we will leave everything is default and we will select run I will do this for both the products So here we have the products with the updated orbits files So the precise orbits are available some days after the image acquisition When more precise orbits are calculated Now we will do the coregistration the coregistration For the TOPS mode includes two steps we have the back geocoding and the enhanced spectral diversity, so we'll do one at a time essentially these both of these steps will ensure the images are very precisely coregistered and that involves use of the orbit information and also carrying out a cross correlation So we select back geocoding Here we can Select the plus icon to add all the inserted products, but we will remove those which we will not coregister so we will only coregister the images with the orbit files included So we will remove all other files, so here we have the images with the updated orbits And we will coregister those two products and We can remove the acquisition date And then we can select run Okay now we will apply the Enhanced spectral diversity And we will leave everything as default I will select run Once we have done this will have very precisely Corrected imagery which is essential for carrying out interferometry So here we have the two images coregistered You can double-click one and here we see only the two bursts, and here's the second one Okay, so this has been, this is a slave image that has been mapped, this has been resolved onto the master Now we can compute the interferogram so we go to radar Interferometric, products, interferogram formation And here we can leave all as default and we select run So here we have our interferogram And the interferogram is in the phase band of this image So if we close all other in viewers, and if we open this phase Then we will see here the fringes the interferometric fringes Each color cycle here corresponds to a Complete phase difference If we open also the intensity and the coherence or you compare them together We can see that In the coherence image there are some areas where we have high coherence and other areas where we have very low coherence And in the areas of very low coherence we do not see clear fringes in the interferogram Only in the areas of high coherence do we see these clear fringes What we are now do is deburst to remove this gap between the bursts so you go to radar Sentinel-1 TOPS Sentinel-1 TOPS Deburst And here we will leave all as default and select run So you can close these viewers and if we look at the Interferogram We can see that the product has been debursted What we'll now do is we will remove the fringes due to topography As the fridges are due to topography and displacement, and if you remove the topography then we're left with the displacement and we do this by calculating the fringes from a DEM and then removing those fringes from this interferogram So we go to radar, interferometric, products, topographic phase removal And here we can use the The default SRTM 3 arcsecond DEM to calculate the synthetic fringes and remove them from the interferogram So we select our debursted product, and then we select run So here we have the differential interferogram And we can open that here And if we compare The interferogram and the differential interferogram, while in the interferogram we see a clear cycle of fringes from the bottom of the volcano in the differential interferogram We see fringes that do not appear to only follow the contours of the volcano But we see more fringes in a certain part of the volcano And if we open this band here this image band topo phase we can see the Topographic fringes the fringes due to the topography that have been calculated from the DEM This Differential interferogram and interferogram itself as well is quite noisy and a common step is to do a phase filtering a filtering will Reduce the noise and it will help with the unwrapping of the interferogram So we go to radar interferometric Filtering, Goldstein Phase Filtering so the Goldstein is one algorithm for doing the phase filtering and here we leave everything as default and we select our differential interferogram and we select run So now let's compare the filtered and the non filtered phases So here we have our filtered phase and here we have our non-filtered phase So we can see that the filtered phase is much cleaner The next step will be to do a geometric correction of this image But first we will take a subset because we have a lot of no data values around The image here where we have masked out the sea And we're interested only in this island So we will take a subset of this island So we will go to Raster, subset And here we will Select a subset only of this small area So here we have our subset And now we will go to Radar, Geometric, Terrain Correction, Range-Doppler Terrain Correction And in the processing parameters tab we will select only the phase So the differential interferogram and the coherence And then in the input output parameters tab we will specify a file name which we will Leave with the default underscore TC terrain corrected suffix And then we'll select run You can close this and now here we have The differential interferogram and the coherence In two separate bands and terrain corrected What we will now do is to overlay these onto Google Earth So we right-click in the image, and we select export view as Google Earth kmz I'll leave this as default and we will click save to create a kmz file So here we have the differential interferogram as a kmz file and now We'll do the same with the coherence so a right-click and we select export view as Google Earth kmz And now we can double click on these to open them in Google Earth So here we can see the differential interferogram overlaying And we see the dense fringes corresponding to deformation so the denser the fringes the greater the deformation And we see that the fringes are over this part of the volcano, we do not see such dense fringes elsewhere You have here the cone of the volcano This displacement Corresponds to the collapse of the of the crater following the eruption and the fringes here corresponding to the displacement here Are due to the lava so the increased height from the lava flow So here we have positive displacement and here we have negative displacement And if we look at the coherence so now let's open the coherence image So here we see we have high coherence over these parts, and we have very low coherence here, and if we Compare the coherence image with the optical image we can see a lot of green here, so this could be forests Where the volume scattering would cause loss in coherence Whereas elsewhere we have a different kind of terrain which may be more stable and where we would have higher coherence We could go further and calculate the precise displacement from the differential interferogram This would require first phase unwrapping And then conversion of the unwrapped phases to displacement SNAP itself cannot do the phase unwrapping but this can be done by other software such as snaphu which is also a free software and SNAP can then import the unwrapped phases from snaphu and continue to calculate the displacement from those unwrapped phases I hope it was interesting. Thank you for watching

Channel: EO College

Views: 17,537

Keywords: MOOC, Radar, EO College, remote sensing, science, e-learning, ESA, space, data, Sentinel, Earth, environment, course, microwave, uni jena, Copernicus, TerraSAR, ERS, ALOS, Satellite, SRTM, Insar, dinsar, volcano

Id: VE38mGI8h-I

Length: 19min 41sec (1181 seconds)

Published: Sun Nov 05 2017