RUS webinar: Rapid Landslide Detection with Sentinel-1 - HAZA07

hello everyone and welcome to the 26th webinar in the Rose series my name is Georgia Cardinal and in this webinar we will learn how to detect rapid lance lights using Sentinel 1 data and there is a snap Sentinel 1 tool box we will analyze the foie grass color hello a landslide that occurred in west Iceland in July 2018 by using the SAR interferometry technique and exploiting the face amplitude and coherence characteristics we will talk a bit about the study area and have a short introduction to landslides The Sentinel 1 constellation the root service then we will continue to the exercise in gross virtual machine we will have the take-home message and a Q&A session will follow the webinar will last around 1/2 hour and please be aware that this webinar is being recorded and you will be able to repeat the exercise by yourself later since this recording will be available both in our YouTube channel and our training website please also note that it is advisable to start sending us your questions immediately as soon as you have one and throughout the webinar since we're here with my colleagues to answer to as many of your questions as possible this way we will not have too many questions accumulated occur the Q&A session now let's have a look to our study area as mentioned the landslide luckily occurred on a remote part of Iceland on July 7th 2018 and based on former satellite data the hillside was known for its instability since 2015 it was one of the largest recorded landslides with an approximate volume of around 10 to 20 cubic million meters here we can also see the area that they breathe from the landslide occupied but let's see some general information about locks lights and destruction in general landslides occur in a variety of environments and can be categorized based on the type of movement and the type of material involved they are common in areas of steep topography and can be triggered by both natural processes and human activities they can lead to damage of infrastructure or blocking of water sources and in the worst cases in loss of life the type of the one we studied today belongs to the debris fall next we can see the classification Illustrated the primary driving force for a landslide to occur is gravity but then there are other factors that affect the slope instability as well in many cases the last slide is triggered by a specific event such as a heavy rainfall or an earthquake the landslide we are studying was likely triggered by the large volume of rainfall that Iceland received during the 2018 summer the ronald angle of the landslide is around 12 to 13 degrees which is relatively long in comparison with other large landslides the last light cross the river below partially blocked it creating a dam and a lake was formed above the debris tank since the landslide fell from an area that showed evidence of earlier ground displacement it might be considered to have been an old landslide deposit the area of the debris tank is around 1.8 square kilometer and the debris is apt 20-30 meters thick okay now let's see quickly some information about the sentinel one data we will use in this exercise we will be using Sir data provided by the Sentinel one satellite the Sentinel satellites are included in the space component of the Copernicus program of the European Commission and the European Space Agency The Sentinel one mission is formed by a constellation of two twin satellites paced at 180 degrees to each other it is an active sensor working on C bond it includes a right looking active phased array antenna providing fast cutting in elevation and azimuth and it provides data with a short repeat cycle of six days and with different imaging modes here you can see the four different acquisition modes which we will use the iw which is meant for land applications and for products of the modes of which we will use SLC since it contains complex data of amplitude and phase that is necessary for interferometric processing in the next slide you can see the 3 subs was the interferometric wide swath consists of and the 9 burst of each one but let's talk now a bit more about saiyr differently when we obtain 2 radar images over the same area that have the same orbit and have the same geometry we can create an interferogram which represents the deformation between two Sara positions by calculating the phase difference of those two products then the parametric phase consists of frenzies and one full color cycle has values from minus 3.14 to class 3.14 within the flow metric processing we can also update information about the coherence of two images this means how well correlated is each pixel of the slave image to the corresponding to the master image pixel the closer to one the value the more coherent damages let's continue now to some information about through service Bruce stands for research and user support for Sentinel core products it is an initiative funded by the European Commission and managed by izi with the objective to promote the uptake of copernicus Sentinel data and support research and development activities the service provides a free and Naughton scalable platform in a powerful computing ermand hosted a suit of open-source tool boxes pre-installed on virtual machines which allow you to handle and process the data derived from The Sentinel satellites what does this mean well with a large amount of data produced by The Sentinel satellites the challenge is no longer data availability but rather storage and processing capacity to solve that ruse offers virtual machines so that you have the appropriate computing environment to handle the data in addition to all that Russ also provides a specialized user help desk to support your remote sensing activities with Sentinel data in a dedicated training program you can find all information about the project and how to register on our websites so I recommend you to check them after the webinar to get familiar with a service and see that VM specifications list of software etc I will quickly show them to you before we start the exercise as I mentioned in the beginning we also have a YouTube channel where you can find all the videos of our previous webinars and you will also find this one in a couple of days now let me show you quickly all the processing steps we will perform in the exercise first of all we need to create a stir the two products that we will use the two Sentinel 1 images so what we will do is we will use a processor that is reading both of them then it is splitting them because we need to select in which sub swath and in which burst our area belongs to because we cannot process all of them at the same time and we don't need to then we will apply the orbits and with this step we will have accurate satellite position and velocity information and then we will create a stir with the batch according and has spectral diversity operators finally from this output we will create energy bill image there we will be able to detect the changes from the first second image so we will detect the landslide after that we will proceed to the insert steps so the product we will rate will be the output the courage third one we will first create the interferogram then we will apply the diverse option then we will apply the topographic face removal operator to remove the face due to topography then the next step is multi looking where we want to obtain square pixels in our images finally we will filter the face and we will have an output so once we have done that then we just need to geocode them we need to create an output and be read in Google Earth and in QGIS we can also subset the area if we don't need it all and finally explore the products so let me now show you quickly the websites so here we have the routes des training dot EU website where if you click in the training you can see all the upcoming and the past events we have we are organizing for example if I go to the past let me go to a past one if you click on more info you can find the recorded video of the webinar plus a to initiation document you can also go to our learning courses and practice on different topics we offer then we have the reus portal where you just create an account here you click on login register you can create your account and once you have it you can log-in I have my credential saved and here you can find all the information you need for example if you go here to the routes library learn by yourself in the routes lecture section you can find all the tutorials from our previous topics we have created you can navigate to the other tabs go to the route software and see the computing environment the training and what we can offer to you so once you have that once you have your account you can go to your route service your dashboard and here you can request a new user service you can specify the area of interest the products you want to use and the purpose of your project and you can have a virtual machine based on your needs once you have it it will appear here on the lower part and then you will click on access my virtual machine and you will access with the credentials you have this page is redirecting you to a guacamole window where you insert your credentials and then you can use the virtual machine as I mentioned we also have our YouTube channel where you can go and find all our previous webinars and you can practice them and you can repeat them following the tutorial ok but now let me continue to the roots desktop so this is how it looks like so let me go to full screen mode so that it is a bit bigger ok here you can see that we have all the software pre-installed and it is delivered like that when you receive your virtual machine the first thing we need to do to start our exercise is to download the data so let me click here and once you click here you have these two tabs by default opened so here you can access the OP and hub and download all the data you need and here you can go in there routes user forum and write any questions you might have so let's go back to the open access hub if we click on open hub it redirects us here so I have it already opened so let me close this window and go here so you come here you log in with your account it's very simple to create one and then you navigate over the area that is of the interest for your study case so here we came here to Iceland to the west part and we drew a polygon we come here we select this one we draw a polygon and then we click here on the left and we come here to set the parameters the landslides occurred on July 7 2000 18 so what do we need to do we need to have a limits which is before the event and a minute which is after the event to see the changes in the area to be able to detect the landslide so the first one we will search for is we will come here in the same zyk period and we will select from this calendar the 23rd June 2018 we will also do the same to the second part we will search for the images one by one so here we select mission Sentinel 1 as product type from the list we select SLC as we mentioned before and we select the iw we also select here the relative orbit number to be 16 because as we said we want both of the images to be of the same geometry so that means that they both need to belong to the same relative orbit once we're here we click here on search and you can see that this hurts give us back one result one image but here you see that this image is of line since one year I'll have now all the products that are in the Copernicus open access hub that are older from one year they go of line so that we save some storage space but once you want a product in order to be able to download it you click here on download product then you receive this message that says that you successfully initiated the retrieval of the product you click ok and this goes immediately in your cart remember with your account you can usually request one product every third minutes to 1 hour the product is usually available in a few minutes after your requests sometimes if the copernicus open access hub is very busy of requests it might take some some more time or if there are many users waiting for retrieving products you might not be able to request your product so you will get an error message but be patient and once you will manage to request it it will become online but remember that when the product becomes online you don't receive a notification you need to go to your cart where you see that your product is stored in there and once it will become online this gray area will become white and you will be able to download it once the product becomes online it remains like that for 4 days so that more users can download it if needed ok so we have the one product now for the second one you will check the rest of the parameters as we have set them before and we will go here on the dates and we will just go and select the 17th of July which is an image after the event we start again and we see that this product also is a flight so we requested the same way once they are downloaded both products we move them to the folder we want to use ok let me close this one we don't need it anymore and now we can start our exercise let's open snap in case you're not familiar with snap this is how it looks like here we have the menu on the upper part on the Left we have the product Explorer window where we load all the products and each one that is created appears here on the left on the right we have the View window which means that whatever product we want to visualize and we click it from the product Explorer to open it opens here in this view and then we can go here to the left to the navigation and we can see in this small window whatever is in this big view window also we have the world map so every product we insert in snap we can see where exactly it is located on the globe and many other tabs ok so first of all let me load the products that we will use so I have it open here so I click on the first product which is of the 23rd June 2018 I drag it and drop it on the product Explorer window once it is added I do the same for the second one if I go now to the world map and I click here I see where the products are located on the globe let's have a better look now in the products let me expand the first one so as you can see here you have some information some folders metadata vector data type point grids Quick Looks and baths the important we need to navigate at is the band's one so if I open it we see here that we have some bats as we mentioned in the introduction every product contains three sub spots so here we have all the three of them and since this product is of a dual polarization this means that it contains information for both pH and BV polarizations we will open all the intensity bands for the three sub swaths of the first product only for the VV polarization so first I click here on this then I click on the second one then on the third one and we wait until they open all the bands hurry up and now so what I will do is I will go here to window tile horizontally so that I have them all in parallel if I go here to navigation tab and I need to have these two options selected this way I can synchronize the cursor position across multiple image windows like here that I have the three of them and also synchronize the views let me just zoom to all of them so that we see them so here if we zoom in we will see that we have nine bursts as mentioned the product consists of three sub spots so this is the first the second and the third if we go now back to our world map we can see that these images are actually reversed why is this happening so this product we have downloaded it was acquired on an ascending path that means that the satellite was moving from south to north so if we zoom in we can see that here between the bursts we have these areas that seem to be black so in the edge of each one of its bursts there is information contained that is being overlaid with the edge of the next one this is to help us to manage better the data so that we don't have big volumes of them they are split in small parts so this problem will be solved in the deburred step now that we have seen that we can go on the second product and if we expand it we can see that it contains exactly the same type of information as well we do not have to open them all so what we know is that our area of study belongs here in the second sub swath on the sixth burst so as I mentioned in the introduction the first thing we need to do is to put those two images together to correct stir them how are we gonna do that first let me close the views okay so we will go here to reader correlation send the little one top scored stration sent in and one top score iteration with enhanced spectral diversity once I open this default one let me expand it a bit I see that some tabs appear over here let's go to one by one step by step as we said so the first thing we need to do is to read the two products that we want to correct stir the first one we see it is the first one up and in the product Explorer window on the read tab number two we select the second one here we see that we get an error it says that the entire image is outside of SRTM valid area and it is asking us to use another DM srtm can cover areas between minus 60 degrees and plus 60 60 degrees of latitude so since this is this area is quite in the north we need to select a different digital elevation model for that one we need to go to the back geocoding tab and select here an appropriate one I will select this one you can also select any other you want and once I click here I can go back again to the previous tab I was and you see there is no error anymore so the next step is to go and split the data to select only the subs what we want and from the southwest only the first we want remember that in both three tabs the information insert has to be the same otherwise the registration cannot be successful so we go here let's zoom in in the area in our case as we said we want the iw2 we want to select from polarizations only the VD and as I mentioned we want to select the sixth burst so I move this part until I see the six here and you see that the rest of them are being removed from the view here and I move the other one as well but if I select only one due to some snap restrictions the the whole process will fail so the minimum it is advisable to use for every processing is two bursts so I select six and seven I go to top sir split too you see here that we have an error that says same subs what is expected so I insert again the same parameters like before and once I do that if I go to the previous tab and then move to the next one you see that the error has been disappeared the next is to apply the orbit files as we said to get accurate satellite position and velocity information the orbit files the updated ones are usually available in a few days after the release of the product in this case since the products were using are quite old there is no problem the orbit files are available to be downloaded but in case you are using a product that is released only a couple of days ago from the today's day you need to select that do not fail if new orbit file is not found option because if we do not select that and there are no new orbit state vectors then the whole process will fail so just in case let's select it then further back to coding as you remember we selected before the proper digital elevation model and we keep the rest of the options as we want optionally you can select another resembling method and type if you want to but you can also keep the default ones and in the enhanced spectral diversity we do not change anything finally we go to the right option the first thing we do is we go here and we navigate to the proper path we want to store the outputs so I go to shared training landslide detection Iceland tutorial kit and I select the processing folder so here you see that you have the orb and the stock suffixes added at the end of the name of the product so we don't need to keep that long name so we can just remove all this part to keep the date of the first product of the master one which is 23rd June 2018 and we can also add the 20 1807 17 which is of the second product so that we know which are the two images we have kureta stirred the second thing I would advise you also to do is here in the name to go and put number two so that you remember in the output which is the interferometric wide swath you have used you have selected for the registration and then you click run so this process is not taking very long but I will not run it now I will close it and I will load the pre-processed stack I had created for you so here it is if I expand it now I will see that in the bands folder I have the information of intensity of Master of VV polarization and intensity of slave so this is what I wanted this is what I needed if I open them now and I go to window tile horizontally we can see that from the whole swath we only have the two bursts we have selected what can we see from here I know where the area of the landslide is just let me zoom into that okay I know that my landslide is over here as you can see there are some differences between the two images as I said we will use the information of face and the information of amplitude the amplitude is here depicted within the intensity bands so we see here some differences here we see an area that has quite high values and here this area doesn't have it why is this happening because here the first image on the left is the one before the landslide and the second is after so this is actually the first indicator that something has changed here and has changed quite a lot and this is the landslide this is not very clear to see what can we do to see it better let me close them we can go here and we can create a nerds bit composition so as we said we right-click here we go to open RGB image window and we want to select on the red channel to have the intensity of the master one which is the one before the event and on the green the intensity of the slave one which is the one after the landslide and we leave the blue LEDs so what will happen here all the pixels that have been only in the first image will be depicted as red and all the pixels that appear only in the second image will be appeared as green so let's click OK and here we see that we have different colors we have some red we have some green we have some yellow and also some black we will ignore the black parts but let's focus a bit on the other colors again let me zoom in in the air that I know that it is the landslide ok so we can see that there are some yellow pixels those yellow pixels what do they mean they actually mean that are the successfully registered pixels of both images but here we see an area that contains a lot of green pixels that means that this information didn't exist in the first one in the one before the event if you remember we put in the green channel the image after the landslide happened so this is actually where our landslide is located we can see it very clearly but apart from that this is still not the final step that we can use this is just to help us visualize and see the difference what we also want to do is to use the face and the coherence of the images in order to do that we need to create an individual gram and apply the next steps so let me close this one and for that we will build the graph so I go to tools rough builder let me explain it a bit the first thing that happens when we open this one is we see this white area up here that contains two operators the read and the write first of all when we create a graph just we go on the right we right-click on it and we select delete so that we can add all the rest operators on sequence and add the right operator again at the end otherwise we will have an error that the graph is incomplete on the lower part we have some tabs so whatever happens up here appears as a tab on the lower part what we will do is we will build step by step this graph and we will set the parameters at the same time on the lower part to start with as we said the read operator will contain the stock product we created before so it is here already selected the first operator we will add is the one to create interferogram so we just right click on the empty white space we go to add raider interferometric products interferogram and we select it you see that it appears here and we need to connect it with a red operator so we go on the right side of the rig and we drag and drop the arrow towards the interferogram operator then we go down to the tab that has appeared and we do not change anything of the parameters the next operator we need to add is as we mentioned we need to add with the first one so that will remove those black parts that are between the two bursts so that we actually put the two bursts together we go to add later Sentinel one tops tops rd burst we connected also with the interferogram and here we see that the polarization available is the V V because this is the only available we have in our product next operator we add is the topographic face removal we right-click again we go to add radar interferometric products top of face removal again we connect it with the previous one in the tab now if you remember we need to set the digital elevation model and as we said assert am is not available in this area so we go again and we select the one we selected before for the registration next step is to obtain square pixels as we said in the beginning so we need to multi look it because this is elongated and we need to make it square pixels so we go to add radar and we select multi look again we connect it and here we can see the source bands available we do not touch anything but here we come in the number of frames looks and we put eight so that automatically the mean square pixel becomes twenty-eight so the original sorry mods contain some inherent spec lines with multi looking will use the speckle appearance and we prove the image it's a credibility so that we obtain the square pixels the next step is the face filtering we need not filter the face we go to add radar in the parametric filtering goldstine face filtering so in this one we can leave the default parameters but in this case I will just enhance the filter a bit and put it to 128 so finally once we have applied all those steps we need to add the right operator we go to add input output right so that we have our output so we go again here again we go to the folder we want to store it processing I select so here you can see that the interferogram burst differential interferometry multi look and filtering options have been added to the name of the product before we run it we should go here to save we should navigate to the folder we want I go again in the same one the landslide detection Iceland tutorial kit in the processing and here I select a name for the graph to save it I save it as draft and I click Save and I want to save it so that if I want to apply those steps again for a different in for a different pair of images I do not need to create all the time the the graph I just go and I lock the previous one that I have already created and I run the processing okay I'm not running it right now I will just show you immediately the result so let me close that okay I expand it and I see now let's say the differences from the previous one in the previous one we had the intensity bands okay the intensity band of the master and of the slave and now we have an intensity band that contains both products from this processing the face band has been created and also the coherence band first of all let me open the coherence band okay if we go to color manipulation here we can see the values and also if you go to the pixel info and you click and you navigate with the cursor over the image you can see the values of the coherence here on the left its pixel has so if I go over a white area you see that I have quite high coherence like here 0.9 if I go to black one you can see that I have very low values of 0.2 so with black are the areas with low clearance now the image is multi locked which means it became in square pixels so what we need to do is we need to go again and zoom in right in the area of the landslide what do we expect to see in that area since the changes from the first to the second image are quite significant we expect to see some loss of coherence because the pixels wouldn't be able to read to correct stir properly so let me go back to the product Explorer window and let me move to the area over the landslide okay so here it is here this is the part that we see that we have this black area which corresponds to the area of the landslide now someone will say that we have many many parts that have low values of coherence and we have black areas and how can we know that there is a landslide happening there there are many parameters that we need to take into consideration for example any areas that contain water all the water bodies will always have very low coherence and they will also appear as black so many lakes will appear as black but as I mentioned in the beginning you had the two intensity bands the one from the first product before the event and the one afterwards and we could very clearly see that in the second product the values of the intensity over the area that the last light was were very high something that was missing totally from the previous one while all the other features they were existing before in both images so this is one way to see that here we have extreme loss of coherence if we go again - pixel info we will see very very low values so this is the area of the landslide okay we saw how the coherence can help us find these changes but how can the face help us in this case if I open the face band and I go to window tile horizontally and I zoom in again to the area of the landslide I can see that here there is this area that we have also loss of face I will show you a bit later more clearly this result but as a first indicator why do we see that happening this is happening because the displacement between the two images has been very very large so fringes couldn't be created as we can see here so we also lost the face due to this huge rapid change that occurred in the area I will close this one and this one to continue with the next step because as we said what we need to do is to apply the geocoding step the terrain correction so that we have the products do located so we go we select this one and we go to read our geometric terrain correction range doubler terrain correction and we select it here we see at the shortest one we have the last product created and on the name the TC the terrain correction has been added again the first thing we do is we go and we navigate to the folder we want to store our output so I go again here in the processing and I select it what we need to do now is we need to have two outputs because in QGIS if we want to add it and combine it with if we want to open the Bands in QGIS and combine it with other data with GPS measurements of GPS station on the area or any other kind of data we can have it in any coordinate system we want since UGS can read everything but we would prefer to have it in UTM but for using it in Google Earth we need to have it in geographic coordinates so first of all if we go to the processing parameters we can select in the source bands only the bands that we would like to be trained corrected so we select intensity phase and coherence remember we select the proper digital elevation model we leave the rest parameters as they are and in the map projection we see that automatically we have the terrain correction for geographic coordinates so if we click run the output this one can be used for visualization in Google Earth and then we click run remember you can go in the i/o car meter stab and here you can add for example the lot and loan for latitude and longitude so that you know which is the terrain correct product we have if we go now here back again and we go to the map projection and we'll select it we can go down and we can select the UTM wgs84 automatic one this way you see it automatically takes the zone 27 we go now to the i/o parameters and instead we write a UTM so that we know what is difference between the two terrain corrected products and we click run and this one we will use it for a QGIS just let me load them now both of them so here I have the first one in geographic coordinates and here I have the second one in UTM if I open them if I open for example the coherence of the terrain corrected on geographic coordinates this is how it will look like if I open the next one of the coherence again you will see that it has been also reprojected but it has some slight differences if I put window tile horizontally you can see that I cannot see them the one next to the other since they have different projection system if I open the coherence and the face once we see that they are quite long and we don't really need to have all these image the first thing we do when we split the product is to select only the verse we need so that we minimize the area that we are processing to save our storage space now you see that this is also quite long so what do we want to do we want to create subsets for both of them because we don't need all this information you can also leave it as it is but in case you want to create a subset you open the one that you want to use you go to master subset and here you go to the geo coordinates in the geo coordinates in this case we need to put for our area that contains the land flight we need to put sixty four point eight seven one for the north latitude bound for the west longitude bound we need to put 22 minus twenty-two point two four eight for the South land chip out we need to put sixty four seven zero eight and finally for the east one we go - twenty-one nine six two and we click OK before we click OK we see it here on the left this is the area of our land slide with the blue rectangle and then we click OK once we click OK the product appears here on the left with the name subject zero of the product now if I open the band and I open the coherence one I say that I have only the area that contains my landslide my landslide is in here so it's quite smaller than the previous one you have the subset here on the left but this is not physically stored if you want to store it you right click on it you go to save product and in this window that appears you select yes because it is because it needs to convert it to the deem the map format so that it will be able to be read by snap afterwards so we click yes we navigate to the folder we want again and once we are in the folder we want we give the file name we need and we click Save I will suggest save it as subset and then keep the rest named by default as it is and you click Save and this is it you can do also the same with a second product with the UTM one okay what do we want to do now once we have created also the subsets we want to export them how are we gonna export them we need to go here select the product and if I open the face one and the coherence so let me close all the previous views and keep only those two opened I will go for the first one for the face for the face band I will go here and I will go to the color manipulation and I will see here the minimum and maximum value as we see up here the minimum is -3 point almost 14 and the maximum 3.14 if you remember this is the values for a full color cycle and here we do not we do not see it like that so what we will do is we will go to the editor on the basic one and we will select this option that says range from data once we do it we see that it gets the proper range in the values and you see a slight difference also here in the area if you make any mistake we can just go here and you clicked reset the defaults and it goes back to the defaults once you have selected range from data if you want to save it like that you just go to the product you right-click again and you go save product and it's saving all the changes okay let me close them now and I will go here to this product that we have the geographic coordinates I will open the face and the coherence that contains the full product without the subset then I want to export them so I will just go first to the face I will right-click on it and I will click export view as Google Earth claims it once i do that i navigate again to the folder that i want it to be stored in the processing then i just give it a name in this case i named it as face i keep the rest by default and i add at the end the food so that i know if I want to explore the subset to give it a different name and I click Save I repeat the same for the coherence as well here right click export etc if I now want to export it as duties to put it in eqj yes I need to select it here on the Left I need to put file to export duties first I go to subset to ban subset and if I want to export all the coherence I disel echt the others I live on the coherence I click OK here again I navigate to the proper folder that I want and I give the name to the TIFF file that I wish and then I click export product and this is it before we move out from here I will show you another example so let me keep the face opened I will show you another interferometric pair I have created for you that contains two images both of them acquired before the landslide so the first one is again the same one 23rd of June 2018 and the second one was acquired on 5th of July two days before we had the landslide I will load it here to show you some differences they have it stored in the prevent ok so I have here the face opened of the one that we have processed and I open now also the second one if I put them in parallel and I go and i zoom in to the landslide area here it is this is it I will see some significant differences the first thing I noticed is that before the event which is on the right side we do not have any loss of coherence and we do not have any displacement action the second part that we notice is that also on the part on the western of it again no significant changes but what we see here on the left image is that on the area of the landslide we have a loss of face and also on the western part we have some extra fringes that have been created and this means something so what we know for sure for the area is that this landslide was always been moving slowly slowly and what we know is that after the landslide happened there were some shakes some quakes that have been triggered and we can see the formation on the whole area in general while a few days ago there was nothing happening in this area once the landslide happened we can see both the loss of face over the landslide area and the new fringes that have been created and we can measure the displacement so for both of them I have exported already in km set and we will go now to show you in Google Earth how it looks like and how we can measure the extent of the landslide but since we do not have installed the Google Earth on the virtual machines the first thing we need to do is to download these outputs to our local computer so how do we do it how do how can we interact between the virtual machine and our PC we press in our keyboard ctrl alt and shift and this window appears on the left so here on the devices I can go here I click I navigate to the path that I have saved the outputs processing and here what I only need to do is just a double-click on the product that I want to download and the download immediately starts here you have it in your download products if I go back again in the beginning you see here a white part this white part is also useful because this way you can copy and paste information from your virtual machine to your computer computer and vice versa so if I want to copy a text from a word file in my computer first I paste it here then I copy it from here and I paste it again on my virtual machine the same happens the other way as well if I click on ctrl alt and shift this window goes away again ok so now it's time to go to the Google Earth and open the products the coherence and the face I have already downloaded to show you in Google Earth we are now to the Google Earth and here we can see that here is where our lunch light is it is quite visible even today from Google Earth and here what is interesting is that we can also see that here the last light if you remember it blocked the river that was coming down it created a small lake and here you can see the new path the river found and down here it goes again and it connects to the initial bank it had so we can see that all these sediments are new sediments that have been created after the landslide ok now let's put our coherence image so if you see we put it we remove it and we bring it back you can see here with black that is the area of the last line so we have done is we have designed a polygon you go here to add polygons you can name it give it style and whatever you want we have created one around this area and if I open it the last slight extent here it is I can go to right-click properties and in the measurements I can see the area it is 2.1 square kilometers which is quite accurate to the initial estimation that was around 1.8 but remember that it's almost two years that have passed this landslide happened in July 2018 so the current image of course has some changes from the past but ok let me close it let me bring again the coherence of the landslide and there it is you can see it very clearly you can also go in search here in the history and find some past imagery but unfortunately there is an image available from 2017 before the landslide as you can see there is nothing here the river is running normally and if we go afterwards we have one which is after which is on May 2019 we do not have something closer to the date of the landslide event the same can happen also if we add the phase information here you can save the loss of face and also the values on the left and we can also compare it with the one that we mentioned before from the other interferometric pair that was from the two images that were both before the event so you see which are the differences just remember this is an example to show you how to detect rapid landslides and remember always that landslides is something that is happening usually quite rapidly we usually do not have information that it is gonna happen we can have some indications implying that probably something will happen we can measure this movement and also if you combine it with with ground station data you can have a better overview of an area and what is happening the impacts of landslides are quite huge luckily this happened in a remote area but still even today it is quite dangerous for people to go and visit this area around because it is quite unstable and it is useful to know where we had previous landslides so that we can protect ourselves from future ones because most likely there is gonna be another one in the same area since it seems that it's quite unstable okay let me go back quickly to the take-home message and before we continue to the Q&A session with the new Sentinel satellites the challenge in satellite remote sensing is no longer data availability but rather how to store and process all the information in addition to that it is necessary to explain how the data can be used and support users in their applications so the route service is here to solve those problems by providing virtual machines store and process the data and by offering a dedicated help desk support team of remote sensing experts to help you in your projects with Sentinel data so before moving to the Q&A session let me tell you again that you can repeat this exercise by your own if you want to practice adapted and evolved it to your own application and for that one go to ruse desk dottie you register in rows and apply for a virtual machine you will receive the training kit that contains the PDF of the step-by-step guide also do not miss our next roots webinar on another topic using Sentinel 5p data so this was all for my site I will now give you a couple of minutes to post your questions and I will be back to you for the Q&A session thank you very much again for attending your questions will be published in the Q&A document we will release and also the video will be uploaded in the YouTube channel till the next time we meet try to stay home and stay safe bye bye

Channel: RUS Copernicus Training

Views: 7,642

Rating: 5 out of 5

Keywords:

Id: Eb9nEVU6Y8E

Length: 64min 49sec (3889 seconds)

Published: Thu Apr 02 2020