 to tutorial 11 of this course wherein we shall learn about interferometry what it is and interferometric processing using synthetic aperture radar that is SAR products in SNAP. So firstly what is interferometry? See in previous tutorials we have seen that synthetic aperture radar image consists of a real part and an imaginary part okay it is a complex number and the phase difference between two complex radar measurements captured from two different positions of the sensor is highly correlated to the terrain to the topography okay. So assume that you are using a single look complex image that is SLC image which consists of of course amplitude and phase information here by amplitude I am referring to strength of back scattered signal and phase is the fraction of a complete sine wave, fraction of a complete sine wave cycle. So among other factors the phase information is going to depend upon the distance between the sensor, antenna and the ground object okay and we can use the phase information from two core registered images to study more about terrains and this concept is used in digital elevation models. Let me try to reiterate we use the phase information generated from two core registered images to study more about terrain topography and this concept this is used to generate digital elevation models. So let us try to understand how to perform interferometry using radar sat images for this particular tutorial. Before that let me give you a brief overview of the steps we are going to follow through the series of previous tutorials I am hoping that you understand now how to access the data from a particular source okay. So for this particular tutorial I am going to use pre downloaded data sets from radar sat to okay SLC is single look complex image. So we will start our analysis with core registration okay I will explain the meaning of each steps as we are performing it in snap. So first step is we will core register two images SLC1 I am going to call it SLC1 and SLC2 and then we move on to generate the interferogram that is step 2 okay and then we use the filter we do phase unwrapping and finally we convert phase to displacement this is where we stop and this slide gives you the overall series of processes that we will follow in snap for this particular tutorial that deals with radar interferometry using snap okay. So in one of the earlier tutorials we have installed a snap. So now let us try to open snap which stands for Sentinel application platform okay because I am going to import the SLC images into snap. So I am directly going to the product explorer okay right now it is empty. So I am going to open the product going to locate where the files are stored as I mentioned earlier for this particular tutorial sample pre downloaded data sets from radar sat are going to be used okay. So let me open them I am going to navigate to the product.xml file okay product.xml file as shown here okay it opens and gets displayed in the product explorer. So in the meantime let me try to unzip the other product as well okay. These are the two sample data sets that I have extracted and now I want to open it in the snap. So as before I am going to use the snap open product dialog box and navigate to the unzipped data set directory and from there click on the product.xml file as shown here product.xml file. Now you can see both the data sets are imported it is visible in the product explorer okay and as mentioned before let us open and check the bands. It is going to consist of an imaginary component and a real component okay. I can double click on it to view how the image looks like. So this is how the image is going to look like. I can do the same for both the SLC images. I can double click and then view how the components look like. So we have imported two data sets into snap. What are these data sets? Radarsat 2 images that I have pre downloaded for this particular tutorial okay. Now let us try to move to the step that is core registration. In core registration of two SAR images one must be assigned to a master that is reference image and the second the next should be assigned to the slave image okay. So we have both the images in the product explorer. Let us try to do core registration by going to radar, core registration and core registration okay. A dialog box will open up. You can simply drag and drop both the SLC products into the product set reader that you see in the screen okay. You can simply drag and drop and in case the details are not getting displayed in the table, what you do is you click on the refresh button that will make the data appear on the product set reader okay. So now the data sets are appearing. Now you see the data sets displayed. There are other options like create stack and cross correlation and warp and write okay. So for create stack, cross correlation and warp I am not going to disturb any details I am going to keep the same parameters just to show you how the screen looks like when you are performing this tutorial okay. War. I can click on write, provide name of the core registered output file and also provide the name of the output directory where it is going to be saved. So I am going to specify a directory where the output is going to be saved and then I am going to click on run okay. So after core registration it hardly takes a few seconds to complete the process depending upon of course the image size that you have used as input. So after core registration the SLC data shall be loaded in SNAP which can be seen in the product explorer. Now often it so happens that processing a complete set of scene that is a large image it can be highly computationally demanding. So we can subset the data set if required. So in this case here you can see I have completed core registration and the output is visible in the product explorer and to make the processing faster I am going to subset the image okay. I am going to select a small region from this image how okay. So first let me view the images, MST is master, SLV is slave. You just need to double click on the image and it gets displayed. Now to subset I can right click and go to spatial subset from view okay and you can see a box that you can adjust over the region of interest from the specify product subset dialogue box. So here you see the box okay. I can move this box or even reduce its size. I can place it over the area that is my region of interest and the pixel coordinates are also displayed you can see on the screen. I am going to spatially subset the data that is what I am doing. Now there are other options like band subset, tie point grid subset and metadata subset and not going to disturb that now. Once I select the region of interest okay I can press on okay and then the subset image subset image is going to be displayed in the product explorer not going to disturb any of these and you can see the subset of the image is displayed in the product explorer okay. As before I can view the bands and see how the subset looks like okay. In order to display the RGB of the subset and core registered SLC we can right click on the output from the subset and then click on open RGB image window. So this is what will be opened. So here you can select the intensity bands for both the image it may be automatically selected in the select RGB image channel dialogue box. If you want you can select the master image in red and slave image in green and then press okay. You can see an RGB view in front of you. You will notice many colors in red or maybe green at places. So what we have done is core registration has aligned both images and the resulting RGB view should mostly look yellow yeah which is what is seen in the screen okay. So we have downloaded the radars at two images, two images one as master one as slave and then we have core registered them and then now we have created a spatial subset and then displayed the image in the form of RGB. Now as the next step let us try to generate interferogram okay. For this we can go directly to the radar tab okay and go to interferometric and then go to products and then we have interferogram formation. Now in this interferogram formation dialogue box I can select the core registered image which was subset as source product and in the processing parameters I can set the parameters as shown below. You can see the processing parameters and once you set it you can click on run. The target product is going to be created and it is going to be included in the product explorer, the output from the interferogram formation. You can go to bands, we can double click the coherence band to check the coherence of the SLC pairs okay I can go to the phase image to see what has been generated okay. The coherence band and the phase image right. So just to reiterate from the radar set to images we have completed core registration. We have created a subset of the image and we have tried to display it as an RGB image and performed interferogram formation. Next is topographic phase removal. Now topographic phase removal is conducted usually using a DEM, a known DEM where DEM stands for digital elevation model. Now this is to emphasize the phase signature related to deformation and in this process what happens is that a simulated interferogram based on DEM that is reference DEM gets subtracted from the processed interferogram okay. So there is going to be a simulated interferogram from a reference DEM that is going to be subtracted from a processed interferogram. So let us see how this is conducted. We go to radar okay and then interferometric products, topographic phase removal, a dialogue box opens up. Now in processing parameters you can select DEM as SRTM 3 seconds, press on run. As you can see it has hardly taken a few seconds and the output as before is displayed in the product explorer. We can observe the output that has been created with a double click. So now let us go to the next step that is applying phase filter. Now interferometric phase can often be corrupted, affected by noise. This can be due to temporal decorrelation, geometric decorrelation, volume scattering and so on and due to loss of coherence the interference pattern may be lost. So hence the signal to noise ratio needs to be increased by filtering the phase and hence this step okay. This is mainly to be able to properly unwrap the phase okay. So we can go to radar, interferometric filtering and I am going to click on Goldstein phase filtering. The dialogue box is going to open up, the source product is specified. I can specify name of the target product and change the processing parameters if required and then click on run so that the output gets displayed as before in the product explorer. Let us wait for it to be displayed and as before I want to view the results how it looks like. So now let us try to understand how to export it to snafu. Phase unwrapping is the next step that we are going to accomplish with snafu. So in an interferogram the phase, the interferometric phase is going to be ambiguous and only known within 2 pi. So in order to relate the interferometric phase to the topographic height, the phase information must be unwrapped okay unwrapped, phase unwrapping. So for that we need to export the filtered and flattened interferogram to snafu. So more details about snafu you can get by visiting the ESA website as shown here. So first what we do is we try to install snafu using SIGWIN and then we try to import the unwrapped phase from snafu. So snafu is actually a statistical cost network flow algorithm for phase unwrapping. It was developed at Stanford University and to use this as an executable file say if you are a Windows user it has to be compiled first so that all the required drivers are installed correctly and in addition you need to have SIGWIN installed as well. Now coming to SIGWIN, this is a functionality very similar to the Linux distribution on Windows. We can visit the website here as shown and then access the 32 or 64 bit version according to your system configuration okay. So I am going to go with setup x86 underscore 64.exe file, I can download the exe file and then I can click on it so that it executes the downloaded installer. So now it is getting installed, I can choose install from internet and then I can specify the root directory where it is going to be installed. I can define the local package directory, use system proxy settings or direct connection as applicable and I can choose a specific download site. Select any one server it does not matter and then the display would show you the progress of installation. Now here one can select search for make okay, select and then press on next and then search for GCC core again select press on next and then wait for the installation to be completed. So you can see make I have selected that and then GCC core I have selected that as well let us wait for the installation to be completed, does take some time for the installation to be completed okay, comes to install snafu. Now if you are interested to know more details about snafu, I would urge you to visit this ESA website where snafu plugin for snaf the details are made available. As before I am going to directly click on 64 bit windows so that the installation is going to commence. So what I am going to do is in the first part of this section you have already seen where to download the snafu file from, it is from the ESA website, European Space Agency website. So as part of the installation I am going to copy it inside the C drive where Seguin 64 folder, TMP folder is present. I am going to copy paste it here okay and then I can change the directory to the snafu source code. What I want to do is I want to follow a series of steps to prepare the compiler. I am going to perform a series of steps to prepare the compiler. I have right clicked and opened it in notepad and then the command to call snafu is present here. I am going to type that command here and this is nothing but the interface of Seguin okay. Let me type the command to call snafu. So now you can see unwrapping has started okay. Now let us wait for all the complete pixels to be unwrapped, take some time. So what is happening? We are trying to perform phase unwrapping, we downloaded snafu and then specified the directory and then we are opening the Seguin interface, typing the command to call snafu and then now unwrapping has begun. So once unwrapping completes itself, the data that is the unwrapped data is going to be written in the same working directory as you can see, UNW phase. So these are the unwrapped data files, it is going to be written in the same working directory. If you remember it was Seguin 64, user local bin in my case okay alright. So let us now try to read the unwrapped phase data into snap okay. So as you can see I have opened both in snap okay, it is visible in the product explorer, the unwrapped data. Now phase unwrapping as such this step is to restore the correct multiple of 2pi to each point of the interferometric phase image okay. So I can click on snafu import, so again to reiterate you know when we are adding or subtracting multiples of 2pi, we are suppressing the phase jumps and in phase unwrapping we are determining the order of each fringe and restoring the continuity of the phase map okay. So to view the unwrapped image I had them in snap and then I go to radar, interferometric, unwrapping and snafu import and then this dialog box opens up okay. Now coming on to phase 2 displacement, that is now we are trying to generate displacement maps. For generating displacement from phase we can go to as shown in the screen radar, interferometric products and then phase 2 displacement a dialog box is going to open up, let us wait here. So you can fill in the details in the phase 2 displacement dialog box. The source product is already available, the target product name you can add the directory in which it is going to be saved that also you can specify. There are processing parameters which you can change if required and the moment you press on run it is going to give you the output, you can visualize it by going to band and then clicking on displacement. If interested you can also use the color manipulation tool okay. So in this tutorial we began with sample data from radar set 2, let me refresh your memory. So we followed the steps for interferogram generation, we imported the images to SNAP, we performed core registration, we subset the data followed by interferogram formation and then we performed topographic phase removal and applied the phase filter and then we exported to snafu. Phase unwrapping was conducted with snafu, so for that we have installed a sigwin and snafu, we have prepared the compiler and then finally we have carried out phase unwrapping and then we read this unwrapped phase into SNAP and finally we performed the phase 2 displacement process to view the output image. So let me hope that you enjoyed this section of the tutorial and I shall meet you in the next class. Thank you.