 And welcome to tutorial 3 of the course Microwave Remote Sensing in Hydrology. So in this third tutorial, we shall learn about Snap Toolbox and how to perform pre-processing of the Sentinel C-Band Microwave Imagery. So in tutorial 2, it was L-Band, LO's Pulsar data and in this tutorial, it is Sentinel C-Band Microwave Imagery. So let us get started. So firstly let me give you an introduction about Snap which stands for Sentinel Application Platform. It is a tool for earth observation processing and analysis. And Snap has processing tools for all the 3 Sentinel toolboxes, Sentinel 1, 2 and 3 respectively. And the Snap installers are created with multi-platform operations like geocoding or ortho rectification or band arithmetic operations or projections. So all these can be performed using Snap. So what we will do is as before we will start with how to install Snap and then move ahead with the exercise. So let us see how to install Snap. So to download Snap, we have to go to the step.esa.int webpage and then you select where to click based upon your correct operating system. I mean Windows 64 bit, 32 bit, Mac as well as Unix. So as I mentioned earlier, this version of the tutorial is for Windows users. So I am going to click the Windows 64 bit version and typically the download it has begun it is almost 875 MB in size but then it will take just a couple of seconds. So while we wait for the download to get over, let me also tell you that it also has something known as the SMOS toolboxes. And these are worth exploring if you are into working with SMOS data. As I mentioned earlier, there are release notes of different projects which are available here like Snap, S1TBX, S2TBX, MOSBox, ProBavi, Toolbox and so on. And if you are interested, you can just click on these and explore a bit. So now the download is complete. So I click, of course, remove protection at this stage and as I click the .exe file, I follow the steps to complete the installation. Again it is a very quick process and it just takes a couple of seconds to complete. So I follow the wizard and there is license agreement. I click yes, then snap user data and basically following the steps to complete the installation process. Click next and select destination directory, yes and then yes, next, configure snap for use with Python, yes, I check it and you can see the installation has begun, okay. So let us wait while the setup installs on the system. As I mentioned before, it just takes a couple of seconds. So I want a desktop icon, yes. So now I can search and open the Snap Toolbox. So this is how my screen is going to look like. So here I am using version 8 of Snap. If there is any updates that is required for Snap, it shall be completed now, okay. So I am going to press yes because I want to check for updates now, okay. The following plugins will be updated, okay. You follow the steps, license agreement, yes, update and the plugins start downloading. So now that we have Snap installation completed and while the plugins are getting downloaded, let us find out how to download the synthetic aperture radar C-band imagery, okay. So for this, we are going to use Copernicus Open Access Hub. So previously this was known as Sentinel's Scientific Data Hub and it allows one to freely access the Sentinel-1, Sentinel-2, Sentinel-3 and Sentinel-5 P products. So let us see how to download the data now as before. So for downloading, we click on the open hub here, open hub and then insert the search criteria and on the portal we can pan into our desired area of interest. So as before I am going to zoom into Mumbai and in particular IIT Powai campus and for downloading data from the Copernicus Open Access Hub, we need to sign up and create an ESA user ID. So for me I have already created my ID. So I am just going to type in the login ID and password and login to the site. So now I want to click on the advanced options. Say I need data for Sentinel-1 and I can provide the sensing period. I can even select whether I need data from other sentinels such as Sentinel-2, 3, 5 P and so on. So let me go back to the advanced criteria and give the sensing period, okay. So which means I need data from this date and year to this date and year. So let me choose back date, say first Jan 2018 to say 2019 sometime in March, okay. So here sensing date refers to date when data was actually captured by the Sentinel and ingestion date refers to date when data was inserted in the Sentinel databases. So towards the left of your screen, you can see the data being displayed one by one and I can zoom out and get an idea of the thumbnail, okay. The footprint is shown here, okay. So when I click on download, the offline product retrieval gets initiated and you know once online you can be able to or you shall be able to download the same and let me tell you that the file size is you know a little more than 1 GB. So it takes some time to download, okay. So as I mentioned earlier, you click on download, it just takes some time to download. So while it downloads, let us try to understand few basics about Sentinel. Now the Sentinel-1 mission consists of a constellation of two satellites, Sentinel-1A and Sentinel-1B. Now both of which share the same orbital plane and each satellite carries an advanced radar instrument on board. So subsequent to each mission, we have Sentinel-2, 3, 4, 5 and 5B and again if you are more interested the mission guide for each shall provide high level description about the missions, objectives and the ground segment and data products and orbital characteristics, etc. So Sentinels are developed by the European Space Agency or ESA and Sentinel-4, 5 and 5P precursor where P is for precursor, they serve specific needs of Copernicus atmospheric monitoring services, okay. Having said that, there are a few terminologies that you are going to encounter as part of this tutorial. So let us try to quickly clarify them first. So here the selection of data can be on the following criteria say the satellite platform or the product type or polarization, sensor mode, relative orbit period and collection. Polarization is what we saw in tutorial 2. Now let us try to understand what is product type here. So as you see on your screen, SLC refers to single look complex wherein each pixel shall contain 2 information, one is amplitude and one is phase. And as level 1 product, we also have GRD that stands for Ground Range Detected. Now this is a multi-look system. So what is multi-looking, we have already covered in tutorial 2. So this GRD is a multi-look system wherein each coordinate is oriented along the ground range and flight direction. These are SLC and GRD, they are level 1 product. We also have something known as a level 2 ocean products or OCN, abbreviated as OCN. Now you know the data that we have downloaded for this particular tutorial, this is the name of the file, you can see it is pretty lengthy, is not it? So as mentioned in tutorial 2, by now I feel you have a fairly good idea about what is polarization, what is HH, VV, HV and VH. So in addition we have some abbreviations, for example, SH stands for single HH polarization and we also have dual polarizations and partial dual polarizations which are written here. Just a few abbreviations, but by now you do understand what is polarization and that 8 stands for horizontal and V stands for vertical. Now you know by now our file would have downloaded and before we start to work with the file I thought it will be a good idea to understand the pretty lengthy nomenclature of the sentinel files. So let us quickly try to understand what it means. So typically the first 3 letters that is S1A is a mission identifier, sentinel 1A mission identifier. The next term is the mode or the beam and then we have the product type which we just now discussed SLC, GRD, OCN, level 1 or level 2 and followed by of course the processing whether they are of level 1 or 2, in this case it is GRD level 1 and then we have something known as a product class where S stands for standard and then A is for annotation. And then we have the polarization and the starting date that is year, month, day, time in hour, minute, seconds and of course the stop date or time followed by something known as the absolute orbital number, absolute orbit number, mission data take ID, product unique ID and product file extension. Now that we have broken down the nomenclature of sentinel 1, let us try to understand a few basic terminologies. Let us try to understand more about SAR image distortions here. So assume that the platform carrying sensor is an aircraft. So the sensor is mounted on an aircraft. So what you see here is the flight direction and I have shown something known as the nadir, the nadir which means directly beneath the platform. I have shown the azimuth here, azimuth is nothing but the along track dimension parallel to the flight direction, azimuth. And then we have something known as a swath which is offset from nadir and nadir is nothing but directly beneath the platform. And of course we have something known as a range that is the across track dimension perpendicular to the flight direction, a simple line diagram which illustrates what is nadir, what is azimuth, what is range and what is swath. Just to make things clear, again we are going to assume that the sensor is mounted on the aircraft. So the platform is aircraft. So now the look angle which is denoted by two crosses. The look angle is the angle at which radar looks at the surface. And ground range distance is the true horizontal distance along the ground corresponding to each point measured in slant range. Now you will ask what is a slant range? So slant range distance is also shown here which is the radial line of sight between the radar and each target on the surface. So we have the ground range distance, we have the slant range distance, we have the look angle and of course we have the incidence angle which is nothing but the angle between the radar beam and the ground surface. I will give you some time to pause the screen and sink into the terminologies. So we have understood what is look angle, incidence angle, ground range distance, slant range distance, nadir, azimuth, range and swath. And remember in this case we are considering the platform to be an aircraft and that the sensor is mounted on the aircraft. With this background, let us try to understand about image distortions because in tutorial 3 we will cover image distortions. So assume the platform that is aircraft is flying over a mountain, say the Himalayas or any other mountain and assume that the target that is whatever you see in the blue shaded triangle is the target. Assume the slope of the target is very steep such that the top of the mountain may be closer to the instrument than the base of the mountain which means q is closer to the platform than p. In such a case in the final image that is formed the top of the mountain is going to be mapped nearer to the nadir than the base of the mountain, is not it? And then the mountain shall appear as though it is leaning over while it is not. So this effect is what is called as layover and it usually results from side looking geometry of imaging radar, side looking geometry of the imaging radar. Now there is another distortion known as foreshortening that you see towards the right of your screen. In this case assume a radar beam reaches base of a mountain before it reaches the top and foreshortening occurs because in a radar image the display is going to be something like this. So SAR image distortions, layover and foreshortening. Now typically this is what you see before correction I have given you a sample image of how it looks before correction and after correction by correcting foreshortening just as a sample example. Now we do have something known as a radar shadow you know where in shadows are regions which are completely black, completely dark and it represents you know region of silence as though there is a complete lack of received information or received echo. So in the diagram towards your left side there is no echo from the region ST. The region is also considered a region from where there is no measured signal no information no measured signal and ultimately those regions are going to look pitch dark. One more SAR image distortion that is shadow. Now let us come back to our tutorial exercise. So for this particular tutorial what we will do is we will try to use the Sentinel C band image in snap toolbox and we will try to pre-process it. In particular we will be covering 5 different steps firstly of course we have to open the file to view the number of bands in which the image has been captured and secondly we will make an orbit file we will perform radiometric calibration flatten the terrain and finally perform geometric correction. So we will see the steps one by one coming on to the step one. So by now our file download would have been completed. So let me open snap and I will go to file, open product and then click on the file. It usually unzips automatically and then it gets displayed here. So I am going to directly select the file it automatically unzips and then it is getting displayed in the product explorer. If I click on the plus I see the metadata, vector data, bands and so on. And if you click on the file you can actually see how the image is going to look like. One can also right click on the file and select RGB image window. As a default setting if you see red R band has been displayed through intensity VV, green has been displayed through intensity VH and blue has been displayed through intensity VV by intensity VH that is band ratio. So we can create an RGB image using snap and try to see how to visualize the image. It is a quick process and you get to see the image something like this in different colors because you have created an RGB image. Now let us go to the second step that is step two wherein we create an orbit file. This file is provided as part of the downloaded zipped image file and in case you do not find it we can also create it using a single step that is you go to radar, you apply orbit file and in the source menu the original downloaded file is taken as input and in the target product menu either given name or as a default measure snap adds underscore or be extension to the original file name. You can see it has already added underscore or be extension. So I can run and then as before it just takes a couple of seconds and once the orbit file has been created it is going to be again displayed in the product explorer towards your left side. Now I see the image being displayed. Let us go to the third step that is performing radiometric calibration. See you should first know what you are doing before you apply these steps is not it. So for that purpose let me give you a small clarity. So radiometric corrections are necessary mainly because of variation in scene illumination and viewing geometry or atmospheric conditions or any kind of sensor noise and response. And to perform radiometric calibration in input output parameters menu and source product option we can choose the newly created orbit file. Newly created orbit file is chosen and as before in the target product underscore cal extension is provided underscore cal. I select the newly created product and I name it appropriately and I denote where to save it. Now in the processing parameters option you can either select the polarization type as vv or vh. I am going to select vv and then you can uncheck the output sigma naught band option and then click run. As before it just takes a couple of seconds to complete the process and once the process is completed the output image is going to be displayed towards the left side in the product explorer. Let us go to step 4 that is to flatten the terrain. Basically the step is performed to correct the radiometric distortions that are caused by different local incidence angles, okay. So what we do is we can go to radar, radiometric, radiometric terrain flattening, processing parameters one could see the band and polarization chosen during the making of calibration file. So here I can create an appropriate digital elevation model, okay and then uncheck the re-grid method option and then click run. So I have the source product written, the target product written and then I can choose an appropriate digital elevation model. So this process is a bit lengthy and it is going to take a little more than 20 minutes for me it took almost 21 minutes, little over than that. So once this process is completed let us go to step 5 which is geometric corrections, okay. So geometric corrections are performed mainly so that the geometric representation of the imagery remains as close as possible to the real world and this is performed to correct any sort of distortions which may arise due to say the motion of the platform let it be aircraft or satellite due to the altitude, attitude and velocity or due to terrain, effects terrain release due to curvature, rotation of earth and so on. Now typically when these variations are predictable or we call it when they are systematic it can be attended to properly whereas if we have random errors it of course cannot be modelled easily or corrected for. So moving on to go to step 5 that is to perform geometric correction what we do is we can write click to highlight the newly created terrain product in the product explorer and then I can go to radar, geometric terrain correction, range Doppler terrain correction and in the processing parameters option I can select UTM WGS 84 as the correct projection type. So here you can see pixel spacing is given, dem resampling method is given, the source bands are given and a number of projections are given, map projections source band is displayed I will give it a few seconds to sink in. So as before we go to map projection and then we select UTM WGS 84 as the correct projection type. So once you click on it the zone is going to be specified, you see UTM zone 43 is specified and then we click run. I have completed this process once and that is why the pop-up shows up again and again. So the writing target product this takes hardly a few seconds as before. So this is the final image that you get that is a single layer gamma VV file and this is the geometrically corrected file that you see here. So to summarize in this particular tutorial we learnt about how to perform preprocessing in the C band sentinel image and we learnt how to install snap, we learnt how to download the image and we also examined a few basic terminologies which is also covered in the lectures that are used in SAR image processing. I hope that you found this tutorial useful and we shall be meeting again for subsequent parts of the course. Thank you so much.