 NPTEL course on remote sensing and GIS for rural development. This is week 7, lecture 4. In this week, we have been looking at certain aspects of extracting data from images and other resources like QGIS plugins, data repositories that can aid us in creating a database for rural development. This is important because unlike urban development and urban scenarios, data for rural development may not be easily accessible. In such a case, we do need to provide data and or mined or proxy data that can support our argument. Saying no data for all the periods has been okay, but now with remote sensing and GIS tools, one should not always say there is no data. There is always some data that we could manage to procure and apply it for rural development. So in the past three lectures in week 7, we have looked at exploring with Google Earth Pro and other resources and softwares that can aid us to collect data. QGIS plugin is one such data set that gives you a lot of data that is procured and provided by volunteers. So let us continue that aspect and in today's lecture, we will touch upon something very, very important for land management in rural areas. This is the basic hydrology diagram by USGS, which is widely used. Hydrology is a study of movement of water. So once you have rainfall, how it moves and distributes itself across the land is called hydrology. There is multiple components in hydrology like rainwater, surface water, soil water, groundwater, oceans, seas, plant water, evaporation, water vapor, etc. Of this, water goes through a cycle. You could see that how water starts in one part. Let us say take any part for study. Let us say we start in the oceans, due to sunlight and heat, water evaporates and then condenses into clouds, further condenses into precipitation. We have types of precipitation, rainfall and snow and then you have sublimation deposition, all these are small, small water hydrological components. Most of it comes down as runoff, either snow melt runoff or precipitation runoff and then you have stream flow, stream rivers, networks, etc. All these happen when water hits the ground and the ground elevation, elevation as in the height differences between the ground leads to water movement. If the land was like this and water falls, it will not move anywhere because there is a gradient, because there is a difference in height, water moves. Water moves through gravity and water moves from high energy to low energy. In other words, from high potential to low potential or high elevation to low elevation. You do not see water moving from ground to tank without a pump. Naturally, through gravity, tank to tap and ground it close. So, that is what this is about. If you have the brown part accurately assessed, you can know how the water moves and then you can take a lot of management practices into action. For that, we are in need of an elevation model. How is the land distributed in elevation? So, these are called digital elevation models. It is a model because like a globe is a model of the earth, a 3D map is a model of the earth. The elevation model, digital elevation model is the model of the elevation changes. In simple terms, it is described as a representation of the bare ground, which is the bare earth, the solid part of the earth. Topographic surface of the earth excluding trees, buildings and any other surface objects. So, it is basically a ground. You do not add the trees, buildings and then say elevation is high. The land elevation from the sea level, the sea is zero and how is the land elevated? Because if you say land is below sea level, the sea is here below the sea level, then water accumulates during floods and stuff. If you build stilts, if you build pillars and then you put your house up, that does not mean your land is elevated. Your land is still low. Only in some places, engineeringly, they elevate the land. For example, Chicago, they have elevated the stream bed so that the water flows in opposite directions. So, they have a stream bed which is engineered in Chicago, where in some parts of the monsoon, the water goes down and some parts of the monsoon, the water goes in the other way, because they can shift this elevation. It is very expensive and cannot be done everywhere. So, normally you allow nature to take full control and the force of nature is very, very powerful. So, the digital elevation model gives how the water moves because of elevation changes and it is the elevation or representation of the bare ground with nothing built on it. It is a 3D computer graphics data in which third dimension is the elevation. So, there are three dimensions. X, Y is the location, X and Y are two plane. So, you have this is a two plane and then you have X and Y and Z. Z is your elevation. It can be up or down based on the relevance to sea level. So, in a raster dataset, if you have a pixel, so what would be in the pixel? It is the elevation, just one value for the entire raster. So, these are some examples taken from NASA. The first example is you could see that the land is flat, but then suddenly there are valleys. Sudden jumps in the elevation like dips in the elevation. So, water would be going and moving down. Soil would erode down. And then the same thing you can see from a point of a ridge. So, how you have small, small hills and mountains taking shape and around it there is low elevation. The third image is very interesting. This is a satellite called Grace, which I use a lot in my research work. And you could see this is how the mass is distributed. It is not a smooth surface. The land is not smoothly elevated, like we assume. It is having ups and downs, dips and highs. And it is because of this, it is squished on some angles. It is not a perfectly sphere. You don't get an orange perfectly without engineering it. So, some part a little bit bulged out and then in a very, very irregular shape. However, the closest shape that we can assume for this is a spheroid. So, this spheroid is having mass at different, different locations, like differences of masses are there. And that mass is because a lot of mass is built on top of a layer, which is the elevation. And from there, water can flow. Water or whatever you want to model. We will see some models. So, digital elevation model is a representation of the bare ground of the earth, which gives you the elevation from as with respect to the zero level, which is the C level. There are multiple, multiple sources of data. And a lot of this has been still updated. It is keep on updating. Some database is given here for your reference. Just look at the names, the regions that have been expanded now, the resolutions have been increased. You have SRTMs, three arc seconds, one arc second. And then one arc is approximately 30 meters to along the equator. It differs where you are. Okay. And then you have GTOCO 30, 30 meter resolution, 30 arc second resolution. The globe is a 30 arc second resolution. CGAR gives you diverse resolutions, which means in some areas, like for example, the Indian regions, you get around 30 meter by 30 meter resolution. And then you have SRTM, viewfinder, panoramas, allows, Germany, and then radar, see the European Union, one arc second, one meter resolution in the Slovenian countries, Canada, Europe, Alaska all have different resolutions. Some of these are national missions where they understand the importance of a DEM. And so they have high cost instrumentations or data collecting pathways. And then they keep collecting data so that they can have a high resolution DEM. These DEM are very expensive or very low cost depending on the resolution. So if you're using some methodologies like satellites, it's less costly, whereas some others are very, very costly. One thing you should understand is the land surface does not change abruptly. So you won't take, you don't need to take this at a high spatial resolution like monthly, annually, etc. Once in five years, once in six years, 10 years, that's what the normal is. If there's an earthquake or a natural calamity, then people would, the scientists would quickly take the elevation changes. Otherwise, the elevation doesn't change. So let's look at some of the data types. So one is microwave remote sensing. They're using the SRTM mission, which is the Shuttler radar photographic mission from the NASA. You could see that it has a main antenna on board. And then there's non-board antenna where it sends data and picks up data from NASA. So these are kind of active sensors where micro radio waves are sent, bounce back from the earth and comes back. So once you know the velocity with which the waves are being sent, and then you know it bounces back at a particular level, then you know what is the elevation. So 30 meter resolution and 90 meters international. So most of the resolutions are very high. The Cartosat 2, which is an Indian mission, similar mission, has one meter resolution. These are the updated resolutions. And you would see that the resolutions are getting better and better every two, three years once, because a lot of investments have been done on this. This gives an example, the SRTM, the US NASA product for Harrow and Cast Hills in India. And you could see that the high elevations are given as pink and violet color, and then the lower elevations are green. So you see how a hill is looking at not a smooth surface, but a high gradient and then there is a slope. There's a slope and a low height. This is very, very important because now we know where soil erosion can happen, where people should not be building houses, rural evacuation areas, naturally high risk prone areas, we should know we should not be building houses, or intelligent farming where it can disturb the sustainability of the system. Then we have, so the satellites are pretty expensive, but then when you come to LIDAR, LIDAR resolution methods, you know that you have really, really high resolution imagery. So this is three centimeter resolution where you could see that a plane is carrying a light detection and ranging called LIDAR. It's an instrument which sends light pulses, and the same as LIDAR. It knows how fast the light particles move, and when it hits and comes back, you know the distance, the distance between the plane and the bare surface, and that is the elevation because the plane's height is known with respect to sea level, and when it bounces and backs, comes back, we know the time taken by the pulse to come down and up, and that gives you the resolution in elevation. So three centimeter resolution is taken by LIDAR, and these are done by NOAA, which is in a US agency, and you can see that how they have invested in taking these LIDAR images. These are now bare earths, so this is not a DEM, but a full picture of the elevation with trees and canopies and buildings. So this is also important because when you have this data and take out the ground data, now you know the height of the trees and the associated particles. These are very important to use in terms of finding new developments or forestation or forestation, how the plants, trees, etc. grow, and where pockets of low elevation develop, which could be leading to flood prone regions, or using a spawn city concept where you can divert the floods and store it here. Let us look at some associated uses before we jump into one example of a DEM. Let us take a DEM. DEMs are always rasters. Why? Because it is not a point data, it is a continuous data, and as I said, it is gridded. Each pixel has a value, and the value is an elevation. The elevation could be in feet, meters depending on the satellite's specification. So now you have a DEM and then you say that, let me take each pixel and then put a filter saying that only elevation above 500 meters should be allowed or it is passed through the filter. So now a DEM which has all elevations is passed through, and then you can have only the high elevation, so above 500 meters. And now you have valley and non-valley. Values are low elevation gradients with non-values of the high, and then you have forest and non-forest, which is the land use, land cover. So if you club these two, if you club the valley, the high elevation, low elevation, and then the land cover, then you have landscape sensitivity analysis, which is different coloring of the landscape, which gives you a valley forest, non-valley forest, non-valley, non-forest, valley and non-forest. So you have multiple permutation combinations, which are very important for biodiversity, plants, insects, animal livelihood, etc. So especially these are very important in rural regions where agroforestry is practiced, because agroforestry means you have to merge with the forest and conduct agricultural practices, not erasing the forest. So these type of activities are highly appreciated there. One more use we will see before we go into the Buon data set. So you can actually take a land use land cover, let us say land use land cover is taken from a Landsat and then DEM is taken from SRTM. And then what you can do is, if you know the elevation and the gradients, you can estimate the slope. So through GIS, you can say, okay, this is the raster, this is the elevation, give me the slopes. So the elevations are converted to slope. So two elevations are there. And if there is a gradient, if there is a fall or a high climbing gradient rising gradient, then you have a slope. The slope can be negative or positive. So these slope values are merged with the land use land cover to estimate mean slope per watershed area and where water accumulates or where land and soil can be protected. Because when water moves, it erodes soil, it erodes plant life, et cetera. And when it comes and stagnates or slowly flows, not fast flow flow of water, but when it flows slowly, that is where life form thrives, fish, insects thrive, and that contributes to agricultural productivity. If it flows fast, then the top soil is removed and you have erosion and the plants are disturbed, the agricultural productivity is disturbed. So the best idea is to slow down the water, capture most of the water and use it for agriculture rather than letting it pass through as a flood water. So there are multiple data sets, but let us first promote our own ISRO data set and we have this link which I have taken from Boogal. I will show you now how to access it and then read through the materials available for creating DEM dataset or understanding where the data is and how it looks like. So let me share the screen for the Boogal link that I have shared. First, I will open the Boogal main page so that you understand that we are going to, anyway, sometimes the links might change, but the main page looks like this, the Boogal main page. So in the Boogal main page, you will go to open data archive and then when you come to open data archive, you will have this India picture that we have looked at in previous examples and you have satellite sensor, theme products or programs and projects. In the satellite sensors, as I mentioned earlier, you can go to CatoSat 1 and you can see all DEM versions. The highest version is 3R, but let us read through it and then let us read through. So this is the metadata. If you click the metadata, it will open in another tab. So this is the satellite global IRS mission for large-scale mapping and terrain modeling applications. So it models the terrain for you and terrain modeling includes the land elevation, digital elevation model. When you say digital surface model, then it will also include all the buildings and trees and stuff. So you can have some terrain which is capturing both the built and trees and areas and stuff, whereas some only take the surface. So here you could see that the resolutions, orbit days, 97 minutes, what time it versus 126 days. So you have the repetition of your scenes and then how the image is being taken on day 1 and day 126. All these products are given and where to access it, et cetera, et cetera. You can click on the resolution and it says that the radiometric resolution header to the best spatial resolution offered from the IR satellite is at 5.8 meters in both panchromatic and multispectral mode. This is the CatoSat 1. Okay. So, but not all areas, you can get it for free for that resolution. Let's see how the resolution stays when it's being post-processed. Sometimes you will have to pay money also. For Indian users, payment may be made, demand draft, et cetera, but there are other ways where through research and education use, you can request Bhuvan ISRO team and through some agreements, they may be considering some data which is for studies and academics. Because a lot of people do a business with this, right? So there are a lot of spatial data companies which are making a lot of money and using free open source data. So to make sure that they can make money, but the cost will be incurred by the user. To make sure that everyone uses it similarly, there is a price. Okay. So this is by satellite and sensor. So we went to CatoSat, DEM version 3.1. Let's see what is the difference. So you can put a bounding box. Let's say I am looking at Maharashtra and I click a bounding box. So I select select. Okay. Please enter MNLN value. If you don't have an MNLN value, you can put the coordinates. We have seen this is not latitude, longitude, coordinates. If you don't have it, you can draw. Okay. So let's say I'm going to draw. I'm going to start the drawing by saying these boxes I want. That's it. And then I say stop. Then you go to tiles. Go to next. So all these data sets are here. So if you want the bounding box, you can take the data from here. But anyway, you already have the toposheet numbers and the bounding box details. You can just click view and the data will be populated. So I'm going to click all the views before I download just to see if they are good. If they are good and the data exists. So you can see slowly it is getting populated. It went too much into. So let me just zoom out a bit. And then we wait. So you see here it says loading. So let it load. And then while it's loading also, we can look at the metadata. So the metadata clearly tells you what is the resolution of the post-processed image. So this is the version three. It is a terrain model. Okay. Cartosat DEM stereotype number one, NRSC, National Remote Sensing Center. And all the data is for, as per the NRSC data decimation policy. It gives you the elevation. Third edition, et cetera, et cetera. The data and the coordinate system, which we had looked at in the previous lectures. So it is a WGS 1984. And the resolution is one arc second, approximately 30 meters. Whereas this original source, the source with which the data is taken is 2.5 meters. But then they have to do some calculations to get at this image. And then you see what is the primary mission about the mission. One arc second is the spatial resolution. And the unit in which the elevation is taken is meters, not feet, not inches, kilometers, meters. So now what you would do is if you like the data, you could download it. So just look at it. It is one arc second. And then this is the third resolution, third vision. So for example, you need the data, you can click on which data you need. And you say download, you need to log in. And then I've shown you how to log in. And then you can download the data. So basically, you can log in. Okay, once you click that, it will come here. And then some password, just to check the idea, you are not a robot. It asks you all these things. And then download happens. I've already have an account. So my download will start now. Okay. So the download list you can see here. So it shows what are the tiles that are being downloaded. And you can again click it. And then it starts downloading. Depending on your download, we start now, it has done. So when we download it, we can put it in our RSGS GIS. And hopefully we will use it in the lecture class where we show the hands-on session. So this is one data. So now I'm going to close this. I'm going to go back to the new selection and then here. So this is one thing. The other is by theme and products. So I'm going to go land and terrain. So I'm going to have the land elevation, right? So it's land and terrain, no biophysical, no ocean. I don't want vegetation in one terrain. So in this all DEM versions, you can select. So now let us select one version and see what is the difference between the versions. We'll select the same sheets. Or I'll also show you the bounding box, which I've already shown that you can take interactive drawing and then the bounding box you can put. But this is very neat that if you don't know the topo sheet, you can go here, press the I button and then press on a catalog data set. So basically this version does not have the entire database or maybe it's not loading yet. But let's see, let's click on this one. And it gives you a bounding box number. So it is tight ID is 24, 3D. And what it says is the topo sheet or the bounding box is there. So let's do D4, 3D, D4, 3D. So there it is. It automatically populates. I click view and the data comes up. So there the data has come up. We can zoom in to see how the data looks like. And the reservation is very coarse. You see that in the metadata version, you'll come down and see that it's still one hour second, but it's not clear. So that is where the third update, the third version is much, much better. This is the first version. So if you look at the version differences, all the other details are the same. But the version and the data stamp is much, much different in the third upgraded version. So algorithms would change. And then you have programs and projects. You can have one hour second. And then you can have all the products. And then just click what versions you want. And then let's say the same versions will select here. Start. Okay. And then I said, stop. Next. View, view, view, view. This is all. So you'll see all the versions, version one, version two, version three, version one, version two, version three. So you can actually, let's remove all these and just see the version one, version two, version three for this specific data. Okay. I'm just going to zoom in for this. And then you'll know the difference between version one, version two, or version three. Also, I can do one thing. So maybe the nearest scale, let's do this first. And then we'll see. So this is version one. You can see that the solution is there. But this white space doesn't look good, because it shows that maybe there's no data. Now you see more and more fingering of the streams and rivers. So what you see as this breaking up of the land mass is the elevation gradient and the lows and highs. Okay. So the lows and highs are normally taken up here. Okay. So I'm going to remove the behind one so that you can see properly. So coming back here, this is the version one, you'd only see these streams, but the smaller streams are not there. The ones you see like trees are streams. Okay. So if you look from the top, there is no trees or buildings, it's only land. And suddenly the black means it is going down. Okay. So going down means it's rivers or streams that are bifurcating. Then I'm going to add the second version. If you add the second version, you'll see some more streams coming up. And it's a little bit sharper than the first image. Now I'm going to add the third version and remove the second version. You could see some more slightly the fingerings of the minute, minute details getting more noticed. So you can see this more noticed. Right. So I'm going to put again, so as you from start, so only some part of the elevations are captured. Whereas you can see this part is also captured, but not sharp. And when you put this, some other parts are getting sharper. So this, so people would normally use the version three all have the same spatial resolution as one arc second, approximately 30 meters. But if you have higher, higher versions, then the clarity of the data is better. Okay. So now you could, you could go back to download list and download the data. If you want, this list contains all the files that you have downloaded 24 hours, you can download already we downloaded our file. So we don't need to download again. So this is how you would access the DEMs from women. There are multiple other sources. In the next class, we will go through the download, putting it in QGIS and mapping it into a map with different colors so that you can represent the total area in elevations, how the elevation is. Suppose the elevation is very solid and straight. It's easy to land manage it, right? You plough it and then you put crops, supply water, etc. But if it is like this, high sloping, then even if you plough too much here, all the soil will come down here. If you apply water, then water will definitely move down. So there is land management that you need to do. Leveling is different. Okay. And losing the top soil is different. So if you say, okay, I'll take all the soil and then make it straight, then you're taking the soil, which is the fertile part that the farm needs, the plant needs. So that is not correct in rural development scenarios. You need to manage it with the elevations. So if you come to some elevated countries like Malaysia and even Mumbai, you'll see some regions, they preserve the elevations. They don't cut it and then make it straight. Why? Because of the elevations, there are multiple hotspots in biodiversity, water, and living creatures. So they want to preserve it. These are very, very important for rural development. If you make it all straight, it is not easy to manage. It's easy to cut it straight and then put a tractor on top of it. But then there should be soil and water, right? So the land mass should not be altered too much. With this, I'll stop today's lecture and then we will see in the next lecture where we can have a hands-on experience on using this data for rural development. I will see you in the next class. Thank you.