 Welcome to NPTEL course on rural water resource management. This is week 12 lecture four. In this week, we have been looking at multiple databases that can aid our understanding on rural water resource management. And also get us data for future assessments of rural water resources. On this note, we will be looking at some more data from the water balance. In the past lectures, we looked at soil moisture, rainfall, storage, and other aspects of the water balance. In today's lecture, I will be showcasing the remote sensing power in collecting data for rural water management. So in the past lectures, we have seen that there are a lot of government data. However, the time and space in which it is available is limited. For example, your groundwater is taken in once in every four months. And it is not spread across the country, but also has some concentrations in some regions. So this may not be applicable to a rural entity where they may not have any groundwater well monitoring system, and or a river flow influence so that then discharge value. So in that kind of cases, it is best to augment, which means join observation data and remote sensing data. That is what IMD has done very clearly. The Indian Meteorological Department where they take observation rainfall and then satellite rainfall data and then they club it into one rainfall data. So now it is a mixture of observation and satellite data. Before that, it was only observation data or only satellite data and there were some issues on both sides. But now when they are merged, then the data product becomes more robust and also has better spatial and temporal resolution. Now, since this is a test, we now know that remote sensing use is pretty good for our system. What do we do? We need to have more access to this data. So for that, ISRO has made a data website called Boone and this requires a lot of GIS and remote sensing, which you could learn from other resources. But I'm just going to focus on the rural water databases available in Boone. So if you look at the progression of this two lecture series, you would have seen that I refer to papers, NGO reports, manuals, World Bank reports, etc. And then I also refer to the databases available in these systems. Then moving on, I also showed you how you could collect data from government and private entities. And now we have come to the remote sensing phase. So the Boone database has a lot of applications and it is made by the ISRO team in the space research organization under its multiple umbrellas and RSC, RSC, SC, SC, where they give near real-time data for rural water management. There are multiple data that come in and in the next lecture, which is the lecture five, we'll be going through to show you how a model can be created for which you need multiple data and that data is not readily available. And so in that occasion, remote sensing data helps. Some data are very expensive to get. You have to buy it and some data, the temporal or spatial resolution is limited. So how do you balance between these two? You add some open source satellite data, which can capture all the dynamics, but it needs hand-holding of observation data. So the observation data you get from your local resources, IMB, CWC, etc. And then you merge it with the remote sensing data to get a good data product out. And they have done like this for near real-time data for rural water management. What do you mean by near real-time is almost from the data of access, it is very close. So in the soil moisture, I would say it is near real-time because they were giving you data within four days. But as ET, they were giving you data in one week. So somewhere we could say that it is near real-time, all these databases. One example of the data they give is NDVI, which is a Normalized Difference Vegetation Index. It is kind of an indicator of vegetation. Is vegetation there or not? So you could use land use land cover, but land use land cover doesn't also talk about water and stress, etc. Whereas this indicator can talk about it. So how they estimate these indicators, why, etc., is beyond the scope. But what I am trying to say is there are data for rural water management in the remote sensing platform. For example, rainfall, slope, how slopey the land is, land use, land cover. If you have barren land versus soil data, all these you can take from Bhuvan. Then you have your NDVI indicators, which can be guiding you towards proper water management. They are big data archives, so you don't have to always download data and keep it on your system, make it heavy. You can always download whatever you need from these databases and then update. Keep updating your database with the data they have. One more private agency is Google Earth Engine, which I will also showcase in this lecture, because there are some data which you can get it from Bhuvan, but there are other data which are available from Google Earth Engine. So Google Earth Engine is similar to Bhuvan, but it brings in data from various satellite platforms. Not only ISRO, but it takes from NASA, from the US, ESA from the European Space Agency, Europe, Japan, etc. So all these data come together in a single platform and that is called Google Earth Engine. So let's look at this Bhuvan. I will first look at it. So Bhuvan has both light and normal connect. First, I'll try to do a local normal Bhuvan website because it has more GUI interface. If it doesn't work because of the internet, we will show you the light. And this is how the Google Earth Engine works. So you have satellite data on one side. It is mixed with algorithms. The algorithms can be from Google Earth Engine or your algorithm. For example, you want to mix data observation plus remote sensing data. All these you can give algorithms. You can type it as an algorithm in the Google Earth Engine and it will run for you and give you output. The output could be a merge product. The output could be a scaled product or a zoomed in product of a particular region. You can also have multiple layers together to give one value, which talks about stress indicators for water management. Let's take an example. So rural water management, stress can be indicated by your water levels in the dam. So water level could indicate stress. Another indicator could be your rainfall. If the rainfall is not there, you can say it is a stress. The other indicator is you will have a lot of water. But if the plants are taking more water because of the crop type, people like eucalyptus, then you will say it is another stress. So these stresses need not be just associated with one data, like, for example, rainfall. It can be associated with multiple data. So we have to be open to capture these multiple data to assess what is the rural water risk. And that helps us to be different than just looking at rainfall. For example, if you just look at rainfall and say, oh, the water is there, so it should be okay. But then when you go to the ground, you notice that the rainfall happens, but everything washes away because the slope is so high, the runoff happens. Or the plants are taking too much water than needed eucalyptus, so you don't have water in the lakes and ponds. So these kind of things can be put in a model after you understand, for which you need multiple databases. Just not having rainfall or soil moisture can help for this exercise because the problems are complex and multidisciplinary. So you also need multidisciplinary approach to capture the data and use the data in your rural water resource assessment work. It could be simple or very complex. Either way, you have to include multiple data. Just not one data is enough. For example, you'll have rainfall but not discharge. Maybe there's a dam which is actually closing the water for which you need the location of the dams. And then the slope of the land where the water is taken from rainfall into the dams, you think? Okay, so now we are going to look at this Bhuvan website. So I'm going to share the website of Bhuvan. Okay, so I'll just first start with the Google search. You can just type Bhuvan Israel. And then the first one is India Geo Platform of Israel, Indian Geo Platform of Israel and you can see .nrsc, which is the National Remote Sensing Center. So as I said, they take the data and then they put it in these kind of platforms so that anyone can access. A lot of people access this data, multiple people from India and abroad. So it is a very useful website for data collection. Let's do the open archive data. Okay, this is much faster than the WRIS website. Data is more but still I think because you saw they have some really good hardware and computers, they would have better access to this. So you could see that you can actually type in a location to zoom in to a particular location and see the data. Then you can collect the data and process it. Once you visualize the data processing and collecting the data is a different source like GIS and remote sensing platforms, which is not part of this course. I'm just going to show you the data availability, how you use it in your work is a different class. Okay, so right now I would just want to show you what data is available so that you could have idea that data is available. I would need to just learn the techniques. So I would first you can go to satellite the first tab on the open data archive. It will give you first satellite what data is available. Since we are starting fresh, I'll join with you guys to start fresh, then we don't know what satellite is what products. So let's not use that time. Let's go to team and products. So in team and products you can put the drop down menu, you can have land and terrain. So just click land and terrain. What is land and terrain give it gives you the elevation gradient how elevated the land is, where are the depressions etc for storage structures, farm ponds, those kind of things we can look at. Okay, so that is all given here with some snow albedo values which we're not using at this stage. Then the ocean and physical products you can click ocean and physical and then you can see heat wave temperature cyclones, those kind of things because if you're in the coastal region, these data would help or is needed very much for understanding the water coming in like in terms of rainfall, temperature, currents, all these things we can look at. The most important one we'll be looking for this class is the land and vegetation team and products. Okay, so before we go then I'll just finish this off also. So in program and projects you'll see like there is a national climate change study which has been done and a high resolution elevation map which has been done these are still updated. So you could use it if you need but I'll go back to team and products and then land and vegetation. In the land and vegetation you have products such as normalized different vegetation index, the NDVI which I talked about, or you have the global coverage and the local coverage and then the vegetation fraction. Let's first look at the vegetation fraction. When I click vegetation fraction, then the definition of the vegetation fraction comes up, which actually says what is the percentage or fraction of occupation of vegetation canopy, trees, crops, etc. To the area, so if you have 100 acres and 50 acres are with the crops, then it is 50% is the vegetation fraction. And this fraction is given as a color and the color is painted on this map, when you look at it so you can look at individual products the date in which you want, you can click it and then it says that every 15 days as a map. Okay, the last one we have this December, so I'll just click December, and then view. So you have the India color with anywhere from red to green and red implies that there is less vegetation crop, etc. Whereas your green would indicate a lot of forest cover along the area. So you see here and same like your WRS you can zoom in and zoom out those kind of things. The color is given here December 1 to 15 21. If you use a slider and jump the dates and go to a different date. What it tells you is there's some cloud cover here the white is not like empty spaces, it is not no data it is cover with cloud cloud event in December. And what you can see this these areas are 90 to 100% having the vegetation, which means that full 100 by 100% is given taken up by the vegetation, whereas in the red areas as I said these are the desert regions, etc. You see only 20 to 30% of the land being converted to which is actually that is in the most parts of India. You have that in because the car very the southern part is also having less water and high water stress. The vegetation fraction can be converted to an ET value because now you know the area. And in the previous class we discussed about evapotranspiration. So for a particular area you know the evaporation rate, you can simply multiply the area, which is a unit of meter squared, along with your evapotranspiration which is units of meters to get cubic meters of volume of water which is lost. These help widely in understanding the ground water potential, the surface water potential and remaining water resources in rural India. Because as I said there won't be any data collected in some of these regions look at it like all of India is painted and colored with different values, which means that the high spatial and temporal resolution data is helping. This data is actually merged with the previous central railroad board SWID data where they can have better indicators of why this ground water is depleting or rural water is depleting. Because now you have the maps which show how much fraction of the land is vegetated and what type you have to go down to the ground. This is one data product that we saw. The next data product is normalized difference vegetation courage for India, just for India. It doesn't populate yet because you haven't taken the date and as I said it's near real time. When you click the data you can have until December of 2021 which is kind of near real time. In your summer time you would see that during the growing period you'll see that you'll have higher data like just a week before two weeks before data you'll have. Right now because in March and February there's not much crops growing it is the winter season and the post monsoon season so you won't see much vegetation index showing in this maps. Then we have the normalized vegetation index NDVI which is a measure of the amount of vegetation on the land surface. And NDVI spatial composite images are developed to more easily distinguish between greenish green vegetation from the bare soils, bare soils is a brown soil which does not have that much of crops. Right now the data is coming up. You could see that 0 to 10 NDVI means almost barren and this part is where you have your deserts and other regions. Whereas most of this part the alluvial act of 1st etc is around 22 to 5.30 the 20.5 to 30 the NDVI range. Still it's less vegetation anything above 50 which is the green color would show healthy vegetation. So 50s are okay vegetation whereas 80s and 90s are high healthy vegetation. 90 to 100 is mostly the forest and that is why you can see forest here in the western guards some in the eastern guards and also in Assam regions the north eastern India a lot of beautiful forest are there. Right so all this can be analyzed from the NDVI data from the one website. So if you're working on government projects they would love to see these kind of images more than the Google Earth engines which are going to show. Okay, so how do we know what data what method they use for that they have a brochure you can click the brochure and it will open as a PDF file. It will tell you what satellite was used which payload etc was used to collect the data. We'll talk a lot about the technical parts of the data in the brochure. It's like buying this satellite data product but you know it is just free. Okay, the next part is your technical document which basically gives you the methodology on how they collected the data. What is the resolution, the image so this is how they calculated the NDVI. Okay, and then there's a volume fraction vegetation fraction. So all these have been given specifically for this and the data that they use would also be told what data they use to estimate these values properly. Then we have the entire world data but again as I said it takes a lot of cumbersome images to take from the world data to your location so we will not be doing that. We'll just show you what is India data. Okay, so we have seen the data range from almost 15 days and look at it 15 days December 1 to 15 and then the Jan part is not there because they haven't updated it. But normally during the growing period you will see this data available. So now you know the land use land cover type. You have the evapotranspiration. You can multiply the area to get the volume that is lost from the water balance equation. There are other products also that we could show one product I would like to show is the DEM project. So DEM are digital elevation models, how the elevation is occurring in the land. And again you don't need to put a date here because the elevation of the land is a stationary property. It may change because of land subsidence, earthquakes, movement plate movement, sectonic movements, but it doesn't change every day like your soil moisture ET or your MBVI. So all these are non-stationary products whereas here the location and the elevation for that location is a stationary product. You can see this. You can download exact range and the technical document will give you what these colors mean. For example, if you want to see Pune, the map can go to Pune and locate your elevation gradient. But you cannot directly access the values from here. It's just colored but the coloring is based on a particular value. So what you have to do is you will have to learn GIS to work on these data maps or MATLAB, SILAB, those kind of programs where they can take these GIS maps and then convert it into an Excel database. Okay, so we have seen the NDVI and other ranges in, okay, he's asking which Pune there's so many Pune. Okay, there you go. So Pune has been located with dots and that dot means, okay, then I'll just take this box of data to come. For example, I'm going to start. I'm just going to click this box data only I want. Okay, so you could take at least two data, but let me say. So you have selected all these blocks of tiles of data. You can download now then. You can download them for your Pune work because Pune, that's the center, but part of it also the district boundary also lies in the other region. So this is all with Booban. Now let me move on to Google Earth Engine. So all I have to say is Google Earth Engine. You can type it in Google. GE is the short form and you will be pulled up to this. If you have a Gmail account, you'll automatically be having an account on Google Earth Engine. Now we'll just look at data sets. So I'm just clicking data set which is on the top and then waiting for the data sets to upload. Okay, what it says is there's all a lot of data sets. You can go to view all data sets and then there's a search box which helps you to select the data. So it is running, but you can also type it here and then search. So here's all the data set. In the find all data sets, as I said, this is a big, big data set because it has a lot of satellite data from different parts of the country and the world. So you will have a lot of data on this. Let's say evapotranspiration. Okay, so there are a lot of products. There is a different couple of products, ALS, Chile product, isolation heat, actual evapotranspiration, decadal daily evapotranspiration, TERRA data. So there's a lot of data which does this ETE estimate different satellites. So the NRAC may be a different satellite. This can be a different satellite. This is my early warning system. Okay, I'm going to click the TERRA climate. So I will show you how to read the data out. Once you click the TERRA climate, the metadata or the data about the data comes in and then you can see the description and read how to get the data, look at the data. In the bands you can click and see what is each band related to what is the date, what is the data available. As in we want reference evapotranspiration and accumulation of precipitation. So you have ETE, the reference evapotranspiration is given as millimeters, same unit. The minimum is 0, maximum is 4548 millimeters. Then you have maximum temperatures. The good thing about this is when you download this data, you can readily apply it to your GIS platforms so that you can download the data and work on the water budget equation. So here you can see one satellite can give you wind speed, vapor pressure, maximum minimum temperatures, snow water equivalent, surface short radiation, how much radiation comes in. And then soil radiation, soil moisture, data, runoff precipitation, reference ETE, all these things. So all this data is shown. Let me also quickly give an example of the data. So I'm just going to come down and say open port editor. So before that it just went back to the search. I'm going to say evapotranspiration, maybe for the bandwidth and close some of the websites. And then here you have the bands as a soil evaporation, interception, vegetation transpiration. So evaporation and transpiration are kept separately. And then I can open it in port editor. So in editor the map will open and this one doesn't pull down your system. For example, it is a lot of data, but it doesn't actually work on the data on your computer. It works silently behind. And this is the Google Earth Engine, the gear box, which turns and collects data for you. Okay, so I'm going to run this. Again, the code just gives you running of the layers and then the layers are planted. What this code does, how it runs, et cetera. I would request you to read about these Google Earth engines. My point here is to show that there are multiple data available and different platforms available for total water management. The whole world is now populated. You would see that it just runs back and forth because it's a complete picture of all the countries. But for us, let's go to India. So the India platform is shown and you can download the data just for India or other countries depending on what you want to work on. So here it is. We have data for India from soil moisture, open body transpiration, all these things, which are very, very useful for farmers, but you need to localize the solution and give it. Giving an India size image may not get the administration from farmers. You need to give localized advisories and localized data. With this, I am closing today's session. Thank you.