 Welcome to NPTEL course on groundwater hydrology and management. This is week 12, lecture three. We are nearing the end of the entire NPTEL groundwater lecture wherein we have been focusing on different data sets to use in your groundwater studies, models, and research. Please understand that most of the data that I share has its own spatial and temporal limitations for which there is a need to use remote sensing data. So I do not want to leave that rock unturned. So what we are doing today is we will be focusing on the available remote sensing data, which is free that you could use and understand the groundwater behavior. There are limitations on this side also. As I said, you have to take your observed data and remote sensing data and try to mix it to a good new product. To be focusing on something new, let me tell you how IMB works. So IMB has the Indian Meteorological Department has rainfall gauges to monitor rainfall. However, they felt that the spatial and temporal resolution may be limited and could be improved by augmenting with remote sensing data. So they joined hands with NRSE and other ISRO data products and NASA products and made a rainfall grid rather than a point data. They made a grid. How did they do it by merging the observation data and your satellite data to get at good products? Like this, there are multiple other users of satellite data and it is good to see how that works for the groundwater system. So today we'll be looking at the most important groundwater instrument in the satellite that can monitor gravity. So gravity can be monitored and from gravity you can estimate the groundwater hydrology, at least the change. So this mission, very special mission is called GRACE which stands for Gravity Recovery and Climate Experiment. It is the only remote sensing satellite that can measure both terrestrial water and groundwater change. Terrestrial is any water on top of the land and including that and below the land under the ground also it can monitor and measure. It was not made for that purpose but the discovery is such that when they saw the gravity changes and they understood that gravity cannot change within months without a big phenomena happening. For example, if you have earthquake then the mass is shifted to one place. From here the plates move and maybe they get overlapped and so the mass is changed. When mass changes or the weight, the mass changes the gravity pull is more. We know this from physics that as the mass changes the gravity pull on an object changes. So knowing this, these smart NASA and German space scientists what they discovered is they made this GRACE satellite which just monitors how the gravity changes across the globe and those gravity changes along with other data can give you the groundwater change. We will see how through some examples and we'll also see the website and how it works. But please understand this is not a standalone neither is the observation data it's good to merge these two data together and make good products out of it. What are the benefits? The benefits are it produces higher spatial and temporal resolution because your point data is only a well at a particular location and it doesn't tell you the full picture of how the entire aquifer system changes and also the depth because as I said, you have drilling wells and you have shallow wells. Most of the monitoring is in the shallow wells not the drilling wells. However, farmers are also using water from the drilling wells. So that is where GRACE data plays a very vital role. Then you have higher spatial and temporal resolution. The groundwater data is monitored once every quarter which is once in four months. However, the satellite data GRACE is monitored every month and they're also thinking about bimonthly data. So this high resolution would eventually be very, very helpful for monitoring groundwater change. So we have global scale now. Globally, we can monitor because some of the aquifers are too big like example, the Ganges flame aquifers and then monthly you need to see the changes because suddenly there is a big pumping happening and we may miss it if we don't look at it at once. However, this data since it's a very specific specialized satellite data it needs some post-processing and cleaning to arrive at a tangible data for groundwater. So that is one limitations. There are some calculations that you would need to do based on the water balance calculations to arrive at groundwater. What you get is a gravity change. How do you change the gravity anomaly? Anomaly is the change per unit time. How do you change the anomaly to a monthly reading or a groundwater reading is the question. That we will show in a very simple water balance equation. Accuracy is good per millimeter also we'll get accuracy but it needs tremendous amount of ground grouping. So this is the catch. I'm saying that you can use this where you don't have data. So but now on the other hand I'm saying you do have data that can validate the grace mission. So how do you do this? Basically you should be able to merge both the data sets where the locations are known and the groundwater observations are available. Then we could actually merge these products and make predictions for groundwater anomaly groundwater storage change at across other regions. Basically what you get here is a net groundwater thickness or volume change between months. And so if you have a long time period and you know that the post monsoon season the groundwater change is positive and then from the post monsoon the groundwater is going to be depleted either by base flow or by pumping. And then it comes to the groundwater least level or point before the summer or during summer. We call it pre monsoon season. We have post monsoon and pre monsoon. So post monsoon will have the highest groundwater level whereas the pre monsoon would have the least. Let's take Maharashtra case. You have a good groundwater levels post monsoon for example, July, August because monsoon starts in June and then you have very low levels in March, April, May. May is really low. So that is where you need water to irrigate for crops and people use tremendous amount of water but the evaporation is high, transpiration is high. Okay, so therefore we do need good observation data for groundwater. So the good part here is it is not going to compete with observation data. It's going to liase or you get united unison with observation data to produce good groundwater storage change values. Okay, I will show you the mission, how it works. So satellites are normally called missions. So this grace mission theory is given by NASA. So what you see here is the grace mission on and how it works. So this grace mission is launched by a combined effort of NASA and the German JPL and JGFZ. So you would see all these different partners come together and work on a common satellite, which is good for the planet. It is one of a very unique system but let's talk about the satellite and how it measures. Okay, so you have a mass. Okay, the mass in the first stage, it is two satellites. This is the only few missions where the satellite is launched in pairs. You have two satellites and as the name suggests, one is called Tom and then the other is called Jerry. So basically Tom runs and Jerry catches. Tom is the cat. Sorry, Tom is the mouse. Is it Tom? Yes. So you have Tom and Jerry story, right? Tom is the cat and Jerry is the mouse. So Jerry runs and Tom catches. Okay, so here Jerry is running first. Okay, and when the satellite goes on top of the earth, it gets pulled by the mass. So when the satellite sees a bigger mass, then the satellite moves faster. Okay, so this satellite is not recording anything. Only it is moving around the planet and it speeds up and slows down depending on the mass. So the next satellite, which is your Tom is catching with Jerry and whenever the first satellite goes fast, it records how fast it goes. Then when it slows down, it records how slow it goes. This changing accelerometer because it accelerates and then de-accelerates. So basically it is not pulled by the body but mostly pulled by the attraction towards the first satellite. Okay, so now what happens is you have the satellite going in, the first satellite going around the earth and when it goes faster, the second satellite goes faster and then when it goes slow, this goes slow. So this change in speed is measured and the speed is then converted to a gravity change. Because the gravity changed, the speed changes. So this is how the gray system works in a very simplistic manner. There's much more physics in it but I'm just explaining the simplistic manner. So the first forward speed is there and then the gravitational pull is acting on the first satellite. So it goes faster. Then when it goes faster, the second satellite doesn't know that this guy is going too fast. So it catches up, okay? It catches up by the attraction towards here and also the gravity mass and then goes in. So this change in speed is recorded and it is taken as a change in gravity. That's it. There's no other force that can influence the speed of a satellite. It is purely gravity. So here, why doesn't other satellites do it? Because they don't carry a very sensitive accelerometer. Here, this satellite has no fancy things. It is only the only fancy because there's no camera, nothing. The only fancy thing is this accelerometer and it is very, very sensitive which means a very highly accurate accelerometer and that can be changed to a gravity pull. Now, you have the theory. So let's see what would change between months. So this is the first month. The mass is there, the satellites go past and it goes past this one, two, three, four is done. The satellites are gone. Then the next month, again, the satellite comes. So every month, once the satellite comes to the same location. Now, if the mass has changed, why would the mass change here? Because you have snow on the top, maybe the snow melts down. And when the snow melts down, there is less mass on the earth on that point. So if the pull is different, okay? The pull is not the same. So there's less mass, less full and less speed. So now this difference between months is accounted for the snow melt. So think about a region where there is no surface water and there is no change in mass, no snow, no nothing that melts between months. What changes is the water under the ground, okay? So let me draw it just as an example. So we have this as the earth, okay? And this has some water bodies like a lake and then there is trees, et cetera, et cetera, okay? So when the satellite goes on this part and the lake value changes between months, then it is influenced by the lake also, okay? Because the gravity, everything on top of the earth and under the earth influence the gravity on this satellite. What about in a place like this, where in a pixel like this, there is no change in surface water because there's no surface water. All it is is soil moisture, okay? And groundwater. And if there is a pump which is pumping groundwater into the aquifer, from the aquifer into the agricultural fields, then a mass is lost. You are losing groundwater mass. When you pull water out, the weight of water is lost and this can be accurately estimated by grace. So now you think that water which you take out is a level which is depleted and that level is also equivalent to a mass. It has a mass, right? 1000 liters is equal to, you know, the calculations for a weight depending on the density of water, okay? So you have this as one aspect to measure and monitor the groundwater change using the change in mass of groundwater which has been extracted and put outside. So now the extracted water can be covered with evapotranspiration or taken away to cities for drinking water, et cetera. But now it has been caught as a mass change. So when the satellite flow flies on that particular location and there is no change in lakes, there's no change in trees because trees are standing. And also if there's a building, if the building doesn't change, suddenly within a month, two, four, five flows extra added, then the pure change is due to soil moisture and groundwater change. Now we should just remove the soil moisture. I'll come to that part when I discuss the equation. Okay, so this is how the satellite looks at the planet. It does not have an optical sensor where like other data we saw, it is taking hyperspectral images, different bands coming in for assessing the properties of land, properties of leaves, color, et cetera. Here it is purely the gravity. So now you see how the grace looks at Earth. It is not a smooth sphere because it is having differences in masses. Suddenly you have the Himalayas with a high mass and then it goes around the Himalayas and then it dips down into the oceans and lakes and other areas. So it is not a common surface, okay? So what you see here is a changing surface of the Earth and most of the changes are attributed to soil moisture and groundwater change. Unless and otherwise it is a big change in the season. You have snowfall or snow melt and then those masses have been accounted for. So here is the equation that we use. Groundwater change, okay? Groundwater storage as a volume change. I will draw it also for explaining purpose. So think about this compartment which the grace is getting pulled every month, once a month, okay? So right now what you get is a total gravity change because from here, a total gravity change which is now equated to a groundwater by this formula. They give you the gravity change is converted to terrestrial water storage. So what you get from grace is this output, terrestrial water storage change which is any water on the top surface and ground also. The atmospheric cloud precipitation doesn't come. It is purely the Earth. Under the Earth, whatever storage is there and on the surface of the Earth, what is we have? Okay, so what we have here is the gravity is converted, anomaly change is converted to a terrestrial water storage change, which is this, which is the grace real output. Now to get groundwater out. So I'm going to say what is TWS. So this whole change is TWS. Okay, so this data is TWS. Now we know that a soil moisture is up to 0 to 400 centimeters maybe, okay? So 400 centimeters, we have soil moisture data. Below that it is shallow aquifer and deep aquifers. So now what happens is I will subtract the soil moisture change, change, not the exact soil moisture. I will subtract the soil moisture change. So now what happens is the TWS change and soil moisture change is taken out. The remaining is only the groundwater change. It's a very simple hydrological balance that you have because when you have the hydrological balance, let me type it here just so that we can quickly look at it. We had storage change is equals to precipitation minus Q, which is a runoff, minus ET and then plus the groundwater net. Or you can say, yeah, plus groundwater net, okay? So the storage can also be taken to the other side which is your surface plus soil plus groundwater, right? It's equivalent to the precipitation minus the runoff, minus the ET. Here we have in this particular example here, what we have is this part estimate, all this is estimated now, okay? Because precipitation minus Q, minus ET is the terrestrial water storage, the total storage, S, okay? So now this part is taken to be as TWS because the storage change includes the top surface, okay, and then the ground also. So that is the net storage that is becoming TWS as per the satellite's definition, okay? Let me see if I could clear it, okay, good. So what has happened is we have seen that TWS, we write it again, okay? TWS is equal to your groundwater storage, all storages together, okay? Just the storage change, the total storage change. And now just to get groundwater, you subtract it with your surface, is the storage plus the total storage, right? It is your groundwater storage plus your soil moisture storage and then your surface water storage. However, surface water, we said it's not happening. It's only your soil and groundwater. So now you can reduce it to just the groundwater. So the theory is now clear, I hope. You take the terrestrial water storage from Grace, change, it's all a change. They will not give you the raw data, monthly data. They will run it across five years and then remove the average, okay? So what you see is an anomaly, not the actual data. So for some particular reason, they don't give you the monthly actual data, they give you the anomaly data, okay? So moving on, let's see how the data comes. The data comes as grids from which you can take a point and then say this is the value of groundwater storage in that space. How do you compare it with an observation well? You can take an observation well. You can know the thickness of the aquifer by different data. Now you merge the thickness of aquifer data with the level data to get a volume and that is what Grace data gives you at the end. It is an increase or a decrease in volume and how much for it. So you can get it as grids or mass forms, mass concentration blocks. Five to 10 years ago, it was very detailed work you need to do to get this data out. But now they have given you a visualizer from which you can easily download the data and use it. If you don't want to fully download the data, at least you can take these kind of estimates from which you can clearly understand what is the trend of terrestrial water storage. And from terrestrial, if you know the soil moisture, you can actually get a groundwater, okay? So your soil moisture also follows your seasonal pattern and that data is also available in Google Earth, Engine and NASA websites. You can take it and use it. Even the other data sets that I've shown, Moon and ISRO, you can use those data sets to take soil moisture, subtract that from terrestrial waters change and then you will get the net water availability. I'm going to show you the website now. Let's start by a Google page on seeing how this works. Okay, so I've opened a new Google page and all I'm going to type is grace, okay? Data visualizer. I've given the link in the slide, so you can also look at it, okay? So the first one is Grace interactive browsers is there. You can click it to view the different data sets that are available. There are two solutions for the data. There is a JPL interactive and you use the Colorado. The mask on is a recent one. So I'm going to click the mask on and you have two links, as I said, the GSFC, which is a different center that actually gives the solutions are JPL. Most of the Indian data is based on the JPL, so I'm just clicking the JPL. For you to understand, you can take any solution. So basically you have the satellite data and the satellite data is converted to a solution, to a value and that value, lot of multiple centers work on it. Each center has their own algorithm, okay? So the JPL is what a lot of Indian literature has used, so we're using that in our study now. So what you see here is that India and all the other countries of the world, it's you can measure it as regions. So now the regions are Asia, Europe, Middle East, Africa, Australia, so it is based as regions. You can do it as basins because basins would make more sense. As I earlier mentioned, the grace works on large scale. So you need to have large scale area to monitor the groundwater, change an anomaly. Okay, so you see here, you have the basin as a Ganges basin, okay? I'm just going to click the Ganges basin. So the mouse, you just click there and it picks the mask on that particular grid. So I hope you could see the grid. So this is what the satellite, each time it measures one value for the entire grid, okay? So that is where you can see that it is not a very small scale district or a block level analysis. It is a massive analysis. And from there, you can try to understand what is happening in the district, but not accurately, okay? So that is one limitation of grace. But as I said, for large scale basins and Ganges studies, Kaveri basin studies, this data has worked. And I'm going to show you the Ganges basin here, okay? So within the Ganges basin, there are multiple grids. So this grid might have a different trend line, but all of it, since we are using it as the basin, the basin value is given. So JPL mask on Ganges Brahmaputra basin is the name. The trend is minus 1.53 centimeters per year. The terrestrial water storage change is going negative. So this includes the Ganges and the Brahmaputra. People would think that there's so much water in that basin. Why is the terrestrial water going down? The flow is not coming down. So what else should come down the groundwater? There has been extensive pumping in this region. So if you multiply this by the area of the basin, you will get how much water we are taking out. And that could equate well to your 265 kilometer cube per year analysis, much more than that for the entire India. So you could have this trend. You could see how the water, terrestrial water storage changes. As I said, it is an average. So above the average or below the average is what is going to show. So in the net, it increases slightly above, goes down. So this seasonality is because of the monsoon. Whenever you have the monsoon rainfall and goes up and down, when there's a big drought here, you don't see a big push in the upper direction. But with the seasonality, if the level comes down, then it is a pumping scenario. Because the season is going like this. You have a summer, you have a summer in the bottom and then you have rainfall, summer, rainfall, summer. Okay? But if the overall trend is depleting, that means that your groundwater and other water resources there is depleting. Because the rainfall is actually trying to push it back, but it's not getting enough strength. It's not having enough data. So there is a loss. Okay? So and every month, when you move this mouse on the data point, you could see that this is September 13, 2015, August 21. So the peak monsoon seasons are captured eventually. And the value is given and how much it is given. You can change the units. You can have some other trends on and off if you want. And also you can take a different point in the location if you want. You can close this and take another trend. So this trend is by pixel. So the pixel is big, as you can see. And you can just click that pixel and you can see, for example, here, this mask on. This whole grid, which is covering right in the center of Rajasthan is declining. So the recording has started again. So the recording of data is going down, right? So then what we can see is we also noted that somewhere the groundwater is going up, right? So let's take this pixel, for example. Okay? So you can see it is going up. So this actually relates very closely to the CGW media data that we saw. Okay? But what is the difference here? The difference here is it is monthly and it covers the entire whole area in one go. And it also includes deep and shallow aquifers because it is terrestrial and down. Anything that down that changes is going to impact your grace gravity record. So you could also save the data as an image. You can take this as an image for your studies or you can also get the data as a CSV file and then we can change it. So as I said, you could easily get the data and then subtract the soil moisture to greater groundwater. This course, this particular lecture is to introduce this topic. There are a lot of literature, a lot of tutorials on this on YouTube and free open source platforms. Please go there. And if you're interested in Grace, have a look at it. I've introduced the topic. You could learn more. They'll give you a step-by-step how to do these calculations. With this, I will stop today's lecture. Thank you.