 Hello everyone, welcome to rural water resources, week 3 lecture 5. In the previous lectures, we looked about groundwater and spent some extra time on groundwater resource because one of the important factors for rubby season for non-monsoon irrigation for rural water. I stopped with some data collected from groundwater board and government data and we looked at how it is being used to understand critical blocks. So that is one way of monitoring the groundwater availability, groundwater issues, etc. in the country. But please understand that that doesn't help the movement of water, it only documents. We have a level here and now is it coming down or going up? Where is the water going? It is very, very hard to visualize and connect, unlike your surface water. So for that concept, we have groundwater models and these are very complex code heavy driven models where you have equations solved for the groundwater properties and groundwater hydraulics. So the hydrology would look into how water comes into the system for groundwater and then gets relocated along your porous spaces and then moves and establishes the head of potential. So the models are many and plenty in number and I would be talking about one model that I have used in the past, which is modflow and it is an open source model and we will look into how it is being built. So it is based on physical equations as I said, understanding the physics of how water moves from one place to the other and also on the soil properties and the geological properties. So basically it is a 3D model, which first you see from the top down. So maybe I am looking at a map from the bottom and I could see a boundary. You see this is the boundary of your aquifer or boundary of your model where you want to model. Then if you go deep into the model, you could see that it is a layered 3D model. When we were discussing about groundwater, we mentioned that visualize a cross section and inside the cross section, we have layers like a cake. I used the term like look at this as a cake. We have a cake layer, a cream layer, a cake layer, something like that. Similarly, a 3D visual is needed. So these conceptual models are being made by modflow wherein you have the first layer on the top and then a layer which is impervious or it is stopping the groundwater movement and then a deep aquifer layer. So here you could see the first layer is slight green and then a darker green layer, but it is not continuous. It is not continuous. It goes and stops, which means some layers exist not for the whole part of the aquifer, but for partly and it could be because of the weathering of the rock. Then you have two more aquifer layers which are slight gray and darker gray in color. They move a little bit higher than the green color and then you have the darker orange and orange which go throughout the boundary. So what you see here and these images are taken from modflow, the multiple modflow platforms. I am showing examples from aqua video. So what you see here is water comes from the top and then rainfall, water infiltration population, then water gets relocated along the layer and after it gets recharged it goes down, down, down to the deepest part of your groundwater aquifer. So in between sometimes it will not move beyond a permeable layer. So all this has to be captured in your model. How do you capture these in your model? It is by data and field observations. You should know how many layers are present, distinct layers in your location and how long it extends. For example, if I take a bore log here, I found 1, 2, 3, 4, 5, 6, 7 layers. If I go here, I only found four layers. So which means three of the layers are truncating. So we should know by interpolation which are the layers that are truncating. So all this is done by the model. All you have to do is add different locations, the water level record and the stratigraphy record which means the layering of your aquifers. So this requires a lot of field work and this is where the aquifer mapping activities of the Government of India is being done. That's why it takes a long time. You have to take multiple samples, bring it to the lab, analyze what material it is. If it is distinctly different from a different depth, then it is a different layer. It is somewhat closer, then they just merge it into one layer. Most of the groundwater models for study purposes, research purposes have three layers. From the top, ground surface, you have the unconfined aquifer, then maybe one or two impermeable layers. So there's one layer, unconfined aquifer, there's a confined aquifer and another confined aquifer. So almost three would be okay to model. Otherwise, you need a bigger model computing software or even hardware like a good computer to model of these. So I've shown you that just data would be enough to understand just the level but not the movement of groundwater. And to understand the movement of groundwater, there is a need for physical based models. When I say physical based, it is driven by physical equations, empirical equations. Empirical equations are based on statistics and it is just a correlation relation kind of a function. Physical equations actually take the physics behind the groundwater movement. For example, hydraulic conductivity, the change in pressure heads and then they calculate the movement of water. So for physical based systems, there's a need of a lot of data. So there are a lot of issues which we will go through and I am defining what are the issues. And please bear with me when I do extend the groundwater because India is the biggest groundwater extraction. Studies in government reports say almost India's groundwater extraction is bigger than the next competitors combined together, which is India is, which means that India's groundwater use is much bigger than US groundwater extraction and the extraction by China put together. So we have to be more protective of the groundwater resource. Yes, it has helped to improve the economy by means of increase in agricultural productivity, better livelihood options like sanitation and improved lifestyles, etc. But we need to preserve it otherwise it would be back to base one. So groundwater data issues, let's go through some of them. Some of them are not representative. When I was teaching the groundwater level wells, I told that some wells are farmer wells which are representative, which means that when you pump water from a well and then take the same reading or come again a couple of days later, then take a reading, then it is a representative well. It represents the actual system where water is being used for pumping, etc. But in some regions, you would see that the monitoring wells are not disturbed, which means they are not pumped, it is only for monitoring. So it won't capture the dynamics in groundwater because it is disconnected, it may be disconnected from the farm groundwater reality. So that is non-representative. Less frequency because of the cost in manpower and our collection of data. There is very less frequency of data, both spatially and what do I mean, frequency less in spatial? There could be much more number of wells that are spread of wells. We saw that Rajasthan, Gujarat is in the right color, but the number of wells is also less. So some more coverage of wells can be there. So that is the higher frequency, improving the frequency of spatial representation of wells. Deployment of wells, then we come to temporal issues, only four times a year it is being monitored. Is this enough? A lot of studies say that we need to improve it because the pumping doesn't happen only in those regions, it can be a cumulative effect. So what you are studying by looking at a pre-Monson level is the accumulated water use impact on the groundwater level in the post pre-Monson levels. In the post-Monson levels, it is also cumulative effect of recharge. So now here comes the question. So which month is most important for groundwater recharge activities? We don't know because it is a cumulative effect. So we'll have to put much more effort in monitoring the wells. For example, in developed countries, you would see that the wells are monitored almost every day by automatic sensors, but it is pretty costly. But other wells are at least monitored once a month, at least once a month. So low density and concentration, sometimes what happens is a state might have enough number of wells, but they are not spread out equally. Some areas are low density, whereas some areas are concentrated. So that issue should be more or less taken care of. Otherwise, the representatives from the learnings from groundwater, such groundwater data. However, I do acknowledge that there are a lot of cost issues involved. We need to understand that because of the complexity involved in groundwater, both the capacity, a lot of people know about groundwater monitoring. So there's a lot of cost in bringing up the capacity. There's a lot of cost in putting a well and just one pressure transducer. I was saying that you put in a piezometer or a deep aquifer well. It would cost around $702,000, which is approximately 70,000 rupees. And that tape measure I showed you is one lakh rupees. So now you could understand that it is very expensive to monitor these wells. So we need to find better ways. We need to find better technologies, cheaper, cost-effectlier to manage and monitor the groundwater. Some wells are abandoned and are not disconnected, which means if you go and look at the data, some wells would have data, data, and then suddenly no data. Those are the abandoned wells. People stopped recording for some reason, and other ways they are disconnected. Okay, so they are not connected to the pumping region or there might be presence. So these kind of issues because they didn't understand the underground complexity and where to place the wells. Again, I'm not talking about any particular agency. Most of the agencies in India go through these issues. So there's a lot of different agencies that are monitoring groundwater both central state. But we have to be understanding which wells that we are monitoring. Some wells are polluted. When I was in the field, we didn't see a lot of wells polluted. So which means level is increasing, but it is not good water. So we'll have to be very careful if we are monitoring and measuring good water or is it polluted water? For example, an industry might be discharging pollutants and move into the groundwater levels and increase the groundwater levels. So is it a good sign? But with polluted water is the question. So it has to be good water. So always look at if the wells are polluted. Sometimes metering is not possible because of the complexity of the wells. So sometimes the sides of the wells are so rugged that it is not possible for metering. Contamination, pollution moves and sometimes contamination also happens. Some people purposely put in this bad water. So like we see how streams and rivers are black in some areas. Similarly, contamination also can be up because it is a storage underground. So if you push water in, then water goes into the groundwater and gets contaminated. For them, it's easy. They just put the black water inside, but then they're actually spoiling groundwater. So when you go and collect data, please be understanding of these issues for groundwater. Always have to do a data quality check, which means data representing what is happening. There is good rainfall in the monsoon and you see a good water level increase, which is good. So that's a good data quality. But suddenly if you see a ground water level increasing and throughout the drought year, something is wrong. So that's what I'm trying to say as data quality check. Sometimes you will see the same water levels muddying the data set. I'm talking about a state data set that I recently bought for research and we saw a lot of issues in the data. Data errors, duplication, which means one level is the same across all the months, which is not possible. And then sometimes you would see a data, which doesn't make sense because the units have been changed. So please understand that I had mentioned earlier to be very careful with the units. Because if you spill the units here, then your data is off. One side is the groundwater data issue. And the other side is the groundwater management issues. Right now, we have Central Ground Water Board, which is their body other than the state PWDs. It will be good to have more agencies involved. For example, litigation looks at the irrigation department, looks at water supply for agriculture. But it is mostly through the surface water pollution, which is dam, scandals, etc. But a good portion of water is also being used from groundwater. So some convergence of management could help, especially at important resolution of the data, the management activities. Some empirical methods are being used. For example, to understand the infiltration of recharge rates, those should be backed up with physical base models. It is time-consuming to run a mod flow, but on the long run, it does help because you have a 3D picture of your aquifer rather than an empirical model, which is just a 1D. Which means right now, infiltration, runoff. So there's not much dynamics laterally also. So groundwater, when it comes in, it doesn't only move down vertically, it can also move planar, XY plane also, it can move. So along the Z axis, it can move up and down, up due to capillary and down due to gravity, but also it can move XY plane, so that that cannot be captured by empirical methods. Some methods are outdated because how do you account for pollution? How do you account for deep aquifer? There are too many wells that have been abandoned. So all these need to be double checked. There's a lot of one-size-fits-all for groundwater management activities. For example, checkpoints, people claiming that it can recharge groundwater, but it can recharge across all of India is the question. So just because it reaches one area, it doesn't mean that it can recharge across India. So if you remember the traditional water harvesting slide I showed or the traditional water body slide I showed, you saw that in traditionally across India, it is not the same method that has been used throughout. For example, we had airy or ponds in Tamil Nadu and a terrace farming along with water storage along the terrace. You saw that in Jharkhand, Nagaland, etc. So there is a big difference and so one-size-fits-all approach should not be used. And because there are data issues, we should look at different data platforms to be implemented with the observed data. So here I'm showing the only satellite in the world that can monitor and map groundwater, which is the gravity recovery and climate for the grey satellite. It is a combined mission led by the NASA team. And they also have other data from different satellites for other parameters that contribute to the groundwater hydrology. For example, your soil moisture, your rainfall, etc. So those data can be collected from Global Land Data Assimilation Systems or GLDIS. All of these are open source and the Bowen GIS are remote sensing observed data. So these are obtained from the Government of India's website where you could map the land use land cover. So if I know how much rainfall is coming, if I know by land use land cover map what kind of crop is grown, I would know what was remaining for the ground, what will recharge. And all these different parameters can be put together to understand groundwater dynamics. So you can clearly see that I'm not omitting observed data. Observed data is important. It has to be the first data that you would use. But because of the spatial and temporal scale, you are adding some remote sensing data along with it. So this is how Grace would look at it the globe. It is not the same smooth sphere, but there are a lot of disturbances on the land because of changes in gravity. Gravity is not the same across the planet. And using that concept, it measures groundwater. So all remote sensing data are giving you a proxy of what is happening. So by these methods, you can actually estimate groundwater recharge at a higher spatial and temporal scale. Only drawback is remote sensing methods cannot be used for a field level analysis or even a village level analysis. We'll have to do it at a state level. So even state level is not agreed upon, but at least South India or Indian subcontinent level is OK. So it gives you some pictures and you could see here that blue color means higher groundwater or terrestrial water storage. You could see that higher water storage is available along the Himalayan regions because of the snow. Those kind of things we could understand from this. Those who are interested in Grace papers can look at more Grace-related papers. There are multiple papers even for India that has been published widely. On the whole, still there's more data needed and more people are needed to convert remote sensing data into a usable format for which ID Bombay led the Mapathon event last December. And we actually looked at how we can engage people, map different locations in India and also use ISRO data, which is the Government of India's data for research purposes, for analytic purposes and to have a better understanding. And so we don't expect every time a government agency to map and give you a land use land cover, for example. So it is those people who are interested should get on the ground and help these kind of data activities. So this event was a great success where we had 9000 participants just to show that collecting data doesn't require just a government agency. But you could work along with the government remote sensing data platforms to generate some data. And depending on the quality of data, someone can use it. So that aspect is always very important. The key players here were ISRO because they are the Indian Space Research Organization for Indian satellite data, Indian Institute of Technology Bombay, MAICTE, and within IIT Bombay we have the FOSI, which were funding and spoken tutorial, etc. So it was a great success and we look forward to doing this again. Point of showing this here is to say that there is a lot of data needed and conversion of these remote sensing data into products can be done by collective action. For the recap, for week three, we looked at hierarchical parameters. Overall, we had around 10-15 parameters, but we said okay, for water management it is not needed to look at all of the 15 parameters, but more important focused parameters. So we looked at precipitation, evapotranspiration, and in the first week before this. And in the current week, which is week three, we looked at service water storage structures, how water is stored in depressions, then leading into channels, streams, and finally rivers and lakes, etc. Then we looked at soil moisture. So whatever water is stored and after that some water is runoff, whatever the remaining water does move down due to gravity and it gets stored in the soil profile for plant use or evaporation, which is together called as evapotranspiration. So soil moisture we looked into detail and we looked at how to measure all these different parameters. In the groundwater, we took one more lecture extra given the importance of groundwater and river water management. And because now you cannot see everywhere people have any channels or canal command areas for their irrigation plots, but you do see a groundwater pump and a groundwater well. So sometimes one well can be used for five different farmers on a rotation basis. So the groundwater has actually increased the access to water, community farming, etc. There are downsides, which because with one person uses too much, the other person does not have water. But those are the management issues that we'll be looking at when we come to groundwater management issues. So this I would like to conclude the week just discussing about the hydrological cycle. We will now more into the aspect of water management because now you would have an understanding of what are the, what does precipitation mean, what are the leakages or losses to the system? So these aspects when we talk about this class, you will have a better understanding because we have introduced the parameters. Thank you and I will see you in week two.