 Hello everyone, welcome to the NPTEL course on remote sensing and GIS for rural development. This is week 12 lecture 2. In this week, we have been looking at the applications of remote sensing and GIS through case studies and live applications through dashboards and other data sources. In the last lecture, we looked upon the nabad boven collaboration dashboard where we looked at the locations of structures. So I will just showcase some more details that I have extracted just by going around with the software and website. So you can see here that we have Dahod region and I've selected the IGWP project, the Indo-German project, Watershed development project and of all states, Gujarat Dahod we have selected. You can see here Dahod I've selected and a period of 1 Jan 2012, which is the earliest until date yesterday, which is the latest. So we have around 1,936 points of data and you can click on this to see what it is. There is a particular and then there is some water rejuvenation, tribal social welfare society board. A farm pond has been created and then multiple photos of the farm pond has been made. Some processing livelihood options have been given for the tribals and then some slope stability have been done. So pasture land conversion of barren to pasture land etc. So all these are part of, these are bunds to slow down the runoff so that water stays and then rejuvenates the aquifer and soil moisture. So these are good for application and image is very qualitative. So it's an angled image and there's a lot of issues that can come with an image because maybe it is not taking the correct position, sunlight etc. So that is where satellite data added on to that will give a better application. So what we're going to do is we're going to show a study that has worked on this data and produce maps. In the meantime I've also selected Dahod region and for two time frames I have selected the data. So for example you have 1995, 01, 11. So Jan 11, 1995 I have data and then Jan 14, 2022 I have data. I can remove this one, March is out of the picture and then 1991 is also there. So 1991 can go above. So I've selected this in the meantime. I've shown you how to do it. You select Dahod, go to discover data. In the back to search you can search for data and then come to compare and then compare the data. So once you visualize you can go back to which data set you want and then you can add it to the visualization if needed. So I've done it. In the discovery you will find which data set in the search data set and then you will see which options you have and you want to add pins, different advanced options of time span etc. So here in the compare you could see that 1995 is on the top and on the bottom is the 2000, 1999. You could see that the water levels are increasing for the same month. Could be because of rainfall also but look at the NDVI, NDWI which is a water index are getting bigger. What is NDWI? We can see here green means less water, blue means more water. And so we could see that there's a lot of built up area coming up here and that is why you see a green color and then the blue color signifies more water. So on the top it is 995. I'm bringing it down. Now this is 1999 we are seeing and if I reduce 1999 you will see 2022. So in the 2022 frame beautifully you will see more water bodies coming up in Dahod which is a good sign of water being stored on the surface and being used for agriculture and stuff. So this is a crude estimate but let's see what we have done through studies and come back and revisit this aspect. So I will go back to my slide for today. Before we go into that I would like to introduce you the many options to look at applications of GIS do exist. GIS geography has a website in which there's a lot of data, accessories, carriers, analysis and then they are updated regularly some steps etc. I can just show you real quick if we click this we will open the webpage for the applications. So I'm just going to go to the screen where it is opening up. It is a screen one. I'm going to share and now you can see it. So you can see that the 100 applications have been done almost a year ago and just see how many rural applications can be there as agriculture, soil types, NDVI is big. So those who like to do NDVI you can know that it is one of the big applications and then Antarctica which is not part of us and then climate change, forest, agroforestry for rural development and then disaster monitoring, damage after an earthquake, ecology, habitat monitoring etc. So there's a lot and a lot of data around this line and you could see that how you could take aerial photograph also in the military time and then snap it to the current areas also you can do mining can impact rural development. So you can also map where the mining is happening, societal issues, human rights etc etc. So there's a lot of applications again solar panel, solar options is very very important for rural development that is also happening. So we have all these and then as I said we will get back to the presentation where we have this link posted and you feel free to go ahead and look at these options. Please remember that all these options do exhibit a cross-cutting team because rural development is a complex entity. There's agriculture and rural development, ecosystem and livelihoods, domestic use and sanitation has to be addressed and climate change which is hampering more on the rural development scenario has to be addressed. So all these are there and we have to be very focused on collecting data and mapping them so that we attenuate or reduce the impact on the ecosystems. So we have remote sensing identified for each and every parameter in this lecture. So for example, agriculture we use NDVI, NDWI, Landsat Images, groundwater from Grace etc, ecosystem, livelihoods which is mostly dependent on agriculture we have. We also maintained some aspects of some aspects of animal, husbandry and poultry farms, aqua farms etc. Domestic use we looked at groundwater, rainfall availability, farm ponds, water for geology and mission and of course climate change. We have shown how to use climate indicators and climate change scenarios from remote sensing estimates. So we have a lot of remote sensing for rural development case studies. The applications involve evaluation of NGO-based work. So this is what we have done in the first section which is today. We'll be looking at how to evaluate NGO-based work because NGOs are very focused on working on the ground. I've worked with multiple NGOs. I currently still work with NGOs and they are on the ground who are working very, very hard for the rural development, most of them I'm saying. And what they do is they work very closely with the farmers and stakeholders and bring them up in their potential adequate options. Salters are very low so not a lot of people you'll find in NGOs and even we are finding hard to send our students into those kind of streams because most of them want higher paychecks. But the point here is it is very, very related because with less money they cannot afford to build capacities for collecting data, monitoring their impacts. Suppose we have an NGO that is working on rejuvenating farm ponds or agricultural lakes. The idea is they will rejuvenate it. People have the potential benefits and they continue rejuvenating it. But normally what happens is they'll ask the farmers and people are you having water? And if they say yes, great water is coming then they happily move on to the next. But if someone else asks them what is the quantity? How much water has been improved? What is the metrics, the yield that has been improved? That is very hard to quantify for them because they don't have budgets for putting people on monitoring. They have budgets only to do the work. They don't have budget to do the monitoring and evaluation. So the assessment of impact is limited. And this is kind of sad because NGOs need to be accredited for their work and for the knowledge that they develop. If not then the system will collapse. If no one knows that the NGOs are working hard or not, then how will the funds come to the NGOs? So only the top NGOs that are very good, very well known will still establish. So it is important for them to do assessments and impact evaluations for which we have worked with a foundation in NGO called Dahod NMSaguru Foundation, which I have already told. And as I said, they have very less budgets for manpower and models and software to evaluate their impacts. Capacity is very low. We will need to build capacity for that and data acquisition, observation data is also not enough. So in this case, remote sensing and GIS can come very handy. In fact, they can alleviate all these stress on the system and bring a clear indication of their impact and work. So what method did we use? We used just some ground data that they had. I will go through the study very in detail and show you how quickly you could do and get published in a very, very good journal because I hope Masters and PhD students are also taking these courses. And it is very important for PhD students to write papers. What I'm going to show now is a student intern. He was not even a student under me. He was a student intern, just internship two, three months. He worked very hard and this paper through the team came out to be a very good journal paper. So time is also needed where we need to put a time on the assessment period. So you need ground data and time. That is all you need. So when did the structures or the investment or the infrastructure come in? That is the time. And ground data of the locations and that's it. The remaining satellite data can address. One more data we need is the rainfall. Again, even if the rainfall is not available from the ground, we can always estimate it through satellite products because we have satellite data coverage of rainfall for a long, long period. So this article had come in the NRF, the National Resources Forum, which is the official journal for the United Nations, a very, very prestigious journal. I'm very happy to say that a world by two interns have ended up in this, who are the third and fourth authors on this, who wrote along with me and one collaborator from Taiwan, and then the NGO person also. So number four is NGO. So you can see that the NGO person is also involved. So these NGO people normally don't have time to write papers or evaluations because they're always on the field. They have to work hard with the farmers. And since I've been on both the sides, I know exactly like how much time we have to spend on the field. And once you come home, you don't want to open the computer. It's just getting ready for the next day. So we as academics and institutions should support the NGOs who are on the ground working very hard with the people to bring up their livelihoods along with the government. So let's take this case study. We have worked on the Dahod district, and you can see Dahod is part of Gujarat. Gujarat is in blue in color. Then we have Dahod region, which is this part, again, blue in color. And then we have different basins, small smaller basins have been demarcated roughly from the data they had. And we can also take a DEM and then do the watershed analysis. But also what you found is there is a lot of check dams. So the idea here is the Dahod and NM Sathburu Foundation, the foundation is called NM Sathburu Foundation. They invested a lot of time, you can see here NM Sathburu Foundation, water and development foundation Dahod. They invested a lot of time in getting budgets for building check dams. And then they built the check dams for the farmers. They didn't stop there. So their mode is one more higher level where they formed village communities to use and manage the check dams through lift irrigation system and other things. So now we have our system which is blocking the water and then water stays on top, the surface water, and then some water infiltrates, percolates into the ground or lack of it, where most of the water is still there and getting pumped out using lift irrigation schemes to the farmlands. And most of this region, our tribal region, and most of this region were not under agriculture for the past 100 years. And that was because it was initially a forest. So if the forest evolved with whatever rainfall it had and soil moisture, but now a lot of forest has been cleared and the clear land could not sustain any growth because of limited rainfall. The rainfall is very less around this region. We'll have the data to show what is the average rainfall. It is somewhere around 400 to 600 mm with some odd peak discharge and peak drought years. So we have all these sub basins classified after we had the location of the check dams. So only the check dams wherever they are, we mark the boundaries and we found that this specific boundary, the Hadaf river basin had more check dams. And we had selected that for further analysis as follows. So what we did is we first plotted the time. As I said, we needed time of ground data. So when did the check dams come? So approximately 1990, they started and they started keeping on increasing the check dams. So this is a cumulative graph of the number of check dams. You can see number of check dams increase steadily from 1990 to 2000. And then there was some slow development, not much development here. Maybe they were building larger check dams and it took some time. And then slowly picked up again until 2015. So we used data until 2015. We wrote the paper around 2019 and it got published around 2021. So the idea here is we had collected the locations and the time of the check dams. So now you have 1990 to 2015 the check dams. So if you look at the rainfall and see how the indicators of soil moisture and plant growth happens, you will definitely see an increase because of the check dams because the check dams have a principle of storing the water and then letting it percolate and then the plants taking it out. So if you know that the check dams are coming into existence, especially this part 2000, 2015, a lot of big check dams came in. You could definitely see that for the same rainfall. The dashed line is the rainfall. So for the same rainfall, you will see higher peaks of the NDVI because they are giving more water in the storage rather than as runoff. So they're converting the rainfall into runoff and then keeping them in the system. So you could see here as the results. First let's see what is the NDVI value range. We have minus 2 plus 1 and the plus 1 is dense vegetation around 0.5 to 1. We have set good vegetation and as far as 3 to 5, 0.3 to 5 and then 0.1 to 0.3 is soil or rock structure, barren land and then we have less than 0.1 is no vegetation or water bodies. So you could see that the NDVI certainly increased a lot. All these are the different check dam basins that we saw here. So these are the 1, 2, 3, 4, 5 basins which are given here also 3, 4, 5 plus rainfall. So rainfall is looked at this axis, right hand axis while the other all have NDVI in the left axis. So you could see that the dashed line rainfall ranges between 400 as I said 400 to 1,200 is the peak here but most normally around 800 and 600 the rainfall level comes up. And what happens here is you could see the response of the rainfall on the NDVI is changing. So if the rainfall is the same you would almost expect a similar NDVI because the water is being taken by plants and plants grow. But if you have storage then the water is being stored more in the storage tanks and those have access to plant growth soil and other components. So you could see here for a rainfall of around 1000 you have only the NDVI around 0 to 0 or negative 1. So most of the negative parts are here in the early stages. So check dams are coming but slowly they will improve the quality and then you could see that suddenly they start to peak and then go away from each other's obeisance especially more on the positive side even the rainfall is coming down. So that is the impact. So the impact here is even if the rainfall is coming down after let's say 2012 to 2015 and then 2020 even the total annual rainfall is coming down in this areas the NDVI is still increasing which shows that the structures are there who have buffered who have stopped the runoff from the system and kept the water into the ground water so that it can be used for crops. This is a very key finding in the exercise. Why? Because we assumed that NDVI would be constant without any change with rainfall and the check dams will have slight indication on the soil but it was not the case. Plans were happily extracting more of the water and growing healthy. You can see the NDVI is very healthy around 0.2, 0.3, etc. around sparse vegetation. So here you don't have dense vegetation because of the rainfall region and also it has been always like that for a long time barren land. So converting a barren land to some kind of crop land is very, very difficult but slowly it is happening thanks to the efforts of the NGOs, NMS of Guru Foundation and the check dam idea that they had. It was not super scientific. It was something that worked in the region and they just used it which is a basic, basic science. You don't need rocket science to save most of the world problems. You just need good science, basic fundamental science. So here the farmers are extracting more water. So how do you reduce it by adding more water structures that can capture the runoff and put it into the ground? You could see here this is another testimony of what is happening. We picked two years of same rainfall, 1991 and 2017, same rainfall. So someone should not ask us, oh, you took rainfall high year and showing high NDVI. So we wanted to make sure that the 1991 and the before and after are the same. So around 650 is the average rainfall and then for 2017, 2018 and 2016 it is almost the same. So there is no big flood before this event and this year and then we saw 2016 is okay but 2017 was better in terms of average rainfall 650 and 657 millimeters is almost the same. So let's assume that both are same. Just 7 millimeter difference is not big and now you could see a spatial distribution of the location of the check dam. So these are the location of the check dams in one of the sub basin which is Khan Basin and you could see that how the color has been changing from negative, the negative values are red and yellow to blue. So blue is happening a lot because and the water flows from top to bottom and then you have a lot of water that has been stored. So you can see a big, big blue color here which is being used for recharging and growing the plants. So NDVI value is really high. There is a lot of crop growth and crop diversification also which adds to this finding of increased NDVI. So we saw a graph of how it changed the average value of NDVI change across the Dahor district which is the previous one and at different timescale. So we have different sub basins and then different timescales from 1998 to 2020, we have the NDVI. But in the spatial resolution image, we picked two years, two years with similar rainfall pattern and precedence conditions and we looked at if the NDVI has sharp change and you can definitely see a bigger change in the top basin where most of the runoff is going to be held back and then stored for improving the soil moisture which improves the NDVI value. For NDVI, again we have the negative values as not water present in the soil and are not healthy water levels. So you can see here from the explanation of what is the NDVI, you could see that the NDVI value ranges from minus to plus and then vegetation has smaller values. So vegetation and built up areas have small values whereas the higher values are blue and that is what we are also seeing here. So here with negative values which is mostly the built up and barren land whereas the soil water capacity has been increased drastically and that you could see definitely in the drought years even though the rainfall is coming down, the NDVI value does not come down fast. Here we have used zero to one as the water body water content, the positive values are reflecting water content whereas negative to zero is bright surface with no vegetation or water content. So this is just barren soil, land built up and you have less water content in the soil. So inside the soil there is more water content and that is even true during the less rainfall year which is being supported throughout this image because of the use of satellite data to understand this phenomena. So you can see down by the difference of water index from 1990 to 2018 for the study sites in the whole of India and it clearly says that due to the increase of check dam the NDVI values have increased and also when the rainfall comes down the NDVI value does not plummet down at once because there is water storage in the water structures. So it will help to improve the NDVI and NDVI value which we see in this small exercise. This is very, very important because nowhere could you see that the NDVI is about peaking the rainfall data in a high extreme. So for example it's a positive rainfall, it goes positive but here the rainfall is going down and you should expect the NDVI and NDVI to go down but you could see that both are either stable, just stay stable or it goes actually up and the up is because of you have good water storage which is still recharging the surrounding area from the check dams. So with this I think we have looked into one exercise where we have this change in the rainfall pattern but still it doesn't negatively impact the NDVI value and that is purely because of check dams because check dams do have that potential of doing it. So this is a spatial recognition and as I said the NDVI may not change drastically because when these small, small vegetation shrubs are grown. If you go to NDVI, this region, the tribal region of Bahut, you'll see them mostly growing pickles, chillies, some fruits and vegetables like tomatoes, cucumbers, etc but not high value crops and crops that can cover the entire land. So NDVI would not be a great estimation but NDVI which talks about the water content is and you could clearly see that everywhere it was red during the 1991, 650 mm rainfall period but during the 2017, 657 mm rainfall period you could see that it has been improved from 0.03 to 0.1 to 0.5. So all of these areas converting back to positives and the positives is definitely due to the increase in recharge from the check dams and spatial distribution of the water to all these locations and all this was done without even going to the field because I know the field, I've been to the field but the student had only very limited time of 2 to 3 months to work on this. So we gave them the location of the check dams, gave them the theory that the check dams, the hypothesis, the check dams improve the NDVI or the null hypothesis, it does not improve the NDVI value for which we had extracted the values and then plotted it and then see visually comparison and also a very quantitative way of comparison because now we have the values, pixel values. So we can quickly say that 80% of these pixel values are above the red values. So that clearly indicates that the soil water content and the surface water body storage across the basin has improved drastically because of the check dams. You could also see more and more of this happening in the downstream areas because water can calculate from high elevation to low elevation and that is what you're seeing that the water comes down and then gets reallocated into multiple sectors. So those who cannot do this as a full exercise, this exercise is done like you download the Dahor map first step and then you can put it in the location of the check dams, download the rainfall, download the Landsat bands. Basic bands was for NDVI we had NWIR minus red and then NWIR plus red whereas for this NDVI you could use a different band which is given here. So normally it's different registration index you can have for Sentinel 2 I think we use Sentinel Landsat, Landsat images because those are older images so B3 minus B5 by B3 plus 5 and for Sentinel you can use B3, B8, B3, B8. So really good, really good data set that can help. These are very, very important indicators and that's why Sentinel Hub is also having it and you can clearly see that the difference between the images is very comparable so I'm just going to keep this image to the side here and then this can come to until Dahor half and then this can come here. So you could see that on the right side you have a Sentinel image from 2022 in JAN. So almost JAN let's say the same rainfall is happening, similar rainfall but you have better water bodies, number of water bodies and the number of water structures are really high which is really good to see whereas here you can see that it is all white this is no not enough water and this is also not enough water somewhere good water is there and then you have these big water bodies but all of this is being increased drastically much more deeper water levels are there and that is what is reflecting in the NDVI values. So with this I think we have showcased one study the time taken is very less two months is needed if you know how these GIS works and honestly I'm saying if the student knew how to work with this dashboard he would have done it within a month also so you can just include your area of interest download your data or upload your file for your boundaries and then after you log in you can just extract these data sets as needed and then you do have a data protocol to sign up and say I need this data and boom the data comes. So you could actually do wonders with this area of data and as I said you can add your shapefile of it so first basically we can make the shapefiles in GIS or Google platforms and then save it as a shapefile bring it here and then download the data so it's always easy to download the data from these sources it's legit it's a user space agency so you won't have big bugs while storing the data just be careful you don't have pop-ups which could have been done by previous installations of other softwares not QGIS okay so you can download this image as a jpeg you want to show options or not and then with no georeference or you can georeference it okay so you can download and then georeference also as I said you can do analytical basic download image option is there maybe they're asking you to pay initially these were also free to give free access to this but now you have to pay so the point here is you can quickly do these maps you don't have to take and subtract two images to calculate the values you can quickly do this and then estimate the differences based on the water that is being stored in the checkmaps and to be honest there's not a lot of these studies that have been done like this and one of the reasons that's why this has been accepted in a very very prestigious journal whether I would stop today's lecture and I will see you in the next lecture where we'll talk more about some applications from the group and how you can use RS and GIS for other signal thank you