 Welcome to today's NPTEL course on remote sensing and GIS for rural development. This is week 10 lecture 4. In this week we have been looking at online platforms and GIS platforms with remote sensing data that can help us to quickly download NDVI data and other crop vegetation index data. Across the research literature, NDVI ranks number one for having a lot of remote sensing based studies for assessing the vegetation, health of vegetation, area, acreage, etc. Because it is very simple to use, very efficient and open source. We also have vegetation fraction which a lot of these platforms house and before we get into the other indicators, I wanted to showcase the four different platforms that house NDVI when we saw in the previous lectures. In the last lecture, we looked at Google Earth Engine which keeps on updating and today we will be looking at the NASA's data sets. Why NASA's data sets is it has high benefits on spatial and temporal resolutions when compared to the other data sets that we have discussed initially and because of the high global coverage and spatial and temporal resolutions, there is multiple scientific articles on it. As students, whenever you want to collect data and form a hypothesis or if you would like to support an idea, the best way to do it is first do a literature review. In the literature review, you find papers, recent papers and look at what methods they have used and you will find some technologies that everyone can use, example NDVI. There are other higher indicators with better resolution so-called or better analysis, however, they might be expensive. We need to make sure that everyone can map at least in the initial stages and that is what NDVI does for you. It's very simple, it is NIR minus visible red by NIR plus visible red and almost all open source satellites nowadays have these bands and even the older satellites. So let's move on and we have actually looked at ISROB1 in the previous lectures and then Google Earth Engine. I had also indicated that there is some coding that you could do. As I said, I will refrain from teaching how to code because that is not part of the current exercise. There's a lot of forums that you could go and find the codes. You can just Google and say that Google Earth Engine making charts and then you'll find a lot of videos. What is missing is how does that relate to a particular topic because that is domain expertise. So in this lecture series, the entire remote sensing and GIS for rural development, not only am I giving you the access to remote sensing tools but also making sure that you know where to use it. Again, remote sensing tool and GIS is started in civil engineering, earth science engineering, geography, remote sensing as a class by itself, satellite technologies in our rural development courses I teach it. Now even policy teachers and law teachers it. But where we would be different, this NPTEL course is very different is we align it to a particular objective which is rural development. And of the rural development, we assess that croppings have a very, very high impact because most of the population depends on agriculture. So we had more focus on it. We will slowly look at other options also. But please remember that other rural development indicators and indices that we initially started with rural infrastructures, rural roads, rural schools, rural hospitals, all these have very, very less data and remote sensing is one of the best that can help. Still, there's much more to go. Whereas agriculture, at least you can see the plants. You can see the croppings, farming, carvers, etc. Suppose there's a building and it is covered on top with trees. For example, IIT Bombay, if you take a satellite or a drone image, you'll see a lot of trees. So you cannot count how many buildings it is. Because the buildings are under the trees. So the aerial imagery will not be just enough. So there's other data that is also needed, as I spoke about synergized data mapping. So after week 10, I hope to again revisit the synergized mapping and showcase some data that has been used widely for rural infrastructure mapping. So let us start with the NASA data sets. I will explain the Earth Explorer and also go to just DIC data set that we have already looked at in the previous lectures, just for the vegetation fraction we look at. So let me share the Earth Explorer website. So this is how the Earth Explorer website looks like. And we have already used these in our hands-on exercise. I haven't logged in, I'll keep it unlocked for now. So right now you can see that we have multiple options. So when you open a Google Earth Earth Explorer, I could open it again for you just in case. Let me open it again. So if I open the Earth Explorer again, it's the same link that I have shared. You will see that it opens on a particular location and in South Dakota because they want to center it in the US. So all you can do is if you move your mouse on the frame, you will see a hand. You click it, it will hold it. It's called pinch and then you pinch and then you move. So or you can, the best way to zoom out to a particular location, zoom out as much as possible and then just drag it. It's easier to go to India like this. So what happens here is we need to show where we would like to work on. So here we can zoom in by moving the mouse in front. Okay, so let me just put it back on Maharashtra because we have used Maharashtra for other NDVI indicators. So you can zoom in more if you need and then, yeah. So let's keep it at Pune region, Nasik region. Okay, so good. So we have this Pune and Nasik region and then as I said, what we'll be doing is there's a lat long that has already been given. We'll be using that, but more importantly, on the top, it's the search criteria. You can use a shape file that you have already used. You can download it and put it on the system and then use it, but also go to GeoCoder, which is kind of a little bit advanced. So let's skip that part. We will go to polygon. So you will draw the polygon where you want the area to be disclosed. So when I use this, you can use a map by clicking the previous map that we selected. But again, we'll use a new coordinates, okay? Or if you do a circle, you can click on top and then zoom out and zoom in, like for example, like this, and then you can put a radius and then the circle is created, let's say 1000 meters, that is one kilometer. So you can see a big circle coming up, right? Or you can clear the circle and then just sort of back, right? I'm going to clear all my polygons, coordinates, again, putting it back to Pune region, right? And then you can have a circle, predefined area. You can add a shape file after you log in, but we'll use a polygon. So once you click the polygon, you'll have to click coordinate. So this is one coordinate. Let's say we can just use Pune, this sort of Pune. You can also see the grids, right? So you can see these lines. These are each tile of the data. And it will be used for searching the data if needed. Okay, that will be enough. And then this is the coordinate system we have. Then you can come down here to say that, what is the cloud cover you are okay with? 100% cloud cover doesn't make sense. So let's keep it at 75%, okay? And then result options, you can see how many results you want to see. Let's say 10 is enough. And then the date range is, we'll go just for this recent year, okay? So or December, because we use December 2021 in the move-on, but we'll use December 2022. And then we will say, actually we can go to September also. September 1, 2, December 2022, okay? And then we can say search all months. You can say search all months. And then click on the result options. You can click all if needed. And it is also good, okay? Okay, this is good. Now we can go to data sets. So just pick what data set you want. So this is on the top also. And here you'll have Plethora of satellite remote sensing data that you could use. I'll just show you some of it, because just for NDVI, we'll come straight to NDVI, but I also wanted to explain this slide. Aerial imagery is just a photograph picture of the location. For example, you're looking at a post flood analysis and you want to see the impact, the damage of buildings and all, which is not an indicator-based approach. So for that, you can use these high resolution images, just as aerial images, okay? These are aerial images, not only taken by satellites, most of it is flights. So you can see here, flight imagery, Antarctic flight line maps. And then most importantly, these are all flights, whereas this is space photography from clouds. And then we have AVHR is also a different sensor placed on satellites, CEOs, legacy is there. Legacy means it's kind of outdated also. Commercial satellites, these are the two commercial satellites that Earth Explorer has bought for you or have a subscription. It's not the real, real high-end satellites where they have now iconos and orbital view. For example, high-end as in, they will not give it for free, okay? So the freer versions are the meta versions or the lower resolution versions. And then there is declassified data, which is something that it was classified once and now it has been declassified, some data on the borders and et cetera. There's a DEMs. So we have all these DEMs. We do not have the Indian satellites, but if you come down, you have your satellite. So no other big satellite name is there except NASA and the European satellite regions explicitly here. So you have RISO-SAT, both the AWIFS and LIS-3. These are good aerial imagery and a lot of analysis can be done using this. So you have digital maps, the National Atlas maps and then digital line graphs are there, Earth observation systems and then Fiduciels, global fiduciel maps, CMIs, all these are sensors. Then the land use, land cover. You can have a global land use cover, land cover trends, photos, et cetera. Landsat is the real important one because it has been a legacy. 1960s still date, it has been taking images. Now we have at Landsat eight and nine and you can get all these Landsat images. As I said, legacy is the older versions. You can see from 1984, 1960s, 1972s, et cetera, et cetera. You have data. 1960s, you won't get much of India, but you'll get across the other regions. So there's a collection level one. You can see that these are the collection level ones, Landsat one to five, four to five, seven, eight, nine. So the one to five is the older versions. You can click on this to get the collection info. It will open on a different page and tell you what these data Landsat one includes. And this says 1972 to 1992. So it's 20 years of data at 60 meters resolution. This is the oldest versions, very good versions I would say. And 1960s is kind of reconstructed data. It's not actual data, but it's still there. So that is 1972. So just let's look at how the resolution has changed in these earth explorer data sets. Now you will see the level collection two, okay. Landsat four to five. So the previous one was one to four, one to two. Okay, it's asking me to take a survey for life, but not now later I do it. So then we have the level collection two, which is Landsat four to five at 30 meter resolutions. So the previous one was at 60 meter resolutions, which was Landsat one to five. Then the four to five is at 30 meter resolutions. There's the four and five versions at 30 meter resolutions. And then this is also going to be the Landsat seven collection, which is really successful at 30 meter data, but it is multi-spectral. So the previous ones were just normally red, green and blue, whereas the multi-spectral data came into existence much later. And then we have the eight and nine, the recent ones. You can see them, the metadata for it. These are in some locations, very, very high resolution. And it has also the thermal infrared sensors and at 30 meter resolution. So the Landsat goes best for 30 meter resolutions, but the sensor has been updated. So now we have thermal infrared sensors and actually somewhere around bi-weekly to monthly, you get the data. So again, Landsat we will not be using for this part because we want products. We want products that are being taken from Landsat models, whatever it is. So we will go to the LC map. So these are two specialized maps and the NASA collections of DEM, MODIS. And then we have vegetation indices. If you click on the vegetation indices, you have the MODIS derived in the indicators for vegetation. And then we have the water reservoir, et cetera, et cetera. Eco stress, all these are related to rural entities, NASA, DEM, vegetation index, phenology. Phenology is mostly on the plant types and those kinds of things. And then we do have DB, IIRS collections, which also we will be using for our vegetation access. So you can see here, these are the vegetation mixes. Okay, and then the radar is more important for penetrations. So these have, it penetrates through the ground. So these mostly will have the soil moisture and land elevation data, much, much higher resolutions. UAS, unmanned systems are there, DEMs. So these are grown kind of images. And then we can see point cloud ortho. Okay, let's click this one. You could see that the unmanned systems also will carry drones, unmanned aircraft systems. So we have, these are high, high resolution and we have 2008 too present, but only small areas. Again, you cannot fly drones across the entire region. So you can see here, there are some taken in the New Mexico, which is New Mexico is not in Mexico, it is in the United States. So you'll have some of these data here. Okay, and these are the unmanned aerial vehicles, we call them or UAS. And then we have aircraft vehicles or systems also, they would say. So the A differs in how you use it. And then we have the meditation monitoring, which we will be using now, and we will be using the EVIS, DIR, NDVI, because we want NDVI. I'll just show you what is happening. So if you click on, let's say, yeah, NDVI, this one, it will say that it doesn't get updated or no longer produced after about 2022. So until then you can use it. So if you want to use the recent ones, don't use these data sets, but you can build a legacy of data. For example, from 1972, you can use Landsat data. And then from 1999's, 2000's models, and then from until 2022, you can use a particular models and then jump into Landsat again. So it's okay because the sensor is actually sensing the data. Okay, so we will close this. All these are kind of outdated except this one. So I'll just click that one. LST is Landsurface Temperature, which is important to show the stress on plants and land. So we have this, and then I'm just going to click Results. So we picked a date, we picked a date range, and we also picked the type of satellite that we want. And here is what we get. So we get 25 images for this particular area for one month. Right, I'm sorry, September to December. So it's around 15, 15 days a data set. So what are these is, this is a thumbnail to show the footprint of the data. So if you click it, it will show you that the tile, the tile and tile where the data has been collected. You can take it out and then go to this one to show the data set for that region. So I'm going to zoom in. So this is a pre-visualization so that you can look at the data before you download the data to make sure that it doesn't have errors or it doesn't have any issues with the resolutions or too much cloud cover, for example. It is still downloading. So that is why you would see the blurry image. So this is also good in terms of the satellite data. You can see that a lot of satellite data is there. And all these dates are there. So the end date, start date is there. So this is somewhere six to 15. So as I said, within every 15 days, the data comes in. So the start date was in September, it's a November six to November 15. So this is the November month of data. So you can compare between, not readily here, but you can compare in the previous region. So this is 2023. We did not give 2023, but it also populates it just for our need. And then here it is 2020 to eight and nine. So this is the last week of the analysis that we wanted to see. We have 25 images here. And so the entire map cannot be downloaded. That's what this is saying, but you can download only the maps that are available with this link. So if you want to download, you have to log in. So it'll ask you to log in, and then you can download this data. We have already showed you how to log in and download the data. So while it is getting resolution increased and stuff, we just pick one month. Okay, let's see if we do have January. So let's do January to Jan end. And then I'm just going to do the circle, apply 1,000 meters or one kilometer. You can change the units here, kilometers, miles, et cetera. So it applies that to the region. Now you see there's a lot of housing there. So I don't want just the housing. So let's say two kilometers radius. And then I apply, so it gets bigger. And then I have this cloud cover is okay for now. And then we can go to the results. We have to see the data set. Oh, EBRNDR is just clicked. And then results, there you go. You have all these results. And you can actually see them as a full tile. So you can see how, if you want to quickly look at it, we can look at within the month how it has changed. So from one to second February month we have. So you see that the entire India is almost green with NDVI, high NDVI in this basin. And then we can also see the previous results. So you can download this and if you need, you can go back to the clear the results or go to search criteria again. Let's just take the summer month, the previous summer, which is May, May 1 to July and then the data set results. You can see that now we have pushed the date to June, July, August, those terms. Okay, so if you look at May, which is the fifth month, these are the fifth month and then you put the NDVI on. So like this. So this one should have been capturing the image but there's a lot of black space, which means the data is not good. So please look at the data before you download it. So this one we can remove saying I don't want to do it because you will spend your memory and taking all the data. So now you can see here, all these yellow spots are not growing and the Ganges region is also not growing. So NDVI is very, very less. But when we go to the monsoon months, I'm just going to click this one. And now you can see all the green. What is the white? It is the cloud. So the cloud cover, if we have increased and said above 50% cloud cover do not show, then all the data, this time will not come because in my region Pune, there's a lot of cloud cover. You can see here, if I zoom in and if it is full of cloud cover, it will not take this image. So this particular image will not be showcased here. So you don't have to download this image and then work on it. So this is above the Earth Explorer and NDVI, readymade NDVI products. I also wanted to show you the other NASA product which is GEIS DIC. So GEIS DIC is also used for a lot of other data sets as we have seen in the rainfall, grace data can be taken from here, et cetera. But you can do the same as browse by catalog, okay? So you can say browse by catalog and say what data you want to use, measurement, temporal resolution. You just see how big this data base is, okay? So let's see measurement and see how many variables are coming. So all these can be taken from this database. You can take carbon monoxide, lanside, land use, land cover classifications. This just goes on and on. It's a really, really big extensive data set. But if you already know what you want, you can click NDVI and then you can pick a date range, okay? So you can pick a date or just leave it and you can pick a bounding box. Why is this important? So that you have an area of interest rather than downloading for the entire world. So I'm just going to click on the pencil and then draw a box. So I'm going to draw a box along India and then there it is, the bounding box for India has been kept. And then you just click it back and then say search. So when you do search, it will search and give you for your bounded region which is India, the box I clicked and then I drew a box. You just have to click on the pointer and then draw the box and then you will get these values. So here what you could see is two data sets are there. In this data collection, there's only two data sets for NDVI as marked as NDVI. It is the NASA's modest images and you could see the resolution is monthly and the spatial is one by one degree. So which is around 100 kilometers resolution. It's not that great but it has a long, long time series from 2000 to date and a lot of people have been using these indicators. So you can see here that the image, you can just click on this to see the full image just for verification process. It gives you the date, time. You can also download this image, save this image for your reports if you want. If you're working on a preliminary report quickly, just took what? Two seconds to download this image, right? So from here, you just said, okay, I want to see this image and take it for this particular month also July, 20, 10. So these can be used as a proposal writing in those kinds of images. Okay, so if you go back to data collections and then see what data is available, let's go back to GIS, DIC, the full website. These two also I'll give. Access GIS is mostly to use it with proprietary software. Visualized data will go to Giovanni. Giovanni is another dashboard within the Earth Data Explorer that is only used for visualizing the data and then making real time analysis, okay? So we will get into that pretty soon and then we'll have browse data at different spatial, temporal resolutions, project, et cetera. So if you also wanted to, as I said, I also wanted to say grace. So you have the grace data, you have different versions of grace. Let's see if the bounding box is the same. The bounding box has gone. You can actually type in the values here or you can draw the box again. So click on the box symbol and then say like this. It can only be as a box. You cannot put a India boundary and take it out. So it's normally 63, 4, 5, 87 and 40. So it's 100 actually. Normally I use 63, 500, and then 40. And then you can just say this one and then grace data is available. So you can see here, the earliest data available on this is 1920. It's reconstructed data, but it's still good. And then you can see grace data available. So groundwater and soil moisture conditions from grace is available. Time resolution seven days from 2003 to 2022 November. And there is a lag. Basically it's a model data. And then you also have per day GLAs estimates and grace estimates of data. So I can also click and showcase one of this data set and then show you how to access it. So since we started with NTV, I also will do, but since we have this, so the cloud enabled is where the data is stored. You can also have an online storage for your images. And then you have this grace data. So just a thumbnail to see how the data, the grins are present. And then groundwater storage percentile for August. So you have 100 percentile or point two based on the average values. So you can see more on the metadata here and then documentation data citations, et cetera. So this grace, we did not see all these, but I will be showing now for the NDVI. So we will do the NDVI again for the same box range. Date is fine, whatever date is fine. So we will say, okay, do you want one of these? I will say Terra Modis I'll be using. Okay, so here we have the global monthly grid data for Modis, a vegetation dyesers using these products. And then there is a resolution given here. Monthly temporal is monthly, spatial is one degree by one degree, which is good. And then we have data citations who you have to cite if you use the data. There is no pay. Most of people do not even cite these in the publications. It's good to cite it or at least cite the NASA team because they have processed this data, put this up and they're running it. So at the end of the day, they are not asking for money for using it. But if people use it, then the publications, they can show that so many people are using it so that this program can continue. The government looks at how many people are using it and the only proof they can show is publications. So if I can write a letter saying that I use it, but I don't publish it, then what does it use? So please cite it in your work. So you can use citations, you have it here. So some documentation of this indicators, same like we did, if you click it, there's a PDF which opens up about the satellite, about how it is being used, et cetera. There you go. So we have all the resolutions and then a big report on how this data was taken. The reflectance of red, percentage of reflectance and what it means. So if it is water, how much reflectance it is and then grass, how much reflectance it is. So you have cloud reflectance points in NIR spectrum from Landsat and different land use land power types. So you have different reflectances based on the land use land power types. So a lot of these are done and NDVI has been taken as this one, same thing, NIR minus red by NIR plus red, basically from the following equations. So if you want a theoretical knowledge about NDVI, you can look at this also. Then when you load it, please cite it. There's a lot of information that has been used about the satellites, where it has been placed, data calculations, estimations, those kind of things. References, again, you'll have multiple references for this data set and then data calendar. So the data calendar gives you on which month, which date the mission was taking data. And then it also shows you that there's any data gap because of instrumentation, et cetera, they will show you that in the data gaps if needed. Okay, so how do you access this data? That is another question people will do ask. So there are two types given in this particular data set. Not all data set have two, some may have four or one depending on the storage. You can click on first, open archive data. This is a folder kind of data, which means you will go here and just click, click, and then take the images out. Instead of downloading it from the, drawing the box and then taking it out, you can just go to the data set. So for example, here we know that this data is from 2000 to 2016. So when you did online archive, this comes up and it's open source, it's secure system, anyone can download it. How to download files from this HACI DPS service? You can read and understand. I'll just show you a quick demo. Here there's no login needed much you can see. So from 2000 to 2015, 2016, there is folders. So folders are kept for the data and you just click on a particular year. So let's say I'm going to go 2009, and when you click 2009, what happens is there is a read me file, before I'll just show you, there's a read me file, these things to show what this data is about, read me. So read me means it's a metadata about the data. So you can just click on this, it will open the modus, what these products are, file format, resolution, everything is given, which is like the metadata for you. And then it says SDS-1 means NDVI. So the product, if you want to download, so when you download this data, all of it will come out, but you want to only use the NDVI. So for which you say, I just want SDS-1. So all these are included, right? So if you go back, okay, I accidentally closed the entire thing, but I'll open it, open online data archive, and then you can see all the folders, I've clicked on 2009, and you can do the HDF. So that is the format, file format, gridded format, which is available. You can just download all of it if you want, but I'll just show you the convention, how it is given. Modus is the name, modvi, okay? And in the modvi, we have the 2009 01. 001 is the month. So you have the month given as January B. And then 005 is the level of the version, the version of the data. So maybe they would have added multiple criteria to clean the data, new algorithms, improvement, because they don't stop with version one. So they make it better every year. So we have version five. So all you could see is they go year, month, and date. Since modus is a monthly data, as we could see here, it is at monthly resolution. There's no point of putting a date. So they do not put a date, okay? So this is just the XML file. It'll just populate here, the XML file if you need it. But if you want this, just click on it. It will ask you to download. First you have to sign in, log in, and then you download, okay? So, and then you can just put it on your GIS database platform, and then you can model it. So this is the raw data that comes out. In the raw data, they have also made these documentation for NDVI data. In the next lecture, I will start with the visualization of GIS-DIC. So I'll just keep it here ready for you. You can also play with these two links, but we will start with the visualized data, and then we'll also clarify on Sentinel Hub, which is also a good, beautiful dataset platform that we can use, but mostly for European satellites. They do have NASA satellites, but they want to promote the Copernicus system. The Copernicus is a database for European satellites. So with this, I will see you in the next lecture, but feel free to go and look at these different data collection, how to do it, image gallery, and then mission guidelines, mission is the satellites, and then see the recent news on how these datasets have been used widely globally, et cetera. So we're talking about the highest downloaded dataset in the world. The most used datasets in the world are from NASA, and it is for all, okay? So a lot of collaborations are there between this dataset and a lot of countries, including India, and so please feel free to use it. And I do like the catalog level, and then also you can limit to what temporal resolution you want. If you want temporal resolution at even 99 minutes, these are model versions, don't worry about it. Three hours are there, one day, so I would say seven days onwards is really good, because every day taking a dataset is not important for rural development. Seven days to 15 days is good, and then 15 days monthly, annually, seasonally is there. So you can see from here. So you have monthly, these two are also monthly, then you have quarterly, then annual, six years, eight years, 36 years, day rule, okay? So you also have the project, which satellite you want to use? You can pick from satellite missions that are there, Landsat, Aqua, Discovery, satellite systems, et cetera, GLDIS, these are projects, and then the processing level. As I said, there are multiple versions and levels. The source of the data, these are the satellites themselves. So you have NASA, DROP, NASA, and then resource set won't be here, but it was in the other NASA webpage, right? And then we have the measurement. As I said, you can take the measurements. What do you want to measure? Soil infiguration is also part of your soils that we use, or subjects. Subjects, if you click vegetation, you will see what vegetations they have, and in the vegetation index you will have here. So what measurements they have are also here. All these are vegetation, and then you can see, you can find sort them here by, for example, let's say you can click on more, all the measurements come up. So I'll say vegetation cover, vegetation index, vegetation water content is what I need, and then I close this, then now only these are filtered. Initially there was 36, more than 36 data sets now it's reduced. Let me reduce it further by saying that resolution. Come down here, you have resolution, and I want only monthly. If I click monthly, only 14 data sets will come, right? Okay, so this is how you could reduce the number of data sets you want, and then filter it and use it for your analysis. So this I'll stop here. I've given you an introduction of this website, and how it could be used, what type of data. I'll see you in the next class. Thank you.