 Hello everyone, welcome to the NPTEL course on remote sensing and GIS for rural development. This is week 10, lecture 5. In this week, we have been looking at different platforms, remote sensing platforms and remote sensing derived data for NDVI. NDVI has been a very strong indicator for assessing the vegetation health and crop acreage. Number of times the land has been cropped or net irrigated area and it is at a very high resolution compared to the observation data both spatially and temporally. So, instead of getting data for once a year with a big, big lag time, here we are getting data within a month and at 15 days intervals. So, in the past we have looked at Bhuvan's data sources, NASA's data sources, Google Earth Engine's database and now we will look at one more aspect of NASA and then look into Sentinel hub. So, in the past as I said Bhuvan data is good for the Indian regions. It has some spatial and temporally issues in terms of resolution. It is limited to 2021 whereas we are in 2023. So, approximately a year and a half lag time can be noted. Whereas Google Earth Engine catalog gives you data for within a lag month, one month which is you can get also data from Jan and Feb how we are in March. March 2023 you can get it for Jan and Feb and NASA's USGS focus mostly on the NASA data that is available and we also have the Sentinel which is the European Space Agency ESA data set. I will go through the other part of NASA's visualization and analysis tool which is very important for understanding the groundwater and NDVI issues. So, without further ado I will open the GES disk webpage. So, we have completed the NASA Earth Explorer but while I was showing how to download data and access the data I also wanted to show some parts of the NDVI data that another portal can give which is the Giovanni. So, I will continue again with the Earth Explorer. So, I am going to click this and open the Earth Explorer page. So, in the Earth Explorer we said that we have all these data sets that are available for NDVI and then we can download it as marking and done. But in the GIS disk database what we noted is if we do NDVI and then put our bounding box on India, the default date can be given so you don't have to put a date you can just click and then these two data sets come up. In these two data sets we clicked on the first data set and went into online archive. So, the online archive as we said we looked at the different folders and within the folder there's multiple data sets. You wanted to know how to download it we can also read these links. Then you can also get data. It will ask you what type of data you want and then we will find the data range etc. So, it says download method is get original files and then you can define the range. So, let's just do a quick 2015 is the max they have for this type of data set. You can see that these are all blacked out there's no data. So, Jan 2016 you can have 30, 30th December you can have or 30th December that's fine. Okay, so we have the whole of 2015 data. It's just for the sake of this exercise we'll do it and the file format is only HDF. Okay, so this is the get data button you don't have to go through the download HTML and stuff. Okay, so if you say get data what it does is it runs and talks to the database in the US and brings all the data to this page. And you can see that the monthly data has been there. Instead of going to each folder and then clicking, you can also do this as one. So if you just click it, it will ask you to log in and then get the data. The login is to just show how much people use the data etc. The file list are valid for two days. Suppose you have slow internet and bandwidth, you can still have this page in your login. You can go to the login and download data sets and then this will be there. So the link has been created automatically for you. It says if you have already account create an account link this gsdispy account and then download. Download the list of links or how to get a widget widget is again an automated process for getting the data. It's kind of advanced. Let's not go into that but download the list links. So this is the list links that you can download and then it's a text document. It'll give you the link for downloading the data with your login. That is a download instruction and download list is there. So these are the list links and stuff and what are the parameters you wanted to select? You wanted the modest vegetation indices. So the both indices are there. The vegetation fraction NDVI is there and then the date range you have given as 259 to 201512. That is what it gives. You can refine and we download it and then get the data. So this is the one way of getting it. So three options are given online archive when you go to the folder you download it and get data is also there. The earth data search is also good. The earth data search opens the dashboard at once and in the dashboard it has already searched for you because it is linking to the GES disk database, this database. So what happens is from here it has linked the data to the earth data search box and that particular refinement is not there but you can still get it. So this is the data set and so there's one match for your data set which is NDVI and for your particular link. However, you could see that the date is still at 2000 to 2015. So you'll need to refine it. You can click on the metadata to see what the data is about. So this is the version 5 and then the data is given for the whole globe and then some other keywords. What is the format SDF format etc. So you can go back to your search and then say I want to download. So this is to add collection to the current project. You can add the collection and then it will be there in your account for download, my project, my download. Okay, so but before that let's also show how this can be used. I'm going to say I just want NDVI and then a lot of NDVI is coming and then let's say I'm going to press enter. It will show you what are the earth data search database tags with NDVI. So one thing which is clear is coming out is that there are multiple databases, same the same data or process data in online versions. All of these are open source. All you have to do is just click on a particular link and get the data if you have the login credentials. So here what we are seeing is there are features available from the cloud, custom handles or map imagery, instrumentations, platforms. Do you want airborne platforms or land based platforms based platforms, let's say airborne. And our base is mostly the hyperspectral from a drone and other imagery. And then very, very small lakes. For example, big trail lake is done using the drones and very, very small area coverage is there. Okay, so if I just take it out. So here you can see jet propeller unmanned vehicles is also there. See here, click it and then click it. You can see jet propeller uncrewed aerial vehicles. Or unmanned is also there. So the UA however they want they can use it. Unmanned aircraft vehicles or unmanned aerial vehicles. And here they say another term which is new also for me, which is the unmanned. Let me see if we can read it. It's not coming up. Let it come up a bit. Yeah, so it says uncrewed. So unmanned and uncrewed. People not there. That's all it says. Okay, so these are different platforms. I'm going to click also all the platforms can come up. So you see all platforms are there instrumentations which sensor with satellite sensor you need camera center multi-spectral hyperspectral and all the systems and then organization is NASA. NASA is the Oak Ridge National Laboratory processing levels. And then projects if it's particular projects, NASA's project data format. All the formats are openable in QGIS. So that is the beauty of using the open source system QGIS. All formats can be opened or converted. So you can really convert these formats in QGIS. I haven't seen that strength in proprietary softwares yet. Latencies how how delay you want the data 123 hours is there and then you can use it. So for example, you can click the 16 day grid data, just click on it. It will show you how many grids are there for every 16 days. And then you can download every data set that you want you don't have to download all the 1400 data sets because this is 15 days. And then sometimes it'll double duplicate data. So this is from 2023 to forever. So 2020 it's still going on. Right. So what we can do is you can actually filter by date. So here you can put the date. Let's say I'll put November 1st to December 21 December 31 and then apply. And so now it's going to get recess. So now from 148,000 to now 42 collections, which is good manageable. You can download each one. So you just add it to your download page or you can just quickly download it after you have the populating the link. And then it will let you download the file. So as I said, you can download the file after you have the login. And that is what it's asking your login passwords and stuff. So you do have to have an account. So make sure you create one account pretty soon for this database. So there's a lot of data that can be taken up for NDVI. And one more thing I would like to show is ready made a product where you can do some quick analysis. Okay, so let's do visualize data. If you click on visualize data in GIS this thing it opens as Giovanni. You can also go to Giovanni just this link it will take you there. Giovanni is an application. Okay, let's see if we can redo that again. Because there is a note which comes up says it what it is so this application allows to visualize the parameters and you can have a help page to look at how to do it. I'll quickly do this but you can definitely read it for your future understanding and then you have to have a login. Okay, so let me close this here. You can hear this is the type of analysis you can do. You can do a time average map, time average overlay map, map accumulated comparison between two maps, selected area average time series, average, etc, etc. So this is a analysis type. So let's say time average map or time series. Let's say time series, time series area average. Yeah. And then we're going to say that I'm going to have 2022 just to make sure the data is available. We'll just say first jam to you don't have to put the time time is not needed to 2022 December December 31, and we're going to say the bounding box I said the numbers for India if you don't remember it. It's okay we can just use the box and then click the box, the box symbol is different here. So you can just say it's around 64, 63, 40, 100, and four. Right, so 64, 63, 45, 90, which is 100, and then 40. So you can click this box again. And this is a set of shapefile if you have a shapefile I said for India Maharashtra boundary you can add it here. And then you can do this and here as I said you can type in dbi, and it has come up in dbi, and then it says search. So when you click search, all this will come into account so from one year time period for the India platform, this data is coming up. And it has monthly resolutions of 0.05 degrees 2002 to Jan as I said clearly, Jan fed data is still getting populated. So which is very, very good. So we are now in March, mid March, so within one and a half months you have the data, whereas you don't have to wait one and a half years in other data sets. Right, so you could is 12 months exceeds the maximum number four can only be processed. So we'll have to reduce it, we can reduce it as monthly so monthly is 12. So it says 12 months you have selected we can only do four months. So let us reduce this time now from, let's say June, June, May, May, May to, so that is a monsoon period to May, June, July, August, so we can do August, right. So four months, and then it populates the same thing again, and now it says okay guest limit we are a guest, so we can only do this much. And if you have an account you get more. So this is what I'll be clicking on to understand what this data is. Okay, so if you want to read the metadata for the data you can click the link, it will open here. And you do have a good analysis of vegetation derived from a model, and then multiple types of vegetation indexes, you have greenest fraction leaf area index climatology, etc. Okay, so once we have this begin date end date you can say plot data, the plot data will plot the data as I said we needed time series area averaged. So it is now running you can see it is running, it is not using your computer's memory or software. Even the code so when you click the boxes automatically the code is developed and sent to the NASA's computers, and there it is actually processing so this processing is there. So your internet speed and your memory power does not get affected. So here you are analysis for India scale it just took 10 seconds. Maybe my internet was fast the uploading and downloading is fast I just check the upload download speed before we started this class minus pretty good. But in a normal situation using your mobile internet Wi-Fi, you can get this within a minute. Okay, so you can see here from May, the NDVI is increasing. That is what it says, the unit less NDVI doesn't have units, and it starts from minus one to plus one, the time series average of NDVI for the entire India at 0.05 degree resolution. Everything is given here, you don't even have to type this in your reports, but you did it because you actually plotted it the boundary, and you said this is the time series, this is the NDVI I want to see. And this can be done, this Giovanni can be used for multiple, multiple parameters, not only NDVI. I'm only showing NDVI because this week is linked to NDVI. So I'll be showing this. So you have different options, you can show the title show the caption or remove it if you want. But let's keep the title, let's keep the caption which is here the user edition was defined as this is this, which is India. If you can take the latlongs of this and put it on the latlong calculators, you can see it is India, and then the visualization results. You can download this as an image, the data can come as an input in image, you could see that the time series average area average output was taken. So for the entire India, all the pixels to do this, it takes a lot of computing power, but you have done it within a couple of seconds because of these supercomputers linked to NASA. So, beautifully, you can explain that oh, from 1st May, 1st June, July, and August, the NDVI is increasing, that is because most of the monsoon happens here. So in Maharashtra, the monsoon onset is June, let's say June 1, and then June 6, the first week, so June 6 normally it comes. So you can see that after that it peaks starts to peak up, and now the plant is growing healthy well. So let's do another one. Let's do Jan, Feb, March, April to see how it comes back down. And then you can say back to data selection, and here you can go to, let's say, Feb, 2, 5, and then I'm just going to go to results, we'll go to the original results, I'm going to plot data. So when you do plot data, so the previous exercises here, you can see that input plus download in lineage, everything is there, but we'll keep it out for now. This is the history, and now our other file is downloading, getting access, successfully ran the time series average for the entire India within 10, 15 seconds. And now we have this. So can you understand this now what's happening is, initially the winter crops were slightly growing, there was irrigation happening. And after irrigation, it plummets, it shoots down because of people harvesting it, and the summer kicks in in March. So you see from February, March, April, May, so may slightly there's another round of irrigation, because a lot of people do some groundwater irrigation subsistence farming, etc. And there's some other summer monsoons in some region. To reduce this, what you do is go back to data selection, you can go up and just put a bounding box near, you can zoom in. Okay, so let's zoom in. Let's take this box out. You pick this hand, and now you can move. Let's say you want this Maharashtra region, this part, this style is enough. Let's say this style is enough. So I just draw a box here. Okay, and then I close this map, and NDVI is fine. The data range is that. And then let's say. So this is the third time series we're doing. That's why it's saying three. Again, this I'm not using GIS, but when you download this time series, this is an analysis, this is a plot, you can actually put it in your reports. If you want to do it from scratch, you will have to download it. So if you don't take Maharashtra region, you see how it goes really down. So for my region, I know there is no summer monsoon. There is only really drought at that time. So it just goes down. There is no, even irrigation doesn't happen because there's no water, groundwater is going down also. So when you take the entire India, it's a different ballgame. Like for example, this one, the entire India is different. Let's click this one for analysis. So we have the same time frame for entire India, it starts at 0.33 and then comes down and then goes up. Whereas for Maharashtra region, it is just going down. This is purely because we do not have a monsoon in the summer. It comes only in June. So until then it's just really, really dry region in that part of the world. So you can download the data, etc. You can change the type of results you analysis you want. If you don't want time series average, you can see a time series map. So now what you're going to see is a map for that area. I'm going to say plot data. So this is not a time series data. This is not points, but you're going to see a map takes a little bit more time because you are launching. So look at what it's saying. It's launching the work, attaching the data files, the cache from here to there and then doing the time average map. It takes a little bit more time because now you're going to see a map that wants to come up successfully and the image. Now it's computed visualization is being created. So this is the GIS step, right? You'll do the analysis, you do the visualization and then you plot it. And all of this is done for you automatically. There you go. So this is the region we said and in this region. So the entire thing is average to one value in this time series. This is an average of this. But now if you see it is a every pixel in count, right? So every pixel is taken. You can zoom in to see the pixels and where the green color is. You can download this as a geotip. In the previous download, you only see image, but here geotip, KMZ, PNG, net CDM. All these are usable in GIS as a raster. So now it's a raster data. You can download it and then put it up in your database. So again, it lasts for your online link for downloading the data. But make sure you have the link already. When you start, just log in and start. So this is about Giovanni. As I said, you can use multiple, multiple data set. All this will be stored in the cache memory. And once you go to a new page, all this will be deleted. So all the history will be deleted. And this is really cool analysis that you can do within a couple of seconds. And the units would be different here. You can see that 0.3921 whereas here the units are different. But again, as I said, there's a scaling which needs to be done. You'll have to look at the data. What is the range and then scale it. So the value is minus one to plus one. It cannot go above and beyond that. So this is a scaling that they have used. So I would recommend using the Giovanni. You can go back to data selection. You can collect different disciplines and then do the same thing. Not NDVI. You can save a station fraction. You can say measurements, what type of measurements you want. Let's say soil moisture. You can do soil moisture also. Soil moisture, soil moisture is there. So if you know soil moisture is very high, you don't have to irrigate. So that is the understanding of soil moisture. And for that same location, we can say that this soil moisture is percentage at 25 kilometers, pretty large compared to the NDVI. And then it'll give you at different depths also. So you can say at 0 to 4, 40 to 100 centimeters depth. So 0 to 10, 10 to 40, 40 to 100, 100 to 200. So 200 centimeters is divided into four data sets and then given. This is also meter cube by meter cube. So it's kind of, you can say unitless, but normally people express it as meter cube by meter cube. Let's say 0 to 10 percent, which shows the initial part. Okay, this is daily, right? So you can see here that the soil moisture data that we clicked is daily. And it says only four days you can click. So let's see what data range they have. You can see that the most recent one is 2016 for that data. Is that correct? Yeah, it has 1979 to 2016. So it stops at 2016. So we cannot use, if you want, you can use it. But I would go for more 2023 data so that we can have some soil moisture. So there's a soil moisture 2011. You can limit the data here in temporal resolutions. Monthly special resolutions, etc. Right. So these are 24 to 2022. So 2022, yeah, so this, this we can do 2022 November is available. Let's click this one and click the other one. Yeah. And then let's say date cannot have so much date. So we'll have to say two and then only daily, right? So daily, you can say, these are the two data sets and NDVI and taking out. So yeah, let's pick a date and plot data. It would plot the data within February, whatever data is available. And then plot the maps for that particular thing. So it says scanning data for that particular data set. Giovanni is very, very important for understanding different sectors. So you could see that this is 100 to 200 centimeters underground soil moisture value at 25 by 25 kilometer grid. So this is pretty big, but it's pretty useful because these are based on ground penetration data. So even if we have this, you can tell the district collectors that if there is need to be water released or groundwater is going to be used can be told by these images. So with this, I will stop the Giovanni exercise and NASA exercise from US level. Now let's go to Sentinel hub as we discussed in the slide. So this is the link that we will be using the Sentinel hub. So we're going to go to the Sentinel hub now by opening a new tab. So from here we will Sentinel hub. Sentinel hub.com, which I've given the link in the presentation. So it says a cloud API for satellite imagery and you can use for explore hub request a trial. You can go to the explore the hub part. So here is the what the data is about. I would like you to take you to the EO observer, which is the Earth observation browser and all these data sets are there, what data is available, etc. We'll just jump into the data section and you can see that there's Sentinel, which is European Landsat NASA, commercial collections, DEM, Copernicus again, a service database from Europe. Modis, MSAT, US, etc. So you can also bring your own data and then do it. So we'll go back and then launch the Explorer. EO Explorer is what we need to open. Please note the browser extensions, etc. So you can accept the, I will not use the tutorial. So just go here and then use it. Okay, so here we have already done a couple of exercises, but let's see what we want. You can you can say that you want, these are the different indicators you have your agriculture vegetation. Let's go to vegetation. In the vegetation, you see certain data sources that are existing and up to 2023 19 March, which is kind of today where we are doing the range. We can say February to March is good. And that is the data that is available. So you can say that you can say advanced search for Sentinel, etc. Or we can say agriculture and then agriculture is only Sentinel to and then we can get this data. So commercial data is also available. If you want, you can sign up and then do it. We have to pay for some of it. One of the highlights is what what agriculture regions then recent news articles they have written published using this data. So let's go to search and then this data find you will just search. Let's see how much data we have. So this is 19th of March we are on and you can see this is just three days ago this image was taken. And you can already access it here. So this is very, very interesting and cool. Okay, before that I should have done Pune just to keep it in Pune, Maharashtra. Yep. So we are in Pune. So you can see Pune is coming when you click Pune Maharashtra. Okay, so Pune region. Okay, let's say songly songly why because it has a lot of sugar cane. So you can see the DVI for the sugar cane. Right. So now, now we're in songly. Okay, so you can see that these are the other tools that are available. You can also do the plot as we did in Giovanni. You can plot the data but first we have to select the layer. So here we have 26 data sets based on the data date and look at this you have already cloud covers or you use it we don't want to use all these cloud cover data. So normally the best data sets are coming on the top. So we will go to back to search and then search now for Pune songly region. So now the data is getting updated because songly we didn't type initially so now songly and we have around only five results. And this is good. This is really good because March 10th this data has been taken so let's do visualize. And so the Sentinel to data has been visualized against Sentinel to data has multiple bands. Okay. And this true color is made up only the three bands for three and two, which is giving you the true color. And if you want we can look at what are the bands Sentinel to bands, and then you will see a list of the bands in Sentinel to that are available. And this is good because that is where the European Space Agency is having. And then you have all these Sentinel to has 10 spatial resolution bands B2B3B4 and B8, etc. And then 20 meter spatial resolutions these are the other 20 meter special. So some some bands are high resolution some bands are low resolution. Another special resolution is B2B3B4B8 B8 is kind of your red will check what it is, and then yours to meet 20 meters is multiple bands and then you have these bands also. So you can see here it is given here as B8 is if we can zoom in that would be great. So B8 is here B3B4B8 so B8 is in the red visible near red near infrared along the red side, and then B2 is your blue. And then all these are visible so visible is what we can see with your the with your colors we can see, and then the B3 is your green light green you have and then the before is your orange kind of red and then the B8 we cannot see it by human eye, and that is your red color. So this is mixed in the composite. So now if we go to the Sentinel hub it says 432. So the 432 is 432 are mixed so red, blue, green are mixed, and then the primary colors are mixed to make this image which is the true color. The false color is 843. So 843 is a false color it uses the infrared region data. So eight data is red which is visible near infrared some part of it is visible, and that is why you could see the red, it gives you the growing period growing color etc. So let's go to true color you can see all these land parcels sangly is very very known for sugarcane, so you'll see a lot of sugarcane and sangly coming up soon, and then the indicators so these are the indicators, you have NDVI, EVI enhanced vegetation index, the normal rise difference in the vegetation index, and then the moisture stress agriculture is B11, B8, B2. So you can see where the agriculture is happening along the area. And then you have savvy soil adjusted vegetation index. So just by clicking the NDVI just populates the metrics the indicators just populate. So that's the beauty of using the Sentinel hub. So every platform has its own, you know, use and benefits because they don't want to redo what others are doing. We are more interested in NDVI so beautifully NDVI comes and for that particular date which is just 15 days that we are looking at. So let's look at the dates so this is 2023 10, nine days ago, not even two weeks, and you can also say how we can compare. So this is March, and you could see that there's a lot of parcels of land that is being under cultivation. I've been there so I've seen a lot of sugarcane, literally a lot of sugarcane being harvested. So one thing we can do is beautifully we can add to the compare. So I'm adding so it adds to compare here also you can do a login, but to visualize as a guest it's okay you don't have to do all these things. So we can go back to search. So one one one data set we have created, you go to discover you to take the other data sets. And then this is in February. What we could do is we could re-align our back to search, and then say maybe pick a previous date in 2022. Remember, and then say search, you can say search for the same dates. Number 15, right, so say search. And then you get the number months at the bottom. Okay, so you have the number months. You can sort it by date or the best data. So this data is good. So all these data have some issues cloud cover, some white is there we don't want that we can use this color. Okay, so clean image, you can say visualize the same visualization comes up. Let's say in dbi and then the dbi gets populated. And I'm going to add it to the compare. So compare is we want to compare live the two and dbi data visually, and then we can download it if we want to download so free to sign up so that you can download all these data sets. So here we have the compare and let's say this left side, what I'm going to do is I'm going to keep the left side image. So this part of my computer is going to be my in dbi from November. And this part is going to be my in dbi from current date, March. So you could see that the the data set is converting more to green, because during November you still have a lot of moisture in the ground soil moisture that contributes to agriculture, and that can be used widely for the sugar cane, whereas ground water recharge aside, March is almost setting into the summer. So you will have less vegetation growing and that is what this color difference is saying. Remember, this is the same bands, we are not changing the bands, we are taking the end dbi. Let's go here to see what this end dbi is about, you can see here it is a normalized vegetation index range minus one to one, it gives you these these colors. And it has been more info is here, b8, b4, b8, b4. So b8 is your visible near infrared, so NIR, let's say NIR minus red, b4 is red, it's not orange, it's red color. So b8 minus b4 by b8 plus b4, that is what the equation we've shown in class you can see here the equation given. So in the compare we have seen beautifully the two data sets being compared. And once you have the education mode can also be turned on where you have better access to some data and theme also can be done. So I think this is normal mode. So in the compare we have this data set, and you can draw in the search, in the search ones you can have. Yeah, in the search ones you can draw and then see where how big the polygon is instead of saying Pune, Sangly you can actually draw and that box can go in so it's kind of like a bonding box. You can upload a shape file to do it same like Giovanni, you can upload a shape file and download the data. And then you can have a point of interest, and then other resources you can also measure and download the image, you can download the image without much problem. So you can say you want overlay maps and article you want an article map is kind of only the basic image so you don't have a higher sprint also you won't have, but at least you can download the image and then do it in particular formats. You can also animate all these you have to log in, do a 3D map, analyze the histograms, colors, etc. It's pretty cool in terms of using a free open source system without login I've done to show some people might have some issues in logging in, but it's pretty safe I do have login accounts, I'll just show you how it is done. And now all these come up right so you can say that you can have just this area I want the, you know, you can measure the area for a plot, say okay is 637 meters per meter, 0.05 kilometer square. You can now download some of the images, geo reference or PNG no geo reference is there. You can also do it with geo referencing that we have taught in class. Some analytics cannot be done. Okay, let's say visualize discover and go back to discover. Go back to search, go back here, go back to NDVI. And now the analytics can be done. So this is basically the area I wanted and it's just calculating the analysis. So you can see here, the green values are here, and you say that is more green, right. So now if I use a different data set from my previous example, say, or you can say just 10, and then let it populate. You will have a better histogram because as in populated. Okay, these are in the Jan month, let's see one in the Jan, and then the population comes up. So you can see now a better high green number of greens, because it is good for so the other side. So, so 6 is green, we'll have a lot of green because there's a lot of NDVI. So this I'll stop. I will see you in the next class with some more indicators. Thank you.