 Hello everyone, welcome to the NPTEL course on remote sensing and GIS for rural development. This is week 11, lecture 4. In this week's series, we have been looking at the concept of synergized mapping for creating multiple data sources with remote sensing data for an updated rural data management. In the last lecture, we have looked at using OSM data and remote sensing data on GIS platform which is QGIS for rural infrastructure mapping, especially schools. In today's lecture, we will start with hospitals and also try to map some more aspects of hospitals which are road connectivity and other aspects. So we will go through the same steps. We will extract by name, extract by boundary which we did in the last time. Since it is going to be hospitals, it will be much lesser than the current setting. So I will use a smaller state for an example. We will select the state and make a layer out of it. We have yet to do the cropping, but I can quickly show the cropping along with hospitals or roads and then we can also search by a village name, as I said, Noni village, for example. So then we have looked at how to check in Google Earth Pro and then as I indicated, there is a lot of internet. So without further ado, I am going to do the hospitals this today thing in healthcare. So let me share the screen for your QGIS. So we can remove these from the previous exercise because let's keep the chain, but we can remove the schools, database and the amenities database, let's remove them for now. And then keep the states, zoom out, we have the states. And as I said, they keep on updating. So for now today, we're going to look at the hospitals. And as I said, we could pick Chattisgarh, which is a smaller state to quickly download the data and use it. Using an older map so that we have long-time studies, for example, Andhra Thelmana was already divided now. But we'll be using this map because for the long-time series, Thelmana was not there a couple of years ago. So we'll use this data set so that we can have all of Andhra to get, then we can differentiate it on a new shape file. So let's start with identification of schools. As needed, we'll go to the vector, QSM, click on the QSM toolbar, let it populate. We can say Chattisgarh or we'll select the layer first. So just let's minimize it and then quickly select this Chattisgarh for us. For some reason, it is reselecting all the layers and then let's collect Chattisgarh, because initially the districts was selected. Now Chattisgarh is selected. What we can do is keep it selected. We don't need to make a shape file, we'll do that in the QGIS platform itself. So now let me share the full screen so that you will see my full pointer of, you know, this is the attribute table for the districts. I'm going to close it. So you can see that in the attribute table of the states, Chattisgarh has been selected. Now we will go back to vector QSM. So now we see it. We will not be using the presets because presets have been kind of not capturing the entire aspect. So let us say amenity first because all amenities can come and I'm going to take hospitals. If you just start clicking, it will come or I would normally go to read all the lists so that we know what data is available. For example, in rural development, we need to know how about the animal boarding, breeding centers, shelters, banks, rural banks are very important for rural development. You can map it. And I would say this is the most comprehensive data set available because it's coming down up, not bottom down, it comes from down up. Child care you don't find in rural regions, but you can find in urban regions. Then you have colleges, driving schools, internet cafes, nursing homes, parking. I'm just reading out some things which could be important for rural place of worship. Photo booth and then recycling hospitals. We want youth centers. So let's go to hospitals and let's see if we have health care also, EHI. So it's arranged in, yeah, there you are. So hospital is there. So we've clicked hospital, right? So let's add one more layer and say if health care is the amenity health, there's no health care, but we can see some other themes. What other themes could be there? We could have, yes, we could have government, right? Health care comes up and then we can also have government hospitals. There's no hospitals. There's no rural. So it keeps on updating. So it's always good to check what else we have. We have hospitals. We have health care and then there's nothing for rural, but we can say government and see if there's any rural entities. There's no rural. So I think we'll stop here. We'll remove this. We'll just say or, okay? And we'll say that we need a layer extent, full states and Chattisgarh. So I'm going to run this just for Chattisgarh, as I said. And then go to the advanced. We may not need the lines, nodes and multi polygons is good enough. And this will run throughout to see which ones we have. So it is good to run twice to see if there is multiple data sets. It will tell you that one layer has been added, which is the amenity hospitals. Now I'm going to do is see if there is amenity hospitals, run query, no ISM objects selected, please select one. So we don't have the object. So let's say node and multi polygons and then it is running on top of it. So mostly if you know the point data, you can estimate the polygon by drawing around it and go to the earth. There's no result. So this also can happen. You'll have a successful query, but no result, which means that yes, you have created nodes, you may have nodes around the area, but for Chattisgarh it is not there. I will not be surprised for reasons that it is a lot of forest area. So let us just click it down and you could see that lot of Chattisgarh around Chattisgarh. It's like a tile. It's like a tile, the data has been there. We don't know how good the data is, but we will surely find out by zooming in. So I'm going to just use my pan. So you could see that these data would contribute to Chattisgarh. We can mask or clip the data out so that we can see what data is there. But as usual, we could plot it. So I am going to do the properties. So in the properties, which I didn't explain in the previous slide for the previous lecture, which is lecture three, we did schools. So in the schools, I did not go into the metadata, but I thought I'll explain it here because it is the same for all. So in the properties, if you click, then it says source a lot of things, a lot of links for the source. But more importantly, you have the extent, what type of data it is. And then the accuracy based at two meters, best accuracy, the coordinate system is good, and this is the more thing. So the license is being given to OpenStreetMaps and the contributors. As I said initially, the access is free, but it's good to assign these people because they have put a lot of time on creating these data sets and map and also maintaining the service. You can see that what are the fields that are there and what type are there. All our strings, there's no numbers because it's just going to be named. If you want areas that we will do later, you will have that also. The source can also show you what is the source that you'd have. It's healthcare systems and you can also have symbologies if you want to change. These are the locals, we can close it. But if you go to the open attribute table, we'll have the fields that we have there. And specifically for hospitals, there's different types of field names. A wheelchair, image, locality, name, city, you are. Veg city, name, geometry, contact address, email, phone number. Website is there for some people, some businessmen would like to have a website. What type of health center is there? This is a primary health care center, PhDs, PSUs. They call for the rural areas. And then there's a district hospital, district main hospital. And then there's a government, non-government hospital. All these are government, nursing is private maybe and the state. So here we wanted more, at least they are not Orisha, Andhra. So these can be removed or just I'll show you if I just click this and then say like this, you have all the, from upside down. I would expect more null to be there. So that's why I went down. So yeah, you have null little. So Andhra Pradesh is there and hopefully we'll jump into Chattisgarh. Yes. So I'm just going to click on this and then drag. Okay, so I click on the Chittar Chattisgarh and then drag. So that you're highlighting only the Chattisgarh region. We can also build a query for Chattisgarh. I'm going to use control. You can also use shift to just select all of them. Let's go down to the end of Chattisgarh. And then hold on shift and then click. So now what has happened is we have selected all the Chattisgarh. You can make a shape a lot of it. And you can see that if I remove this. Clear selection for, just remove this part. You can see that the Chattisgarh selected as within the Chattisgarh area. Correct. So you can have within the Chattisgarh area, you have selected all the layers, all the hospital locations within Chattisgarh, which have Chattisgarh. The others are errors, but they have somehow creeped into that because of the buffer region also. And maybe the locations when they specified somewhere, they would have put Chattisgarh. The open source contributors. It's okay. As I said, there are some hospitals inside also, which without Chattisgarh name, this could be the null ones where you can add the names. But the best way to check it is by using remote sensing. So now we will export this as hospital data. Let's say export. Do you want all features or selected features? I'm going to export all, just so that we can see where they are located. And we're going to say Chattisgarh, smaller name, CH, underscore hospitals. You can also write OSM. Most places in the names as GIS rules use underscore and no special chemicals. You can save this and add the save file to the map. And you can remove the temporary file. This is a good thing about QGIS. It allows you to create a memory in a temporary file so that you don't waste your folders and the space. If it doesn't work, if you don't like the data, you can just remove it. Okay. So I can say, yes, I want to remove the other one. We have this part. Now we will open our Google Earth again. And we are going to add these are the previous ones of the schools. And now we're going to add the hospitals from the different areas. Open, we have OSM, we have CH, okay. Do you want to adjust the style? Let's do once. Okay, let's say, yes, I'll show you why this won't be that, you know. Do you want to use a single color, use random colors? You can say use random colors, icon. Do you want a same icon or different icon? You can have different icons here. You can pick a hospital for the icons. Let's say there are multiple icons. You can always change it later. But yeah, let's put a house, yeah. You can have this for normally a hotel, but we can just use it. And then height is, you can just keep it as same. Okay, so name is there, I can say okay. And you will see, do you want to save it? You can say save or cancel. It will start plotting it in Chattisgarh. So moving from Trichy to the state of Chattisgarh, you will not see the boundaries yet because I have not let the boundaries to come up. Now I click the boundaries borders, it has come up. Again, go to view, make sure the tilt is okay, reset the tilt. And then click on this to see the hospitals. So beautifully, all the hospitals have been marked. So one thing which comes out lately is this region where the most forested region are very, very less number of hospitals. And so we won't be, we will be needing more here because of the rural population and the rural livelihood options that they have here. They have better livelihoods and options they do need hospitals. So we can go here, kind of a bigger city which has a lot of, I'm just randomly selecting. There's nothing that we have in mind in terms of specific areas of interest. You can see that how the hospitals have been mapped. Okay, so let's see if that hospital is correct. We can just click on the field that was created. I've clicked on the bed and you can see it's a hospital node, hospital address, Chattisgarh, address full address. And the name is Sri Krishna Hospital. Let's see, it says Sri Krishna Hospital and Vaccination. So there is a very good association between the accuracy and the names when it comes to commercial places. I say commercial because these are more into the rim of payment services. This is a vaccination center also, maybe during COVID they had a good access to these kind of things. So you can see also another hospital here. It's Narayani Multispeciality Hospital. If you click on this, the name is also Narayani. So these two are not same data. This is Google Earth Pro data. So kind of an OSM data, but as I said in rural regions, this is not being updated. So which I will be showing now. So the city coverage is pretty good, which you can use and which is also needed for rural development because mostly they go to cities for a better access of hospitals. So let me zoom to the layer or zoom out. You can save this as a layer. Just double click zoom out. Then go to view, reset the tilt. Okay, this area, let's focus on the bottom part where as I said, there is a need for more hospitals. Along the city boundaries and stuff. So if you zoom in, you could see that here, the names and borders may be populating, but yeah, this has not been populated. So which means this is Assistanti Baban, the name. It's a mission hospital in Chattisgarh, but there is no data on Google Earth Pro, which means that it has not been updated. So this dataset, which we are using from OSM and mixing it with Google Earth Pro is the best dataset for now for the assessing the health locations. So as we said, there is a need for more and more of this data to come up because if they are underrepresented and not being mapped, how do you map it? Now, so you are a researcher, you want to use remote sensing data for mapping, which is good. So you can actually use these OSM to update the data and put it into this network. So you can see here, this city is a district hospital. So normally this district hospital would be mapped in Google Earth Pro. It gives you, say it's a big hospital, you can see it says district hospital, Bichapur, Chattisgarh, and that will be the same name here. It says just named district hospital, but the city is Bichapur, right? So you could actually get better of this data, update the OSM. So it's not only the Google Earth Pro data can be updated with OSM, the OSM data can also learn from Mix and Match. So your goal is to keep your GIS layer up to date by collecting as much as information as possible, the local languages are there and other languages are there, you could add it. Good. So I will also be happy to do one more round of using the road network because road network is very, very important, we said. And I'll show you how this is also important. So for example, we did the schools, let me go to the Tamil Nadu schools again, or Trichy schools. We'll go to Trichy schools because we'll be happy to, yeah, so these are the Trichy government schools and the private schools that we mapped and how do you know that a school is village, rural entry or not? If it is mostly in these kind of areas where we have a lot of buildings, then that is a city. So it's in a very verbal encounter, we'll say, okay, it's a city or not. And this is a big school. And then we can say it's about Marshall RCA high secondary school. So it is not behind there. There's a lot of rural land, but for sure it's on the highway. So it is a big city or all the way to the city. So like these visual stretches you could look at and then see if a school is government or aided those kinds of things. So this is a big school. You can see it's definitely a school, Holy Cross girls high secondary school. And this is kind of a government aided school. You could see that there's railway quarters and that is why it's a very housing area. What is good here is you can draw a buffer. So for example, you need to know how many houses are around so that they can walk or commute to the schools. That can be done using your Google Earth Pro or create buffers in your panel. I'll show both. So first year, let's say that you need to do a path, assessment of path. You can say circle, okay. So I did the scale and then I went to circle and then wherever you click, it will go abroad. So like this. So let's say I want 200, 300 kilometers. Okay, one kilometer is kind of too big. Okay, so this is 501, one kilometer is this big. You see how many schools are covered for houses in this region, but you can make it small because one kilometer of travel would take some time for them. So let's say 200 meters. Within 200 meters, where can we have schools? So you can just bring it down and then zoom in to see if we are on the ballpark. You can also give 200. Okay, so 200 approximately. So within this, you can see that the housings are very low in number. And this kind of buffering is needed for school locations and having access to school, like in terms of roads, networks, buses for rural regions and stuff. And also the government in most parts have banned on certain type of shops in these kind of radiuses. Like you cannot sell illicit liquor and other things near a school. So there is a radius. So this is how the radius is calculated. You have the location of the school and then you make a circle of radius. So within this, it is more important. Why we are concerned is we would like to see if these are more into the housings and are the housings being located within this area and are the housings helpful? Okay, so this is one way of doing it. You can actually put a point and then blow it up and then see if the schools are coming up. We will also see Trichy schools or Tamil Nadu schools polygon. You could see that number of polygons are very less because not a lot of these polygons are going to be mapped. You could see that careful drawing is not done for most of the parts. So you could see this one. Okay, so if I remove or make this polygon, it says SB, IOA, multiplication and high secondary school. You can see most of the city here. Only some part of the village you can see here. So it's not that big. And also this one is St. George high secondary school. It is not mapped in OSM. So you can actually update the OSM. As I said, my village school is not also mapped and people can map this and provide data with this. So all the 2000 plus students are enrolled in this NPTEL course. Think about all 2000 contributing at least 10 schools wherever they are from. Okay, so across India, our students are attending this course. I hope and I aim that all these students do contribute to mapping schools because that would be a really, really good database. So there is a database on paper, okay? With geo locations, current geo locations and area, maybe it's not there. So this kind of mapping you could actually do and provide the policy makers for rural development and infrastructure development. Without knowing the size, the status of schools, we cannot have much benefit. Let's give an example. We may be able to take a smaller school just for the sake of an example. Yeah, so this is technically along rural areas. So all these are agricultural land, you could say. And I am going to make the color a little bit, go to the properties, style color, you can make it, opacity is 100%, let's do 50%. So now you could see that Amrita Vidyalayam Chichi is the school. It looks like a boarding school with a big campus, badminton court, but along a village, you know, entity. And then there is a Uraci, which is kind of like a village region. Mahakalyam and Coil is there. So it does look like a village because of this agricultural land, I'm saying. So most of the agricultural land is there. All these are water bodies. And then you do have a lot of temples around. So houses are also not very close, which is like urban centers. So these could be a, definitely more of a rural entity. I don't know why this motor school is also labeled as a school. We can see the properties just to say, oh, it's St. Francis School, and then there's an archery. So maybe the boarding schools normally would have bought an area outside because it's lesser costly and then do it. So this is the other thing I wanted to show that you can do a time series analysis and to just see like how the school was developed or what was there before the school. So here I see that you could see that all of this was agricultural land. These are agricultural land that can happen. And before the school, there was pure agriculture happening. You can see that all of this agriculture, but then they bought these plots. And you can see exactly that maybe these plots were from a particular owner. These plots were from a particular owner. So they merged these plots together to buy the land and then the school has been built. So in 2002, the school was not there. And again, land use, land cover change, we say 2007, the school was not there. And then slowly the school construction will happen 2010. You can see that the land is being cleared and still some agriculture is happening here. But slowly as the school is developing and the land is developing, you can see that it is being occupied for the school premises. So this is one of the reasons where you could have better use of GIS and remote sensing to look at a particular school, how it has evolved. And most importantly, what is the coverage? How many villages does it cover? And is it good for covering the entire region for school kids? So this is also kind of a rural entity with less number of housings. I'm going to push it to the current. So it says covering global senior secondaries. When it says global, it's not government. So it could be another private institute. So it looks like a lot of these schools have done well in terms of having international names and or having a big campus for a more targeted to the populations of urban. And so the Bishop Heber is there. It was a very, very famous school. And a lot of people from the villages do commute here. So we do have this. So I'll stop with the school part here. We will quickly run the roads for Tamil Nadu again, at least at Ruchi. And we'll overlay it here to show how these are connected to the schools. So I'll keep the TN schools and the TN government schools dots here. And what we're going to do is we'll go back to Google Earth, open our vector plugin, click OSN. And then we'll say roads. Let's say what comes up. So just say roads. It will be highways, streets, relations, roads, roads, unknown type is there, good. And then you can add, you can say some other road type. Let's say bicycle road is important for the small pathways for mostly Europe. But here also small pathways are called bicycle roads, minor roads we want. Unclassified is good. We'll add it. And then we'll also, OK, when we do the preset again and again, it will override it. So let's see if we could add a road. Yeah, so we've added one, one more. So type is route, type is route road. And then I'm going to do amenities. And it's not there. We'll do an R, amenity, and then say road. Highways, look. Government, do we say road? Yeah, maybe. So we do have small roads and small things in type. So maybe we can say type. And then road. Road is already there, it won't come. Highway, rural. So for some reason, all the rural types are taken out. But it's good. I think we already had this as a query. We do have schools, health care. We have run, it's fine. OK, so this should be OK. So type is route, type is road. We could have someone as highway also just for case. And then what are the values in highway? You have bus stops, construction, cycleways, emergencyways, footways. So in these things, we have links, roads. Let's keep it as roads. So we have R, R, R. And then as I said, we'll do a layer extent of Ritchie. And let's run. And come down to say that in advance, I need lines. Not nodes. I don't want nodes. I don't want polygons. We'll just see if lines are available because roads is a line or a pathway. I've run for other places just as a backup because as I said, the quick OSM has just been updated. Yes, good thing we have run it. Successfully, one layer has been added. This is a merge layer of all the three searches we have. So all the three will be there, but as a merge layer. So let's zoom into that layer. And you could see that it has been zoomed in. But I don't see the lines. Maybe we'll see the other person. Oops. You see that? So these are the roads. For some reason, the color and the symbology is too thin. So we'll go to simple. Maybe stroke is too. And then the color can be black because we do have a lot of apply. OK. The big properties is 2.5. 1.5 is good. OK, apply. OK, that is good. And now if I put Tamil Nadu, you can see more. So this is only Trichy. So we said only Trichy. And it is certainly not Trichy alone, but some parts of Trichy. So it is in and around Trichy. It is not capturing the rural roads. So this is where I said infrastructure and road mapping is very, very key. We do need to have it. So I've also added done an exercise on roads. So I just added then this is particularly for a thing that we did initially, highways and stuff. You can zoom into that layer, which is we, if you remember, we did this for a particular place in Maharashtra and Karnataka Park, where we did a georeference and image, and then we extracted the data. So like that, also, you can add the data. This is from that lecture where we georeferenced tile and then extracted the road network. But we can also take highways and other aspects as and when we did. So you also have wells and tanks that I have done, which is water bodies. So someone was asking for surface roads and water bodies, which we will be looking at. So for now, as I said, we can select this road type, move it to export, save feature as, go to OSM, my folder, then say maybe touchy roads, no space, shapefile, add it to the shapefile, remove these layers. You don't want these layers. And then let us see roads. We can also zoom to layer. Yeah, we do have this. It's a good thing we'll open in Google Earth, bro. File, open, cheat roads, open. You want to use it, like I said, no. Again, view, reset, tilt. And there we have it. So I'm just going to open this so that you can see the individual polygons, polynines. Let's click on one. Yes, image has a lot of clouds. OK, we'll remove this so the best lines come up. You can see how beautifully, and this is a village area. So you have this road. Three, I'm on two wheelers, this is a highway. And there is a road through the village. I'm saying village because of the agricultural lands. You see the land divided into parcels. It is not accurate. However, it is beautifully there because it goes through the system. And it is actually dividing even to the small lake and pond. They have a road. So someone has actually meticulously driven a bike or something. So you can actually do it as a bike layer. You can click on a waste point and then click, click, click, click on a path on an app and then collect the data. Again, there's multiple videos on how to collect waypoints. Once you collect the waypoints, you can export it here into OSN. And then there is a beautiful layer. If you can see clearly, these are not your average roads that are laid with track. These are mud roads or non-matt roads. But these are very important. I'll show you why. Because when you go here, I see for this path. So this farmer is growing this land. I hope he is growing. We can definitely look at the images in the past to see, yeah, there it is. So if we can find a growing season, yes, it is green. Definitely there is growing something. This is a summer period, post-summer, not out of rainfall, don't have crops. Yes, so you can see the rows and all cut. So that is the cropping. So now if the farmer is having a small crop, and in this image, we had better view of these crops. So there is a crop. Now the farmer has to harvest it and send it to the market. And the market, everything is dependent on this road connectivity. So the good thing here is from here, we can quickly estimate the time to go to the highway, and which is the best route. So from here, you can say, this will take around line and path, yeah. So from this plot, you can say through the highway, if the roads are connected, you can go through 600 meters, or if you say, I'm gonna clear, and then say from this to this straight road is 300, 400 meters. However, it is not straight. You will have to go through a path. I'll show you why, because it's not connected. So now you have to use a road and say, okay, from here to here, there's a road. And then from here, you have to take this road, this road, this road, and then go to this is Maligai and that gives you 605, 108 meters, 100 meters extra. But you need to incorporate all this on the travels, the budgets, the time. If you know and been in villages, you'll know that how fast you get the produce to the market, the price is different, especially flowers. I've been taking flowers when I was a kid. So when we take flowers in the village, quickly they'll ask us to collect the flowers and give it in a bag, and the bag is being transported to the market. So early, early market price, let's say a kilo is 200 rupees the flower. As time goes up, the same kilo of flower goes to 100 rupees. So the farmer is at big loss. He'll take around 10 kilos a day and all the loss is there. So this is how the connectivity does keep a very, very important aspect for rural development and livelihood. And more on that, the profit margin is changing. The second aspect I also wanted to say is if the produce is quick. So this is just a small thing, a road that is connected very near, very near, 600 meters to the highway. But think about a lot of other options where the cropping is here, for example, here. And the road network is not there. So they have to go through the mud roads and then go to the highway. It takes more time. Let's say it takes two hours to get to the highway and from there it goes. So all throughout this, the crop, let's say a fruit like kiwis or other things that, okay, let's say smaller fruits that everyone uses, slightly tomatoes. So tomatoes may get damaged and the damage is a loss for the farmers. So this is where if the road network is not there, rural network connectivity for the farm is not there, there's a loss in time and livelihood options. So with this, I will stop here. Today we have looked at hospitals, healthcare and roads. We can also look at how the same analogy can be used for how the hospital is connected. But for now, most of the hospitals are in the city in the OASM database. I hope to see more data being populated, especially after this class, on the hospitals in rural villages. I'll see you in the next class with more on crop mapping and agriculture. And then we will call OASM over by week. Thank you.