So, I hope this will help you visualize and understand a little bit more about spectral relationships between different types of objects and spectral signatures as we'll see, and how that relates to spectral resolution. So, what we have here is a graph of percent reflectance versus wavelength. Percent reflectance is just how much light comes, it's a ratio of how much light hits an object versus how much is reflected. So, you can really just think of it in for us, for our purposes here as to how much light is reflected, and the wavelengths are the different parts of the spectrum. So, blue green and so on. So, if we have a certain surface material in this case lemongrass, you'll see that there's not much light reflected in this part of the spectrum, but there's a lot of light reflected up here, and then it drops down again here, and then it goes up again here, and then down. So, this is a way of being able to chart out. So, for different parts of the spectrum, how does this particular material behave? We could compare that to a different type of material like a maple leaf, or a first spruce tree, or dry grass, or dolomite which is a type of rock material, or clearwater, or clearwater versus turbid water, which has sediment in it. So, by the way these are all curves that I've gotten, these are real. This is real data from the USGS website where they talk about landsat. You can go look it up yourself. So, the idea that I want to get across here is that we're trying to find unique combinations between different types of materials, and what that's what we can see here, is that for example, dry grass does not reflect as much light in this part of the spectrum as does a first spruce tree, and a maple leaf reflects more than that, and a long grass reflects even more than that. So, if we were just looking at this part of the spectrum, and there's turbid water, we would be able to tell all of these things apart from one another based on the numbers for the cells for that. If we do that for different parts of this, this is how we can build up this idea of a spectral signature. The idea is that if we can look at the combinations of reflectances for these, this is what will help us be able to uniquely identify different types of materials, and just like your own signature is meant to be unique, what we're hoping for is that different types of spectral responses as we call them are going to be unique, and that will allow us to identify a spectral signature for each type of material. So, with Landsat 8 for example, the light areas here are the bands that are actually sense. So, we're not able to sense the entire spectrum continuously, the sensors are only sensitive to certain sections of the electromagnetic spectrum, those sections are referred to as bands. So, for example, here's band one which is quite narrow. Here's band two in here. This has band three, band four, band five. So, this is- I'm hoping a way for you to be able to identify is that will just something like band four as I was just showing you, you would be able to tell these different types of materials apart, because they have different responses in that band, in that particular part of the spectrum. They're not very different, at least a lot of them aren't very different than band three. So, band three would not be very useful for separating out those different types of materials. Let's see what else. Band six some of them are a little more similar although dolomites way up there, that would be pretty easy to recognize. Here some of them are more similar, here they're similar. So, the idea is that if we have all these different bands, we can try and see well, which bans worked for this type of material that are different than other types of materials so that we can identify that spectral signature and be able to identify what that thing is on the ground. Here's the panchromatic band for Landsat 8. So, this is a wider band. So, we're able to get a higher spatial resolution, and we're not able to necessarily tell things apart quite as well, because it's such a wide band. Yes, for this part of it you may think well you can tell them apart there, but not so much there. Remember, it's only getting one number for that whole band at once, it's not able to-, this isn't a smaller band. Like this is the whole thing, you can't just say, "But I just want this part." Well, if you want that part, that's a narrow band, that's a higher spectral resolution, and that will lower the spatial resolution. So, if you want to higher spatial resolution, you have a wider band, and so you have less spectral information that you're able to get out of that. I just put this slide in for your reference, these are the band numbers and names, and wavelengths for Landsat 8. It's probably will lease for the type of work that I see the most popular satellite and sensor. So, you'll see that they have different names. So, Landsat 7 did not have a coastal aerosol or also known as an ultra-blue. Landsat 8 has that. They re-number them a little bit between the two. So, a lot of historical date as Landsat 7 or even earlier, Landsat 5, but Landsat 8 is the most popular current one that's available. So, you can see I'll go through each of these, but you can see that each of the wavelengths is good at particular things. So, for example, near-infrared is good for emphasizing biomass content and shorelines. There's a serious ban that's good for detecting certain types of clouds and so on. So, you don't necessarily have to memorize all this, or have them at the top of your head immediately right away, but the more you work with this type of data, the more you see these different kinds of combinations, the more you'll get used to being able to work with these, and figure out what's going to be best for your application.
