 Hello everyone, welcome to the next lecture in the course remote sensing principles and applications. We have been discussing about the about the spectral reflectance properties of few common earth surface features in which we started discussing about the spectral reflectance curve of vegetation and what factors that influences it. In today's lecture, we are going to continue with the topic of spectral reflectance properties of vegetation and also we will move ahead with other commonly occurring earth surface features. So, till the last class I told you that the spectral reflectance curve of vegetation can be broadly divided into 3, the visible NIR and SWIR ranges and we are also seen in detail the factors that influences the spectral reflectance property in each of these portions of electromagnetic spectrum. Today we are going to get introduced to your concept of what is known as a red edge. So, what exactly a red edge is? A red edge is defined as the transition portion where the reflectance suddenly increases. Let us look at the spectral reflectance curve of vegetation here. If we can observe in this curve after the red portion something around this 0.68 micrometer range, the reflectance suddenly increases and it reaches like a very high value. So, there is a small transition zone like small portion of bandwidth where this transition occurs that is from a very low reflectance in the red portion to a very high reflectance in the NIR portion. This transition zone or this portion of the spectral reflectance curve where there is a sudden transition of reflectance happens we call that as the red edge. So, the red edge actually has certain applications in vegetation monitoring that is if you look at this maybe we look at like example within the previous lectures yes. If you look at this particular slide we can see that as the vegetation's nature changes like here in the slide we discussed about how the spectral reflectance will vary when there is change in water content in the leaf right. Here you observe what happens to the red edge as the leaf dries out basically it will undergo water stress and once it undergoes water stress then the reflectance in red will begin to increase because we have seen that as vegetation undergoes any sort of stress it will begin to reflect in the red portion of the spectrum. So, the red edge actually like now begins to happen like the reflectance in the red portion now has suddenly increased here. Similarly there is also a change in the NIR reflectance. So, as the vegetation property changes be it some sort of stress or be it senescence or be it like abundant growth of vegetation whatever happens it affects the red edge and also it affects the wavelength range at which this red edge occurs. So, essentially speaking the red edge is the the transition zone or the increase in reflection from red portion to NIR portion in the electromagnetic spectrum of vegetation. Also one more important thing to notice is the portion at which or the portion of EMR at which the transition occurs. Typically speaking the sudden increase in reflectance from red to NIR will happen around the red to NIR transition zone that is around say 0.67 micrometers to 0.72 micrometers within that like very short range where the wavelength transitions from red band to NIR band this increase in reflectance will occur. So, this red edge and also the wavelength position at which the red edge occurs will help us to know certain properties about vegetation. As I told you the red edge position that is the wavelength at which the red edge occurs it will change based on the condition of vegetation. As the vegetation undergoes some sort of stress or when the vegetation undergoes some senescent cycle like when the vegetation matures then the red edge the wavelength at which this transition occurs that portion of wavelength will start to move towards shorter wavelengths. We call it as the blue shift that is say let us assume the vegetation curve something like this. So, let us say this transition occurs something around 0.68 micrometers the position at which this reflectance changes. If the vegetation undergoes some sort of stress then this curve will become something like this the reflectance in red will increase similarly the portion at which the transition occurs at which the change occurs will move towards shorter wavelength. Now this wavelength may not be 0.68 it may be let us say 0.65 micrometers. So, this sort of shift will happen towards the shorter wavelength we call this as blue shift of the red edge. So, whenever when blue shift will occur blue shift will occur whenever the vegetation is undergoing senescence or when it is undergoing some sort of stress that is when the red reflectance increases then the red edge the portion the wavelength at which the red edge occurs will move towards shorter wavelength. Similarly, when the vegetation is under like active growing phase like it starts from like very small plan 2 it suddenly increases to like a large canopy when that thing happens then the red edge portion will move towards longer wavelength that is let us say the vegetation was something like this before after sometime like after maturity it may become something like this that is the this curve the dark red curve has now moved to the position of like this dotted curve actually yeah. So, this signifies the red edge is now moving towards longer wavelengths we call this as red shift of the red edge. So, observing this red edge and the position at which it occurs will help us to understand certain properties about vegetation and also for agricultural crop monitoring or in general vegetation monitoring etc. But one thing what we have to remember is the wavelength 3 inch at which the red edge occurs is actually like very short is of very short bandwidth. I told you it may fall around like say 0.67 to 0.72 micrometers and even when the change occurs like when blue shift occurs or when red shift occurs the change in wavelength also will be in the order of say 1 micrometer even less than 1 micrometer the change will be like very narrow will be occurring over very narrow bandwidths of wavelength. And hence in order for us to properly observe this red edge and the position of the red edge we may not be able to do it with our normal multi-spectral systems with wider bandwidths. Let us say one system has like bandwidth of 0.65 to 0.75 micrometers like crisscrossing red and NIR or one may be having like 0.62 to 0.68 another one may be having is next band may be 0.70 to 0.79 micrometers etc. This sort of wider bandwidth sensors which are typically present in our multi-spectral systems may not be able to capture this red edge or the position of the red edge. For observing the red edge we need what is known as like imaging spectrometers or hyperspectral sensors. So, what exactly are hyperspectral sensors? Hyperspectral sensors are such sensors which observes in many short continuous bandwidths that is like in the towards the end of the last lecture I showed you about like the bandwidths in Landsat thematic mapper sensor. Maybe we will just quickly revisit it again. This particular figure in the slide actually tells us the different bands present within the Landsat thematic mapper sensors. So, it has this is band 1, band 2, band 3, band 4, band 5, band 7 and so on. So, this is for thematic mapper sensor. So, this is an example for what is known as a multi-spectral system. So, multi-spectral system is a system or sensor which has a limited or a small number of wider bands that may be or may not be contiguous. Contiguous in the sense like they are not continuous they band 4 to see how in the NIR range suddenly band 5 has moved to NW sorry SWIR range in the wavelength of say 1.5 micrometers. This sort of non-contiguous wider bandwidths we call such sensors are multi-spectral sensors or multi-spectral systems. On the other hand, hyperspectral systems or hyperspectral sensors will observe lot of small small bandwidths continuous bandwidths say it may be 0.412, 0.42 micrometers, 0.422, 0.43 micrometers like this even for hyperspectral sensors they will classify everything in nanometers 410 to 420 nanometers, 420 to 430 nanometers. It is convention to use nanometers to represent hyperspectral sensors. So, they will have large number of continuous bands with very narrow bandwidths. So, the presence of this continuous bands without any gap 0.41 to 42, 42 to 43, 43 to 44 like this large number of continuous bandwidths without any gap along the electromagnetic spectrum also very narrow bands like normally like in Landsat the bandwidth is in order of like few micrometers whereas like even sometimes tens of micrometers. But here it will be in order of tens of nanometers or even shorter. So, presence of large number of such continuous narrow bandwidths on the sensor containing such bands are known as hyperspectral sensors and for observing this red-edge we will be needing such hyperspectral sensors. It is not possible to observe red-edge from normal multispectral sensors. But nowadays certain satellites are having a specific band known as the red-edge band which is occurring in the transition zone of red and NIR very few satellites has this in order to observe this red-edge. But most of the commonly used satellites for which data is freely available to us do not have this specific red-edge band they have the traditional green red NIR bands or such that very few satellites has this red-edge band built within them. But using hyperspectral sensors it will be possible for us to observe this red-edge and also the position at which the red-edge occurs. So, I told you that the position at which red-edge occurs will help us to understand about certain properties of vegetation. So, how to calculate the position of red-edge? There are like many different ways, many different indices are available to calculate this red-edge, red-edge position etc. A very simple example we are going to see in this lecture. We are going to calculate red-edge position using what is known as a derivative spectroscopy. So, what exactly is this? Like we have seen that for a normal spectral reflectance curve the red-edge seems to be the red-edge is the portion at which the reflectance suddenly increases. So, if you talk in terms of like mathematical functions, so red-edge is nothing but or the position at which red-edge occurs is nothing but the point at which the slope of this reflectance curve is maximum or when it increases suddenly, when the slope increases suddenly when it reaches the maximum. So, how to calculate slope of the curve in mathematics? We have seen it using the derivative, the first derivative of a function say if y is equal to f of x, then the first derivative of the function dy by dx will tell us the information about the slope of the curve. Similar concept we apply here, we take the slope of the reflectance curve with respect to wavelength, how the reflectance changes with respect to wavelength and we identify a point where the change in slope is the maximum. So, here is the first derivative of slope, so the change in slope is maximum. We identify that particular point or the wavelength at which the maximum change occurs, that particular point we classify it as the position of the red-edge. So, that is given here. So, this is possible from hyperspectral sensors. This is reflectance in band j plus 1, this is reflectance in band j divided by bandwidth. So, you will be like keep on calculating this and j plus 0.5, the band at which maximum change occurs, we will classify it as the position of red-edge. So, this is one of the very simplest ways in which red-edge position can be calculated. But as I said there are plenty of ways, there are like separate chapters of how to use red-edge for vegetation monitoring and so on, but we will not go in detail in this particular course. The main aim of this particular topic is to introduce to you the concept of red-edge and its importance. Till now, we have spoken about the reflectance property of a single leaf or grass etc. In reality, when a remote sensing sensors observes vegetation from space, it will not be observing a single leaf, but it will be observing a collection of leaves or a bunch of leaves plus the stem, some soil, flowers, fruits etc. So, normally a remote sensing system or a sensor will observe or will observe like a vegetation plus its background like non-leafy things like stems, barks, fruits, flowers etc. So, how these will change the reflectance of vegetation? That is what we are going to discuss in the subsequent slides. First, we will talk about what is known as a leaf additive reflectance. So, while discussing about the NIR portion, the spectral reflectance of vegetation in the NIR portion, I told you that leaf transmits a large fraction of incoming energy and it also reflects other large fraction of energy. Say the transmission will be about say 40 percent, reflectance also will be about say 40 percent. Both are equally high, in compared with the total absorptives. So, what happens to this transmitted energy if a single leaf is present and some radiation is incoming towards it, a part of the radiation or a large chunk of the radiation is transmitted towards it. So, it just passes through without undergoing any change. If it is just a single leaf, it would have just passed through and it would have interacted with whatever was there underneath it. But normally as I told a vegetation canopy will contain more than a single leaf like there will be like 10s or even like sometimes hundreds of leaves arranged in kind of like a stack depends on how thick the canopy is or how big the canopy is. So, what happens to this transmitted energy? The transmitted energy in general will interact with the leaf underneath it and that will again transmit certain portion of it and reflect certain portion of it. So, the transmitted energy from one leaf will act as input energy to the leaf underneath it which will again add up to the reflectance that is finally coming out. So, if you see this particular slide the figure in this slide you can see. So, this is like leaf in the top let us say leaf number 1, this is like leaf number 2. So, what happens? Certain amount of incoming energy is coming and falling over it. Let us assume 50% of it is kind of reflected back. So, this is reflected by the leaf. Again let us assume another 50% is transmitted. So, let us assume like the absorptance is almost 0 like just for the example sake I am telling here. So, 50% is reflectance, 50% is transmittance with almost 0 absorptance that is the case the 50% energy is now transmitted to leaf 2. Now what will happen? This leaf will reflect 50% of this incoming energy. So, that is from the total 50% came in like from this phi i, 0.5 of phi i came in out of which another 50% will be reflected that is one fourth of original incoming energy this will be reflected and another one fourth is now transmitted through. So, this is also will pass through because what happens like this will while passing through again the bottom portion of the leaf will reflect some portion again here and again a 50% of it will be transmitted. So, this one fourth will be will further became half one eighth will be again reflected by the bottom portion of leaf remaining one eighth will be transmitted towards the upper portion. So, in the upper portion we have the first half of reflected portion plus this one eighth of this reflected portion or 12.5%. So, these two will add up and it will finally make up for 62.5% of reflectance that is whatever the original energy came in a part of it got transmitted this transmitted energy gets reflected by the leaves underneath and while this is passing through this will undergo multiple reflection even by the bottom of the leaves also and finally the total reflectance observed by a sensor at the top of this first leaf will be more than what is being reflected by this single leaf. So, the final reflectance or rather than putting final I will say total the total reflectance observed by a sensor due to multiple leaves will be greater than the reflectance produced by single leaf. This is because of the presence of multiple leaves underneath it. So, in this example for the sake of simplicity we took this leaf reflects 50% transmits 50% without any absorption this is just for explanation sake but it will change the numbers will change but the concept is this whatever is being transmitted by the leaf on the top will act as the input source of energy to the leaf at the bottom and it will undergo certain reflection and it will add up to the total reflection recorded by the sensor at the top of the canopy. So, this figure in this slide actually tells us more about the leaf additive reflectance. This leaf additive reflectance is very clearly observed in the NIR band because leaf has the greatest transmittance in the NIR band that we have seen and I also now we would have understood that for leaf additive reflectance this transmitted energy is kind of like one of the major input source okay. So, in NIR band as the number of leaves increases the reflectance increases quite sharply this is also observed in SWR portion to some extent. So, the presence of a large number of leaf or a very high leaf area index what we call like leaf area indexes the area of leaf or to be more specific one sided because a leaf has two side. So, area of one side of the leaf total number of area in a given ground area area of the ground. So, this is what we call it as leaf area index LAI. So, as the LAI increases LA may increase because of when large number of leaves gets added up underneath. So, this is one leaf there may be another leaf like this another leaf like this and so on. So, as this number of leaves grows in a canopy the LAI will increase as this happens the reflectance in NIR portion and also to some extent in the SWR portion will increase because of this added number of leaves primarily because of the transmitted energy from the leaf at the top acts as input energy to the leaf at the bottom and the reflection when the leaves at the bottom will add up the total reflectance observed by the sensor on top of the canopy. I also told you that in addition to this leaf additive reflectance also told you whenever a sensor observes vegetation it will not only observe leaves it will also observe other non-leafy components such as stems, branches, flowers, fruits even the background soil. So, all of these combined together will create a integrated signal in the remote sensing sensor. So, and this kind of effect like whether a sensor is observing leaf or it is observing non-leafy portions etc depends on both the illumination geometry and also the viewing geometry. Maybe we will quickly see an example. Let us take an example of a row crop. Okay, so this is like a land parcel where crops are planted in kind of rows like this. So, each red line here represents one row of crop. So, whenever such row cropping is being practiced the in between portion will normally be left barren or left fallow without any vegetation. So, this will be like the in between portion will be like soil portion. Okay, so let us say a sensor is looking directly from overhead that is like here this is where the sensor is being located. Okay, it is directly looking overhead. So, when a sensor is looking overhead this particular sensor. So, this particular sensor when it looks like this it will look both the crops like this is each row of crop plus the soil in between. So, this particular sensor is now going to observe both vegetation like leafy components also the soil components. In other case let us say the sensor is located here and row crop is standing something like this. So, when the viewing angle is something like this most of the sensors view is going to be limited for just by observing this vegetation portion or the leafy part rather than looking at the soil. And also let us take sun is present somewhere here now. Okay, so sun is present here. So, what will happen the sun is going to illuminate this other side of the vegetation whereas the sensor is seeing the another side. So, essentially the reflectance is going to vary if the sun is present here. So, the same angle as that of the sensor. So, now the sensor is going to observe like a bright portion of vegetation. So, all these things combine together sensors or how the vegetation is aligned whether it is a row crop or whether it is like a randomly distributed like forest canopy or how the solar illumination geometry is how the sensor viewing geometry is all these things are going to create an integrated effect of radiance in the sensor. So, essentially when we do remote sensing most likely we may not get only the leafy part. It is possible only when a very thick canopy is present that is when the sensor is observing over a large thick forest such as Amazon. When observing over such thick forest the sensor might be seeing only the canopy parts only the leafy portions. But when there is like a sparse vegetation such as like some agricultural lands with row cropping pattern or some shrub lands when vegetation is quite sparse then the final reflectance or the radiance observed by the sensor is going to have an integrated effect of soil, leaf and other non leafy part. So, we should always keep this in mind when we compare the reflectance from vegetation with the spectral reflectance curve of leaf taken from a laboratory. The spectral reflectance curve we get from remote sensing sensor may not exactly be obtained from a leaf it may be obtained from several different features. We should always keep this in mind when we do this curve matching what is being obtained from laboratory and what is obtained from remote sensing image when we compare them we should always keep in mind. Most likely we should compare only the pixels which add only like pure vegetation in order for us to do our classification in a better way. In addition to all these we have also come across that vegetation is in the non-lambushian reflector that is vegetation will look differently as the sensor viewing angle changes. That is I one of the classes I told you that vegetation is primarily a backwards character that is it reflects a good chunk of energy in the direction from which it is coming itself. Similarly, the vegetation's reflectance property will change as the sensor viewing geometry changes. So, here an example is given in this particular slide. So, here we have plotted what is known as anisotropy factor. So, anisotropy factor is how the reflectance for a given angle for a given viewing geometry that is like theta v comma phi v and how the reflectance divided by the reflectance when taken from nadir looking sensor. So, this will give us the anisotropy factor. So, how the reflectance will be when taken from any one particular angle that is divided by the reflectance of the same vegetation observed at nadir. If we plot this ratio along with this viewing zenith angle and the azimuth angle then what we will get is a three-dimensional plot like this which tells us at which angles the anisotropy is the maximum. So, what essentially it means is the viewing angle of the sensor has a larger BRD of fx with respect to vegetation. So, a vegetation when observed at different different angles may look completely different and blue and red bands show higher BRD of fx than green and nir bands as per some experiments. So, all these things suggest that observing vegetation may not be a straightforward task when we observe vegetation and when we take certain decisions about vegetation and its properties we should keep in mind several things that is vegetation what we observe as vegetation may not be pure leaves it may be soil or it may be non leafy parts and the difference in viewing angle and the leaf in the sensor illumination geometry might have played a role which would have affected the reflectance that we have obtained. So, all these things we should keep in mind when we make certain important decisions maybe like a very good example or case study I will tell when we discuss about what is known as a spectral indices how to combine these reflectance to create spectral indices and one of the interesting case study about like this greening of amazon rainforest I will explain you when we discuss that in the later part of the later part of lectures in the same topic. So, with this we end the spectral reflectance properties of vegetation in the next lecture we will start with studying about the spectral reflectance property of soil, water and snow. Thank you very much.