 Hello everyone, welcome to the next lecture in the course remote sensing principles and applications. Today we are going to continue discussing about the spectral properties of various commonly occurring earth surface materials. In the last lecture we discussed about the spectral reflectance property of water and what factors will influence the spectral property. Today we are going to start with the concept of understanding the spectral reflectance of snow. So before moving on to understanding the spectral reflectance of snow, we will just get introduced to small introductory concepts about snow. So snow is nothing but a collection of ice crystals or ice grains, air, water and other contaminants mixed together. Everything like ice plus air, whatever everything mixed together will form like snow pack. The snow occurring at different places in the globe will vary a lot in the size. Size means the grain size, how big or small the grain sizes, the density or the packing of the snow, the amount of contaminants present etc. So snow exhibits a very high amount of variability from place to place and time to time. And also compared to any other material in the globe, snow has the highest or brightest reflectance especially in the visible band. Like we might have experienced this, some of us might have seen snow really or some of us might have seen it at least in televisions or movies which is snow appears extremely white and bright in the visible domain. So snow is like the brightest among most other occurring earth surface features. And also snow or its particle size and everything this will be under like constant evolution. That is when snowfall occurs today after a certain period of time, snow will undergo some change like it may become more dense, the grain size may increase or some contaminants may get added to it etc. So snow shows high degree of change in both spatial and temporal domain. So when we want to understand the spectral property of snow, we should realize that the final reflectance or the final spectral property we obtained from snow is an integrated effect of the ice grain, air mass, water molecules and other contaminants mixed together. So it is not only the spectral reflectance of ice but also like mixture of several different things that could be present in the snow. So some of the factors that influence the spectral properties of snow are first thing is the grain size, how big or how small the snow grain sizes, age of the snow. Most likely snow will become larger in size as it ages like more and more accumulation may fall, it may grow larger in size, hence snow age will influence it. Then presence of any liquid water molecules within the snow will affect the spectral property of snow. Then the solar zenith angle in which angle the sunlight is falling on snow, then presence of any contaminants may affect the spectral reflectance from snow and also the thickness of the snow pack, whether the snow is really thick or it is like shallow etc, will all these things will have the ability to change the spectral property of snow. First we will look at the general spectral reflectance curve of snow and how it changes with grain size as well as the wavelength of observation. First if you look at this particular slide, here we have plotted two curves, one is the spectral reflectance curve for fine snow that is snow with fine grains or smaller grain size and then this orange curve is the spectral reflectance curve for snow with larger grain size. If you look at them, first what we are observing is for both the grain sizes whatever be the grain size, the general pattern is invisible up to say 0.7 micrometers. Let us draw some line somewhere here, up to this the reflectance of snow is extremely high in the order of like above 90 percent. So the reflectance of snow is extremely high in the visible domain and also up to say 0.8 micrometers. Only after like when the wavelength enters NIR band after it crosses the 0.9 micrometers, the reflectance begin to decrease drastically and after crossing this 1.4 micrometers the reflectance is quite low in compared to the reflectance in visible. And hence here also we can divide the reflectance into visible and IR bands. I generally make it as IR which comprises both NWIR and SWIR. So highest reflectance occurs in visible band and in NIR the reflectance begins to decrease and it decreases overall we can observe the trend is decreasing as the wavelength increases the overall trend. So in general the reflectance will decrease in increase in wavelength in the IR domain except these 2 like small peaks almost the overall trend of reflectance is coming down or decreasing as the wavelength increases starting from NIR band. So this is the general property of snow or general spectral reflectance curve for snow. Among this the coarse grained or coarse grained means larger grained snow will have still lower reflectance than snow which is composed of fine grains. And hence the grain size how small the grain sizes will increase the reflectance. So finer the grain size higher will be the reflectance, coarser the grain size or larger the grain size lower will be the reflectance. So this will be in general applicable across all bandwidths and this effect will be more pronounced in the NIR and SWIR ranges. Say we will again go back to the curve here if you look at or compare both the curves the difference between both the curves is very minimal in the visible domain. Both the curves are almost close to each other and still we have very high reflectance not much of a difference. Once we enter or once we cross this 0.8 micrometer when we enter the NIR portion of electromagnetic spectrum the differences begin to increase the difference between fine grained snow and coarse grained snow. And as we progress or as we increase along the wavelength we are we can observe larger difference in reflectance between fine grained snow and coarse grained snow. This difference is again highly amplified in the SWIR domain. Say here we can observe the peak here is much higher when compared with peak here. Here also this is one thing and also we can note a characteristic dips closer to the water absorption bands even in case of snow 1.4 and 1.9 micrometers characteristic dips may most likely will be present. So in general in order to understand the spectral reference property of snow snow has very high reflectance in the visible bandwidth and the reflectance begins to decrease starting from NIR and in general the reflectance decreases as wavelength increases. Then for fine grained snow or snow packed with smaller grain size the reflectance will be higher than the snow with larger grain size and this difference in grain size is highly amplified in the infrared domain especially NIR and SWIR domain of electromagnetic spectrum. So this particular slide further emphasizes how grain size influences the spectral reflectance of snow. So the figure given in the left side spectral reference of snow with snow having grain size with varying radius is being plotted. So the radius starts from 50 micrometers and goes all the way up to 1000 micrometers. We can observe as the grain size increases the reflectance keeps on decreasing continuously. The difference is not much in the visible domain but it is much higher in the NIR and SWIR domain. Maybe this will give the figure on the right side and will give a more clear picture like here we can see how the reflectance varies as a function of square root of grain radius. So in the x-axis we have square root of grain radius in the y-axis we have the reflectance. We can see invisible domain or even up to say 0.8 micrometers or the change in slope of this curve is fairly shallow. So which signifies that reflectance does not vary much with increasing grain size up to 0.8 micrometers. Once we cross this 0.8 micrometer domain then the reflectance begins to increase quite fast as the grain size increases and it is like highly amplified as the wavelength increases. We can just compare the, we can compare how the reflectance changes, how fast it decreases for wavelength starting from 1.1 micrometer up to 1.8. Even in 1.8 it is almost like kind of low exponential kind of decrease. It starts pretty high reflectance then suddenly it falls to very low reflectance. This signifies that grain size plays a major role in controlling the reflectance of snow and this difference is highly amplified in the infrared domain of electromagnetic spectrum. The next feature that we are going to see is the snow age. We have already seen that as the snow ages or as the snow becomes older the grain size tends to increase more and more snow may fall on it it may become densely packed or the grain size may increase and when the grain size increase we know the reflectance will decrease. So in general as the snow ages the reflectance decreases primarily because of increase in grain size. An example is given here in this particular slide. So this is the spectral reference of fresh snow with the density of about like 0.2 grams per cubic centimeters. Here it is a 20 day old snow this dark black line with a higher density. So as the snow becomes older it will get more dense and its grain size will increase leading to a decrease in spectral reflectance observed. Then snow may undergo process of melting and refreezing or water may get added to the snow pack. So presence of water in the snow pack will make the snow to form larger clusters that is let us say some small water droplet is present there. So snow particles may tend to form cluster around it and in general became like one large mass or one large cluster. This kind of occurrence or this kind of practice may increase the overall grain size of the cluster like in order to cover the single water drop many number of smaller grain crystals snow grain crystals may surround it and became one larger cluster which will affect the spectral reflectance. So primarily if the snow grain size changes or if snow forms cluster because of presence of water molecules within it then its reflectance will go down because it is more related with the change in grain size rather than presence of change in water. In the visible NIR presence of mere water may not change the reflectance much but if the snow becomes cluster due to the presence of water and becomes one large mass then it may affect the reflectance of snow and also once snow melts and then refreezes the reflectance may not be as high as the first initial phase that is given here in the slide. See here the spectral reflectance curve A indicates the reflectance curve for original cold snow with some density value then when the snow melts the reflectance curve falls to curve B which is like a dotted black line represented here. So the reflectance decreases as the snow melts then the same snow refreezes but the refrozen snow the reflectance will not increase as much as the original condition. So this will increase from the melted state but still the density the density has increased and hence the reflectance will be lower than the original reflectance that was first observed in the snow was completely in a frozen state. So presence of water also will try to bring down the reflectance of snow particles. The next property we are going to observe is the solar zenith angle. So solar zenith angle basically signifies in which angle solar radiation comes and falls on the snow pack. So in general as the solar zenith angle increases the reflectance will increase that is if the snow surface is like this let us assume the sun is directly shining from overhead that is sun is at the zenith angle is 0. That is the case we will we can measure one sort of reflectance curve then as the sun's zenith angle increases then the solar or sorry the snow pack reflectance will increase. So an example is given in this particular slide. So here you can see the reflectance curve is plotted for one particular snow grain size with varying solar zenith angles. So we can observe in general the reflectance increases as the solar zenith angle increases. So this is for one particular snow size without changing it and everything is measured with direct solar radiation. So diffuse by direct ratio is 0 that is whatever is the energy measured or plotted in this particular curve everything is primarily due to the direct radiation from the sun and the diffuse radiation is completely controlled for building this curve and diffuse radiation is 0. So that is why the diffuse to direct ratio is given as 0. This is because objects may behave differently under the influence of diffuse radiation. So in order to understand the effect of direct solar radiation controlled lab experiments were carried out where the diffuse radiation was completely cut out and only direct solar radiation was allowed to fall. And finally presence of any contaminants. So I told you before that snow has the highest reflectance among most of the commonly occurring earth surface features especially in the visible domain. So whatever be other objects when it falls on snow it will reduce the reflectance of snow say some sort of dust, some sand particles, some vegetation, something whatever falls on snow will in general have lower reflectance than the original reflectance of snow. This is highly true in visible domain. So presence of any contaminant will decrease the overall reflectance of snow. A very good example is given in this particular slide. So here we can see like this is like a fresh snow taken during like just after like winter season after everything got like accumulated can see how bright it is. So the figure on the right side it is taken after certain time period in like spring and summer season where a large fraction of snow has melted and due to the spring and summer season lot more like particles got travel and got settled in top of the snow cover. And this changes the total reflectance of snow. So we all know that we might have heard in scientific articles and all like deposition of any dust particles over the snow will lower its reflectance which will alter the energy balance of the snow pack. It may faster or it may make the melting to occur more fast. So the presence of dust will not only change reflectance it also has a larger implications in maintaining the surface energy balance of snow. So this covers the spectral reflectance property of snow basically. But remember one thing whatever the spectral properties or whatever the different factors that control them we have discussed most of the curves have been obtained under strict lab conditions like as I told you in the just in the previous example above the solar zenith angle in order to study the property of snow only under the influence of direct sunlight diffuse skylight was made zero in by controlling other factors in the laboratory. It is possible only in control lab conditions but during field experiments or during field visits we may not see the patterns a very good example is soil like when I explained about the properties of soil I told you that fine grained soils like silt or clay has higher reflectance than coarse grained soils like sand this is under lab conditions but under field conditions this may be completely reverse fine grained soils like clay or silt may have lower reflectance because of various other factors involved. So on field and in laboratory the conditions may be extremely different and we should always apply our due diligence when we try to compare the spectral reference curve obtained during field measurements and with the spectral reference curve obtained under controlled laboratory conditions we should always consider what are all the differences or what are all the other factors that may play a role in changing the spectral reflectance curves obtained under field conditions. So, if we look different remote sensing images and if the image has both snow and cloud then there may be a confusion for us to identify which one is snow and which one is cloud we may be able to identify using some other features like snow may be occurring like a large pack or clouds may have like some specific patterns or it may cast shadow on the ground all those things may act as clues to us when we look at an image but how to differentiate cloud and snow spectrally we will see from this particular slide. So, in this slide the spectral reference curve of snow and cloud is actually being represented. So, here you can see in the visible domain or even up to like NIR bandwidth the reflectance of snow and cloud are more or less similar. So, it will always be impossible for us to spectrally differentiate cloud or snow up to NIR domain say up to say 1.4 micrometers or something. But once we cross the 1.5 micrometer domain or 1.4 micrometer domain then the reflectance of snow will decrease drastically but cloud will still have higher reflectance. So, the best possible band to observe cloud or to differentiate cloud and snow will be the SWIR band short wave infrared band greater than 1.5 micrometers. In wavelength less than this value or less than this threshold reflectance of snow and cloud will be more or less the same and we may not be able to differentiate them spectrally. We may have to do based on other features as I told based on like cloud shadows or something but spectrally differentiating cloud and snow may not be possible in wavelength less than 1.5 micrometers and after 1.5 micrometers they may become more clear and visible. So, with this we end the spectral properties of snow. The next topic we are going to enter is what is known as a spectral indices or more commonly known as vegetation indices. I prefer to use the term spectral indices because nowadays such indices are being developed unused not only for studying about vegetation but also for various other features on the earth's surface. So, in common we will use the term spectral indices but when we talk especially about vegetation then we may use the term vegetation indices. So, what exactly a spectral index is? So, spectral index is singular term and spectral indices is the plural term. So, what exactly a spectral index is? A spectral index is a combination of reflectance in more than one band of electromagnetic spectrum. Here band especially I am talking in wavelength less than 2.5 or 3 micrometers especially visible NAR and SWR domain. So, if we combine the reflectance of certain feature in more than one band such that the combination produces one single numerical value out of it like we may divide one over the other we may add something or we may multiply one with some other factor whatever. Essentially we are combining the spectral reflectance of a feature in more two or more than two different bands such that they produce one number one numerical value. And by looking at that particular numerical value we will be able to tell some information about the physical or other properties of that particular feature of interest. So, one of the earliest spectral indices that was developed is like a ratio index that to for vegetation say a small introduction I will give. Say in olden days when the when remote sensing was primarily used for understanding the or monitoring the vegetation people were having difficulties in comparing the spectral reflectance images acquired during different dates or during different places etc. Because as I told you the spectral reflectance curve we obtained from remote sensing images may not only be affected by the terrain but it may also be affected due to surface topography, solar illumination angle, atmosphere and various other factors. So, in order to remove them people were trying different concepts or different methodologies. Once a simple methodology that was developed using a ratio index. So, ratio index is just dividing the spectral reflectance or reflectance value in the red band or combining the spectral reflectance value in red band and NIR band that is rho NIR by rho red. So, this was one of the earlier developed spectral index we called it as the ratio index because it is a simple ratio dividing the spectral reflectance in NIR band by red band and the primary use of this is to monitor vegetation that was the major aim with which this index was developed. So, why what it will do if you look at this NIR is in the numerator red is in the denominator. So, when we we already know that for a healthy vegetation NIR reflectance will be quite high and red reflectance will be quite low and when vegetation undergoes stress red reflectance will increase that also we have seen like we have seen how red reflectance increase and how blue shift occurs that is how the red shift move towards the shorter wavelengths those things have already seen. So, for a healthy vegetation this ratio will be very large and for vegetation the stress or some sort of diseased vegetation this ratio will be smaller. So, just by looking at the value people were able to get some information about the vigor of vegetation vigor means the healthiness of vegetation and also after doing the ratio people are also able to remove some effect of topography, solar illumination etc. These two are like cortically placed bands next to each other. So, the effects may be more or less equal. So, people when they took the ratio of these two spectral reflectance values they were able to observe the features more clearly or they were able to reduce the effect of topography, solar illumination and various factors. So, the primary aim of developing such ratios or such spectral indices were to first thing to increase the information content about any one biophysical property of the feature under discussion. That is here in the example vegetation how healthy a vegetation is. So, just by looking at this number this ratio value we will be able to tell higher the number more healthier the vegetation is that sort of simple analogy we will be able to reach it and also it will reduce some other unwanted effects due to topography, solar illumination, atmosphere and so on. So, spectral indices helps us to remove some unwanted effects and also brings to the front more clearly some important biophysical characters about the feature of our interest. So, this was like one of the earliest spectral index that was developed. But the problem here is in the especially the ratio index is the values are unbounded unbounded in a sense let us say due to some reason the reflectance in red goes very close to 0. Then the ratio becomes extremely high it is it will approach towards infinity and also let us say some other reasons may or even the any difference in kind of like calibration may take the values to negative values sometimes it will happen like even if you process some satellite data you will understand due to the varying calibration constants used some reflectance value may go to negative region it is like mathematically it will happen not physically. So, under certain circumstances the ratio or the number may behave very widely it is unbounded as red reflectance decreases the value may increase suddenly all these problems will occur. So, from this particular simple ratio index people started developing other indices so that simple biophysical factor can be studied more in detail or in a clear way and at the same time some other unwanted effects will be removed. So, in this particular lecture we have seen about the spectral reflectance property of snow what factors will influence the spectral reflectance properties and also we have started discussing about what is known as a spectral index. So, in the next lecture we will see or we will get introduced to some of the few commonly used spectral indices and the major applications for which they are employed. With this in this lecture thank you very much.