Michael Benson | Satellite Remote Sensing of the Environment

all right so I want to welcome you all to another installment of the North Campus sustainability initiatives north campus sustainability hour we have a website lists all of our past and upcoming events as well as has videos from all of our prior talks so if you miss one feel free to go check it out you can download it we’re also on Twitter at UMMC si today’s topic is satellite remote sensing the environments so normally this is where I would introduce our speaker today I’m our speaker so I’m gonna introduce myself my name is Michael Benson of a graduate a PhD candidate and electrical engineering computer science I work in the radiation laboratory don’t worry it’s not ionizing so I’m not radioactive I don’t think with that said our our the next 40 or so minutes we’re gonna divide into third I mentioned I’m a PhD candidate for those of you that will be attending my defense presentation within the next few months don’t worry the amount of overlap between today’s talk and that are minimal so that incur I encourage you all to attend date TBA because there’s still work to be done that’s a fun joke today though we’re gonna talk kind of take a basic approach to satellite remote sensing of the environment talk about how it works what what it is why it is how it works some kind of basic sense and a specific application that relates to tracking the kind of understand the global carbon cycle which is an extremely important global system so this is what we told you in the email that went out and how many in here received at least one copy of the email for today’s talk show ants fantastic if you didn’t make sure you sign up on the list no I’m gonna divide this into thirds so we’re talk first about remote sensing in a general sense insert different applications then we’ll talk about something the Canadian forest we’re gonna hear a TED talk later so what the heck is remote sensing well it can be defined in the simplest terms as an ability to get information or detect something where you’re not it could be something very very close to you or something kilometers tens hundreds thousands of kilometers away potentially there different types of Road sensing and there are different sensors that are used to accomplish this but first let’s take a step back how do they work well they operate they receive information somewhere on the electromagnetic spectrum well we all have eyes and ears and noses and mouths and we have five senses we can feel taste touch etc well and we can hear in a certain range right audio frequencies about 20 Hertz to 20,000 Hertz vibrations similarly we have electromagnetic radiation at different frequencies and these different frequency areas have different properties we’ll talk about those in a second but we only can see a very very small piece of that spectrum there’s a washed-out picture here I apologize for that but it’s such a gorgeous day how could we actually shut the blinds right but microwaves which you might have a microwave oven this operates in this region I can penetrate through clouds you can use a microwave imager to see something even if it’s a cloudy day infrared waves can detect heat which is a sign of life also terahertz waves can detect molecular composition of things so pretty cool right and their pros and cons teach its most basic form remote sensing instrument to text waves something waves hit it it detects them in some way shape or form and then it produces information based on that some of these are passive systems meaning they just listen or look some are active where they actually send out a wave it’s gonna reflect scatter do other things come back so way to think about that is if you have a camera a passive systems you’re relying on sunlight or ambient light in the room you take a picture an active system well if you have your Flash taking a picture in a darkroom you provide the light that then reflects off your subject and it comes back to the lens so there are different types of senses we’re gonna talk about three but there are many many others we have optical this is probably the most familiar to folks in this room who here’s used Google before Google Maps Google satellite view those are all optical images right and they measure light in near the visual range so you can be hyperspectral meeting different points in the spectrum microwave remote sensing is in the microwave range I mentioned passive sensors well a Radiometer is a passive sensor that detects the microwave

emissions that we all produce and then radars for active systems where they actually provide that microwave radiation and then detect look it bounces back you can also use lasers sadly not on dolphins or sharks or anything else yes that was a Mike Myers reference but you can use lasers and they detect among a variety ways to do this but in essence the amount of light that bounces back after you send out a little pulse of light so here’s some applications why you know why is this even important well for agriculture believe it or not forestry planetary exploration navigation if you’re driving around right a lot of the newer cars actually do have radar systems and radars great at ranging as well as imaging global system on earth so the carbon cycle the water cycle climate and also there’s some government applications some that we might like more than others but GIS municipal state and federal map building terrain all that jazz and also there’s some military applications as well which we’re not going to talk about today but for agriculture you can monitor crop growth it’s useful right if you’re a farmer they give a lot of lands you have an airplane you know fly over give you a very quick understanding of what you have and what’s going on also monitor soil moisture well governments particularly the US federal government likes to give out subsidies but there are restrictions maybe if they give you a subsidy you can’t grow this or you have to grow a certain thing this is an easy way to ensure that you’re complying also economic forecasts and do a quick flyover of the plane states see how much get an estimate for how much weed is growing let’s say and that impacts all the markets so very important and can be both airplane base and also a satellite based planetary exploration now I can talk too much about this but this is not limited to just the earth many of the you know spacecraft that we send out into space do have different types of remote sensors they need to be able to see and it’s one kind of an algae you could think of this image here is a combination of multiple radar images synthetic aperture radar images of a portion of Venus Venus for those of you have seen pictures before has a very thick atmosphere but again at microwave frequencies we can penetrate through those clouds and we can actually see some terrain here now you’re probably wondering well is the surface really that green and red and blue and all that no this is false coloured so ignore the colors but the colors do show differences right different regions different composition all of that very interesting stuff and we can use this to learn about similarities and differences of the earth and other planets say Venus or mercury or Mars or what have you we also use remote sensing locally here on the planet for disaster understanding so there’s an earthquake we can see and get a very good understanding of well how much did things move to what extent in urban areas what was damaged archaeology is something else is kind of cool because at lower frequencies you can penetrate certainly underground to some extent and also in vegetated areas like the Amazon you can kind of cut through that canopy without having to hurt the natural environment so this is an example from National Geographic basically showing an area in Egypt that on the surface looks perfectly mundane but underneath there’s cities so very cool stuff and there’s some action article in National Geographic from 2014 I encourage you to check out interesting stuff weather prediction I guess what other profession could you be in be wrong eighty percent of the time and still have a job but weather prediction extremely important and we have developed models and they’re getting better to actually predict the weather but these models need an inputs they need to know what’s actually happening in the world to say well what’s actually gonna happen in Ann Arbor in a couple of days and one of those inputs is the sea surface temperature well the image you see here is an average map of sea temperature for the month of May in 2001 this was taken by the moderate resolution imaging spectroradiometer or modis for short actives are fun and you can see that in the Pacific over here 2001 was an El Nino year so in the middle of the Pacific you can see sea surface temperature about 35 degrees C in a La Nina here that would be colder you can see it you know strong concentration of warm water that certainly impacts our weather here and around the globe as well so this is very cool stuff and I’d also note that this data is actually available to the public you can get

motus data and other data from the different satellites for free other systems you can’t but there are some that you certainly can so that’s a brief introduction we’re gonna take a quick break now I’m going to show you a couple of videos that were designed by the I Triple E Geoscience remote sensing Society I should note that these were designed more for school-aged kids but they’ll certainly serve our purpose as well the kind of fun and they really do a good job kind of both introducing this a bit more and will lead us into our specific application which I’ll speak to next how do we observe the Earth from space since the beginning of our civilizations humans have always needed to understand the world around them in order to survive this includes forecasting the weather assessing environmental hazards managing natural resources and improving understanding of the climate nowadays space provides us a unique vantage point for observing the earth together essential information about all these natural phenomena is a supply ship and aircraft base observing systems are also very important but their view is limited from space you can see the full picture of our planet to create accurate maps to measure and monitor environmental issues like rainfall deforestation of the land the rising sea level etc hi my name is wavy and in this video series I’m going to tell you how we observe the Earth from space we use remote sensors instruments that can measure radiation which is energy that is emitted by the earth itself and some that reaches the earth and reflects from its surface and goes out into space the way this energy is emitted or the way it changes after reflecting from the Earth’s surface is as important information about our planet for this purpose we cannot measure sound waves because they cannot travel in the vacuum of the space as electromagnetic waves actually do traveling in vacuum 300,000 kilometers each second however to reach the earth they have to pass through the atmosphere and they will only reach the earth if they belong to specific ranges of the electromagnetic spectrum called atmospheric windows these atmospheric windows are in the frequency bands of radio to microwave and wavelengths of infrared to visible light this is why for example we study atmospheric humidity also called water vapor in the bands near 22 gigahertz and 183 gigahertz or atmospheric temperature in the van near 60 gigahertz another example is the infrared band where the temperature of objects can be measured in this case infrared sensors are especially made to see the infrared radiation given off by objects even in the dark so I should know these are videos produced by gr SS I mentioned there are five videos told in the series really watching the first two but the other three and they’re all actually available in a variety of languages on if you go to I Triple E – G RSS org eventually will be linked remote senses remote sensors have been with us for many years they extend the abilities of our five senses allowing us to measure information that we cannot perceive with the five senses the electromagnetic spectrum is the range of all possible wavelengths of electromagnetic energy each object has its own particular signature wavelengths good resonances because that object reflects transmits and absorbs those frequencies in a particular way with remote senses we are able to measure this invisible electromagnetic energy which exists in different amounts at different wavelengths laughing remote sensing means acquiring information by measuring at a distance we use many different sensors which are sensitive to different wavelengths in the electromagnetic spectrum for example some sensors are designed to receive green wavelengths while others receive infrared wavelengths all remote sensors belong to one or two major groups

massive sensors that detect electromagnetic energy that already exists and active sensors that produce their own electromagnetic energy and measure its reflection actually your body is equipped with three passive remote senses your eyes see the reflected electromagnetic light that is emitted by the Sun your ears hear sound which are acoustic waves in air you’ll know smells odors produced by chemical reactions we also have one active remote sensor the voice by listening to its echo you can find information about distance from the object that reflects it to understand better the difference between passive sensors and active sensors you can think about a camera with a flash the camera will be an active sensor when the photographer uses flash in darkness because it illuminates its target and measures the light reflected back to the camera on the other hand the camera will be a passive sensor when the photographer does not use the flash because the camera is not providing the source of energy needed to capture the image in this case the energy comes from an external limiter like the Sun or lights in a room so one of the points note is another difference between active and passive sensing systems is resolution often times we see on TV there’s a it’s a picture and people just keep zooming in zooming and zooming in zooming and enhancing I suppose almost unlimitedly in real life that’s not true there is a natural limit right so every sensor has some limit that it says all right everything less than this amount obviously group that all together into a single pixel let’s call it well active mode sensors have a finer resolution so they can see much more detailed increments whereas those passive sensors are generally much more generally you can see a very large area but the resolution in those areas is relatively coarse so to walk away with from anything thus far that’s something useful to know so I’m going to move into our second phase of today’s talk a particular project I’ve conducted under the supervision of dr Leland Pierce professor kamal sarah von d again so for those of you that are bit more technically minded you can see very few equations if you want more come to my defense let’s talk about the global carbon cycle okay carbon for those that can see the pictures is emitted into the atmosphere from burning fossil fuels fire that jazz and other natural processes as well we convert carbon dioxide into oxygen in a variety of plants well how does that happen they actually take that carbon and ingest it in some sense that carbon then can reach into the groundwater we have a huge carbon store in our global ocean and it’s a big so basically a big cycle the question is well we need to know more about our forests to understand the carbon fluxes so how much carbon is coming in how much is coming out and that’s not an easy thing governments around the world in particular the US government and the Canadian government send out teams of fo every so often to do an inventory of their national forests the Canadian government usually sends thousand people out every few years it can be both dangerous benefits in the winter of the summer remote etc so this is an effort to gain similar understanding from the comfort of not the middle of the art of the Canadian Arctic so how are we gonna do that well we’re going to look at a particular area in Canada I’ll show you where in just a moment and we’re gonna try to simulate what these electromagnetic sensors might see make a big old database of possibilities and then when we look at something real try to look into our database to say well what looks the closest okay well this does well given that we have an estimate now for the type of trees we have their forest canopy height and also how much biomass is there so we’re gonna be looking right there what that pin is in the Boreas region southern study area this is something that NASA studied extensively in the mid 90s probably countable is abouts the countless but Capitol hours went into studying that in a variety of ways both with folks on the ground taking measurements as well as a variety of different remote sensing campaigns so zooming in a bit more hopefully the back projectors you can see a little bit better there are different sites within this region is relatively homogeneous but not entirely

there are under stories there are a variety of different things that could have going on there and that’s something that we’re gonna talk more about in just a moment so kind of zooming in you can see we’re gonna be focusing specifically in that area right there it’s a lot of Jack Pines so what are we gonna do I mentioned that we’re basically going to be looking at actual remotely sensed data generating our own simulated remotely sensed data and trying to compare the two and see what we can do that’s a preview we do pretty well so first off we need to be able to simulate what we see so we’re gonna use a process developed in-house despite another grad student 15 years ago to generate fractal trees using something called Linda Meyer systems so you can actually create a mathematical representation of a tree and then iterate it using a replacement system and actually get a very strong realistic representation of a tree here on the screen you see a red maple of the deciduous tree and then a pine tree the smaller one we can Foley and defoliate these at will we can change we know everything about them where every branches or stem therefore needle every leaf everything electric properties physical properties which lets us do some pretty cool things something I’m not talking about right now but I’ll give you a little note is we also in our lab look to see well how different physical phenomena affect different types of remote sensing in particular radar so using these trees we were able to develop a model for wind how that would interact with trees I’ll tell you I’ve never looked at a tree the same way again now that I know the specific mechanics but how does that affect a radar system or a Radiometer or a lidar system a laser system we can put these trees together and make a forest and we put together specific DNA for to match the Jackpine region the trees that we find up there there’s a pretty good match and I’ll say that we did test this looking at different types of remote sensing up there using our simulators and we got very close results our optical simulator we took advantage of something that’s already out there we’re nirvair hoff you create a sail model basically you English Wow uses four different optical paths and can very accurately estimate what one of those say Google satellites or in this case Landsat five might see we use that to calculate a leaf area index as well as other values and then from that we can calculate well we calculate our leaf area index from our known geometries feed that into the sail model with some other parameters and we can actually get out a variety of state of parameters what each at each band with a reflection of the top of the atmosphere as well as composite parameters such as the normalized difference vegetation index or NDVI for those that are later we have a radar model I’m not gonna go into too much detail here long story short we can approximate what a synthetic aperture radar might see at a given frequency and a given instance angle and we also have a lidar model which is a laser system shoots a pulse down measures the amount of reflection in time I will note here the lidar ‘s have a very limited footprint while the optical systems and the radar systems have a much larger footprint in comparison to orders of magnitude larger so we created these simulators and we generated over 8,000 possible stands say well let’s say we have an 8 meter canopy height and 10 kilogram per meter squared biomass go make a few of those now move on so we generated that long story short it’s a very densely populated area right the for those who can’t see the x-axis is the canopy height the y-axis is biomass long story short pretty well filled we have a digital elevation model of our area basically that means what’s the what’s the terrain doing what’s the height relative to some constant so this is a Landsat image and we’re going to overlay a few things on it this is so this is optical those purple lines indicate the extent of SAR images radar images you can see they’re pretty much covering a wide variety of things this one for example relatively wide well I’m gonna overlay now the lidar data this

was taken I will say was airplane based the last two certainly been satellite based but the lidar footprint on the ground is about between six and nine meter diameter circle so the red lines just flight paths there are a number of shots along these lines but as you can see it goes it’s nowhere near covering the whole area well what would the lidar see again I’ve mentioned it kind of records the reflected light here’s an example we look at the ground response and then it’s been above it would be from the canopy and we can estimate if a canopy height by just knowing how fast the wave travels and kind of the difference between where these Peaks are there are a variety of different methods that is certainly one so we looked at some ground data but there was an issue the folks that went out into the field measured the canopy height in five meter bins meaning I could only say with a five metre certainty well this area is between twelve point five and seventeen point five meters tall which it’s kind of nice but gives you this sort of canopy height and if you we want an accuracy of five meters that can be done currently with the present technology there’s really no reason to proceed so we wanted to come up with a better option so we basically took a topographic map as well as some data from the shuttle radar topography mission back in 2000 and create our own approximation of what the ground type might be here’s the math I mentioned using a linear combination of a few different tree parameters we developed allometric equations for trees themselves and now we had ground data we had information about the stands themselves and then we looked at our large area and Co registered all of our different data so our radar data our lidar data our optical data so everywhere these overlapped we put them together and then looked at each region each stand that was homogeneous and said well which stand in the database is the closest match based on different parameters well what parameters did we use for our lidar instrument we had a mean canopy power as well as the you have powers between the ground and the canopy return for our SARS system our radar we had as polar metrics so variety polarizations to work with and then for our optical system we had a variety of channels right bands as well as thermal channel which is great kind of looking at heat and then a composite channel and DVI so we looked at each of these and said well let’s just go through using a single parameter and see you know let’s find for the thermal Channel let’s look and see which stand in the database has the closest just number right and look at that and then say the one the Stan with the closest number let’s pick that it’s a very very simple approach and we said well okay there’s gonna be some error associated with that and we did that for both the height the biomass and all these different parameters and you can see some things work better than others so for example the thermal optical channel produced a almost 1.95 meter our mean square error for the height estimate not too bad at the same point another optical channel optical channel 7 produced a over three meter Armus error so the thermal channel is a bit more correlated with the height you could say than the other so we went through and selected four of these parameters that indicated strong correlation with the parameters we’re looking to see a fifth parameter the lidar height estimate was used as a bounding parameter so before to limit our database we said well we have a lighter height estimate from our measured data of 10 meters let’s say that we’re only going to look at stands in our database with a lighter height estimate within say 2 meters of that so from 8 to 12 meters it’s a way to reduce the search area and it actually helped improve our results a bit as well doing this we then optimized our are basically if we want to use a linear equation to figure which stands the bestest selection equation so we used a mostly regression to see well what makes the most sense we picked a subset a random subset of our stands 25% and came up with these weights basically saying that the radar horizontal copal channel is extremely important for us at the same point

though the optical channel of the the thermal optical channel is important our NDVI kind of the ratio near IR to the red channel optically is less important and our lidar power is a bit more important than that using these weights we looked at all sixteen of our stands and we were able to actually actually estimate the canopy height to an ARMA Seraph about 1.6 meters and also a dry biomass – about 1.6 kilograms per meter squared which is actually very good and some plots of that just showing you know decent range and all that ideally we would have had more data points but the number of actual homogeneous regions in this area was smaller than we would have hoped there so that’s just a little bit of information about a project kind of taking a lot of math out if you have any questions on that I’ll be happy to answer them in a few minutes but this is a really interesting talk by Greg aster let him it’s about 12 minutes again while this is going on please do if you haven’t filled out the form on your table about topics you’d like to see us to touch on next semester please do we really want to hear from you technology can change our understanding of nature take for example the case of lions for centuries it’s been said that female lions do all of the hunting out in the open savanna and male lions do nothing until it’s time for dinner you’ve heard this – I can tell well recently I led an airborne mapping campaign in the Kruger National Park in South Africa our colleagues put GPS tracking collars on male and female lions and we mapped their hunting behavior from the air the lower left shows a lion sizing up a herd of Impala for a kill and the right shows what I call the in view shed that’s how far the lion can see in all directions until his or her view is obstructed by vegetation and what we found is that male lions are not the lazy hunters we thought them to be they just use a different strategy whereas the female lions hunt out in the open savanna over long distances usually during the day male lions to use an ambush strategy in dense vegetation and often at night this video shows the actual hunting viewsheds of male lions on the left and females on the right red and darker colors show more dense vegetation and the white or wide-open spaces and this is the view shed right literally at the eye level of hunting male and female lions all of a sudden you get a very clear understanding of the very spooky conditions under which male lions do their hunting I bring up this example to begin because it emphasizes how little we know about nature there’s been a huge amount of work done so far to try to slow down our losses of tropical forest and we are losing our forests at a rapid rate as shown in red on the slide I find it ironic that we’re doing so much yet these areas are fairly unknown to science so how can we save what we don’t understand I’m a global ecologist and an earth explorer with a background in physics and chemistry and biology and a lot of other boring subjects but above all I’m obsessed with what we don’t know about our planet so I created this the Carnegie airborne Observatory or CA oh it may look like a plane with a fancy paint job but I packed it with over a thousand kilos of high-tech sensors computers and a very motivated staff of earth scientists and pilots two of our instruments are very unique one is called an imaging spectrometer that can actually measure the chemical composition of plants as we fly over them another one is a set of lasers very high-powered lasers that fire out of the bottom of the plane sweeping across the ecosystem and measuring it at nearly 500,000 times per second in high-resolution 3d here’s an image of the Golden Gate Bridge in San Francisco not far from where I live although we flew straight over this bridge we imaged it in 3d captured its color in just a few seconds but the real power of the CIO is its ability to capture the actual building blocks of ecosystems this is a small town in the Amazon image with the CIO we can slice through our data and see for example the 3d structure of the vegetation in the buildings or we can use the chemical information to actually figure out how fast the plants are growing as we fly over them the hottest Pink’s are the fastest growing plants and we can see biodiversity in ways that you never could have imagined this is what a rainforest might look like as you fly over in a hot-air balloon this is how we see a rain forest in kaleidoscopic color that tells us that there are many species living with one another but you have to remember that these trees are literally bigger than whales and what that means is that they’re impossible to understand just by walking on the ground below them so our

imagery is 3d its chemical its biological and this tells us not only the species that are living in the canopy but it tells us a lot of information about the rest of the species that occupy the rainforest now I created the CEO in order to answer questions that have proven extremely challenging to answer from any other vantage point such as from the ground or from satellite sensors I want to share three of those questions with you today the first question is how do we manage our carbon reserves in tropical forests tropical forests contain a huge amount of carbon in the trees and we need to keep that carbon in those for us if we’re going to avoid any further global warming unfortunately global carbon emissions from deforestation now equals the global transportation sector that’s all ships airplanes trains and automobiles combined so it’s understandable that policy negotiators have been working hard to reduce deforestation but they’re doing it on landscapes that are hardly known to science if you don’t know where the carbon is exactly in detail how can you know what you’re losing basically we need a high-tech accounting system with our system we’re able to see the carbon stocks of tropical forests in utter detail the red shows obviously closed-canopy tropical forest and then you see the cookie cutting or the cutting of the forest in yellows and greens it’s like cutting a cake except this cake is about whale deep and yet we can zoom in and see the forest and the trees at the same time and what’s amazing is even though we flew very high above this forest later on in analysis we can go in and actually experience the treetops leaf by leaf branch by branch just as the other species that live in this forest experience it along with the trees themselves we’ve been using the technology to explore and to actually put out the first carbon geographies in high resolution in faraway places like the Amazon basin and not so far away places like the United States and Central America what I’m going to do is I’m going to take you on a high resolution first time tour of the Carbon landscapes of Peru and then Panama the colors are gonna be going from red to blue red is extremely high carbon stocks your largest Cathedral forests you can imagine and blue are very low carbon stocks and let me tell you Peru alone is an amazing place totally unknown in terms of its carbon geography until today we can fly to this area in northern Peru and see super high carbon stocks in red and the Amazon River and floodplain cutting right through it we can go to an area of utter devastation caused by deforestation in blue and the virus of deforestation spreading out in orange we can also fly to the southern Andes to see treeline and see exactly how the carbon geography ends as we go up into the mountain system and we can go to the biggest swamp in the Western Amazon it’s a watery dreamworld akin to Jim Cameron’s Avatar we can go to one of the smallest tropical countries Panama and see also a huge range of carbon variation from high in red to low and blue unfortunately most of the carbon is lost in the lowlands but what you see that’s left in terms of high carbon stocks and greens and reds is the stuff that’s up in the mountains one interesting exception of this is right in the middle of your screen you’re seeing the buffer zone around the Panama Canal that’s in the reds and yellows the Canal authorities are using for us to protect their watershed in global calm this kind of carbon mapping has transformed conservation and resource policy development it’s really advancing our ability to save forests and to curb climate change my second question how do we prepare for climate change in a place in the Amazon rainforest let me tell you I spend a lot of time at these places and we’re seeing the climate changing already temperatures are increasing and what’s really happening is we’re getting a lot of droughts recurring droughts the 2010 mega drought is shown here was red showing an area about the size of Western Europe the Amazon was so dry in 2010 that even the main stem of the Amazon River itself dried up partially as you see in the photo in the lower portion of the slide what we found is that in very remote areas these droughts are having a big negative impact on tropical forests for example these are all the dead trees in red that suffered mortality following the 2010 drought this area happens to be in on the border of Peru and Brazil totally unexplored almost totally unknown scientifically so what we think as earth scientists is species are going to have to migrate with climate change from the east in Brazil all the way west into the Andes and up into the mountains in order to minimize their exposure to climate change one of the problems with this is that humans are taking apart the Western Amazon as we speak look at this hundred square kilometer gash in the forest created by gold miners you see the forest in green and 3d and you see the

effects of gold mining down below the soil surface species have nowhere to migrate in a system like this obviously if you haven’t been to the Amazon you should go it’s an amazing experience every time no matter where you go you’re gonna probably see it this way on a river but what happens is a lot of times the rivers hide what’s really going on back in the forest itself we flew over the same river image the system in 3d the forest is on the left and then we can digitally remove the forest and see what’s going on below the canopy and in this case we found gold mining activity all of it illegal set back away from the river’s edge as you see in those strange pock marks come up on your screen on the right don’t worry we’re working with the authorities to deal with this and many many other problems in the region so in order to put together a conservation plan for these unique important quarters like the Western Amazon in the Andes Amazon corridor we have to start making geographically explicit plans now how do we do that if we don’t know the geography of biodiversity in the region if it’s so unknown to science so what we’ve been doing is using the laser guided spectroscopy from the cao2 map for the first time the biodiversity of the Amazon rainforest here you see actual data showing different species in different colors reds are one type of species blues are another and greens are yet another and when we take this together and scale up to the regional level we get a completely new geography of biodiversity unknown prior to this work this tells us where the big biodiversity changes occur from habitat to habitat and that’s really important because it tells us a lot about where species may migrate to and migrate from as the climate shifts and this is the pivotal information that’s needed by decision-makers to develop protected areas in the context of their regional development plans and third and final question is how do we manage biodiversity on the planet of protected ecosystems a great question but one that you’ll have to finish this yourself you want to answer so with that the few minutes we have left I will take questions in the minutes have a wonderful rest of your day however if you have optogenetics then you can genetically modify neurons in a way that some specific target cells are only responding to a wavelength different colors so we are developing the next generation