Applications of Remote Sensing in Monitoring Ecosystem Function and Biodiversity

RAN WANG Author

02/24/2023 Added

30 Plays

Description

Research Assistant Professor and Image Processing Specialist, School of Natural Resources, University of Nebraska–Lincoln

Searchable Transcript

Search:

[00:00:00.810]The following presentation
[00:00:02.250]is part of the Agronomy and Horticulture seminar series
[00:00:05.820]at the University of Nebraska, Lincoln.
[00:00:08.310]Welcome everyone to the Department of Agronomy
[00:00:12.450]and Horticulture spring 2023 seminar series.
[00:00:15.930]This is a second week of the series.
[00:00:17.790]Last week Steven Knezevic from here in the department
[00:00:21.572]gave a great talk on soybean responses to Dicamba.
[00:00:26.790]Next week we have one of the department's outstanding
[00:00:31.500]graduate students, Milos Zarich,
[00:00:34.200]giving overview of some of the industrial
[00:00:37.590]hemp research out at North Platte.
[00:00:41.010]So be sure to tune into the website
[00:00:44.421]for details on that as well as other,
[00:00:47.340]other seminars throughout the series.
[00:00:51.020]But today I'm honored to be introducing our speaker,
[00:00:57.060]Dr. Ran Wang from the UNL School of Natural Resources
[00:01:03.960]as well as the Center for Advanced
[00:01:06.480]Land Management Information Technologies.
[00:01:09.450]Ran is an environmental remote sensing scientist there and
[00:01:14.640]as I'm a spatial scientist and as a spatial scientist,
[00:01:18.030]I'm always glad to be in the company of good remote sensing
[00:01:22.950]scientists because as many of you I'm sure also aware remote
[00:01:28.590]sensing is a course of science in its own right.
[00:01:31.650]But there are so many diverse applications that touch on so
[00:01:36.000]many things we all do from of course digital
[00:01:39.457]agriculture, range on monitoring, ecological restoration.
[00:01:45.930]The list could go on and on and I don't think Ran will
[00:01:49.530]probably talk about all of those things today,
[00:01:52.470]but maybe touch on on a few of them,
[00:01:54.930]especially as they have a relevance to to departmental
[00:01:59.250]research as well as here on East Campus. So with that,
[00:02:03.120]I will turn it over to you Ran.
[00:02:04.950]I will note that those of you online please feel free to
[00:02:09.000]enter your questions into the,
[00:02:11.010]it's the Q&A, or the Q&A box and we'll address those
[00:02:14.040]at the end. We also have these
[00:02:18.420]handheld mics that, for those of you in the room,
[00:02:21.420]I'll pass it around to you so you can pose your questions
[00:02:25.260]so that those online can hear.
[00:02:27.900]And with that, I'll turn it over to Ran.
[00:02:30.030]Thank you.
[00:02:30.863]Thanks Dan for the invitation and for the introduction and
[00:02:33.667]I'm really glad to be here to share some of the research.
[00:02:37.890]I don't wanna say I do.
[00:02:39.330]So we do together and we have been working together for a
[00:02:43.680]while so we have a lot of fun and I'll also share some of
[00:02:46.527]the work we do in this talk.
[00:02:49.590]So I'm working for SNR and also the most center.
[00:02:54.034]It is quite impressive that can't remember the whole name.
[00:02:57.179]I couldn't even remember that by by myself.
[00:02:59.070]But I call it here.
[00:03:01.740]Seeing the on scene using this guy we call the hyperspectral
[00:03:06.240]remote sensing or imaging spectrometry to apply that to
[00:03:11.460]detect or monitor ecosystem function and biodiversity.
[00:03:14.730]I'll share some of the work we do during the past few years,
[00:03:19.080]see if it works. Okay, before we start it,
[00:03:22.080]let's get the idea of get this, the definition of remote
[00:03:25.710]sensing clear. So when we talk about the remote sensing,
[00:03:29.370]I mean all the method we use result in contact is the
[00:03:33.780]target. So we sense that from a distance,
[00:03:36.900]like the eyes we use every day.
[00:03:38.520]So that's kind of remote sensing and we use the
[00:03:41.310]electromagnetic energy and we measure the reflected or
[00:03:46.230]emitted energy of the earth.
[00:03:48.060]So that's the sun and the sun gets energy towards the earth
[00:03:51.450]and the earth will reflect some of the energy and some parts
[00:03:54.715]for certain wavelengths
[00:03:56.790]The target on objects on earth emit their own energy.
[00:04:00.330]So we use all kind of this different instrument or platforms
[00:04:04.610]to measure this two type of the energy
[00:04:08.220]at different wavelengths.
[00:04:09.810]So this part is the light. We can see use our bare eyes,
[00:04:15.180]that's from half 400 to 700. So that's a visible light.
[00:04:18.960]But other, this instrument can measure the energy beyond
[00:04:22.470]this kind of narrow range.
[00:04:25.200]So we can go from here to somewhere here for the remote
[00:04:29.370]sensing part and we can do that using satellite.
[00:04:32.820]That's probably the easiest thing when we say remote sensing
[00:04:36.957]is like this figure is showing how satellite in space,
[00:04:40.003]compares this energies.
[00:04:41.940]But we can also use like regional aircraft.
[00:04:45.120]We use drones.
[00:04:45.953]I know a lot of people in this department fly drones and
[00:04:48.840]also we collect this reflected data or the energy on the
[00:04:52.200]ground. Even use this data guide leaf clip,
[00:04:54.630]we put that on the this.
[00:04:57.270]So that's,
[00:04:58.103]I try to go to go use the broader definition of a remote
[00:05:01.500]sensing, not narrowed down to one particular thing.
[00:05:06.450]Okay. Then among these technologies,
[00:05:09.150]there's one I particularly like that they call imaging
[00:05:13.260]spectroscopy. So basically we combine two things together.
[00:05:17.280]I don't know if I can move this program, I shouldn't do it.
[00:05:22.320]Okay, here we go.
[00:05:24.690]So the imaging spectroscopy, we put,
[00:05:27.480]basically we combine two things together. One is image.
[00:05:31.080]So we take image like we use a camera,
[00:05:33.510]we already get a fancy image like this one.
[00:05:37.020]And the second part is spectroscopy.
[00:05:39.690]So for each of this single pixels,
[00:05:42.120]unlike the RGB image we take with our cell phone,
[00:05:45.270]we have red, green, blue here,
[00:05:47.430]we have hundreds or thousands of narrow wavelengths
[00:05:51.060]because that's spectral and that's a definition
[00:05:54.450]for this thing.
[00:05:55.283]And we'll talk about this again and again in the next 40, 45
[00:05:59.250]minutes. That's a technology we use here.
[00:06:03.390]Okay, the first thing I want talk about is the
[00:06:06.570]remote sensing of photosynthetic activity.
[00:06:09.960]So the idea is pretty simple.
[00:06:12.090]Let's say we have this kind of flat towers on the ground
[00:06:15.664]and using these flat towers,
[00:06:18.514]it can measure how much CO2 goes through the instrument in
[00:06:22.410]the middle and then absorbed by the ecosystem like the
[00:06:26.730]grassland or forest.
[00:06:28.965]So here we see the CO2 goes through that and then we measure
[00:06:31.620]the CO2.
[00:06:32.453]Then we have a way to evaluate the productivity of the
[00:06:35.460]ecosystem. And then we have these studies
[00:06:38.130]all over the world.
[00:06:40.106]We have crop fields in Nebraska, we have crop,
[00:06:41.940]we have the start crop fields and we have grassland,
[00:06:44.717]we have forests and border forest.
[00:06:47.550]The idea is pretty simple if we can link this guy,
[00:06:51.207]the CO2 measured by this ethical variance instrument to
[00:06:56.130]either our environments and that's an optical sensor we
[00:06:59.820]measure here if we can link this two,
[00:07:02.400]we build a relationship between remotely sensor data and the
[00:07:06.630]flux data,
[00:07:07.463]the CO2 to the productivity measured by this ethical
[00:07:10.470]variance data.
[00:07:12.390]Then we can apply that to aircraft or satellite remote
[00:07:16.170]sensing data.
[00:07:18.270]We let this thing fly over the earth,
[00:07:20.910]we got the estimation of the global photosynthesis of
[00:07:26.130]productivity. The idea is very simple.
[00:07:29.550]So here today I want to focus on this side because we are
[00:07:33.450]here, right?
[00:07:34.283]So that's me in Nebraska,
[00:07:35.790]the corn fields to illustrate this idea of the work we do.
[00:07:41.040]So, some people may start think about, okay,
[00:07:43.380]you mean remote sensing?
[00:07:44.640]You must want to talk about something about the NDVI.
[00:07:47.940]Cool.
[00:07:49.230]Let's start from NDVI.
[00:07:51.300]So take, take advantage of two bands
[00:07:54.570]like reflections from two bands,
[00:07:57.030]one is the nearest red band and the other is a red band.
[00:08:00.477]And we combine them together.
[00:08:02.370]And then vegetation has a very different NDVI values
[00:08:05.670]from bare soil, from other background,
[00:08:09.630]let's say other targets.
[00:08:11.280]Then we can utilize this information and use that as a
[00:08:14.010]greenness index to pro predict or monitor for synthesis
[00:08:19.980]of JPG or whatever. It has been used for 30 years.
[00:08:24.210]So I don't want to bore you guys with a technology we
[00:08:27.030]developed in the seventies or eighties. That's not fun.
[00:08:30.247]Even this guy here, the chlorophyll index.
[00:08:33.930]Well this index was actually developed here at the UNL in
[00:08:37.951]the early 2000s. Late nineties to early 2000s.
[00:08:40.710]So it's very useful to predict or to monitor chlorophyll
[00:08:44.130]content for the crop fields at leaf and canopy value.
[00:08:48.240]So that's what we have already been doing, right?
[00:08:51.390]So we have the airplane fly over the crop fields,
[00:08:55.050]we calculate NDVI, we calculate chlorophyll index.
[00:08:59.790]Cool, we can do that 20 years ago, but it was new.
[00:09:05.587]So we do have something new, something fancier.
[00:09:08.580]So the idea these days is we use the technology or we use a
[00:09:12.690]new metric called the chlorophyll fluorescence.
[00:09:16.470]So the idea is when the sunlight come to the earth and when
[00:09:20.850]the sunlight gets to the plant, it lets the plant live.
[00:09:24.510]That's where the magic happens.
[00:09:26.790]So the sunlight is for system will turn the sunlight into
[00:09:30.476]something useful for us.
[00:09:32.370]That's the thing we call the photosynthesis.
[00:09:35.070]But this sunlight will actually have three different
[00:09:38.190]directions.
[00:09:39.660]One of the directions will just go to the photosynthesis,
[00:09:43.920]go through all this processing here at different scale,
[00:09:47.160]the small scale from within the cell and then to the
[00:09:51.750]canopy scale.
[00:09:53.340]And then the other,
[00:09:54.960]another way for the plant to deal with the light energy is
[00:09:59.850]rather go at heat, but we can see there is a third one.
[00:10:04.290]So that's a chlorophyll fluorescence.
[00:10:07.710]So if we can measure this chlorophyll fluorescence, we can,
[00:10:10.997]we can build a direct connection between this signal and the
[00:10:15.660]photosynthesis of plants at different scale from very small
[00:10:21.360]medical scale to canopy scale.
[00:10:23.727]And this technology is actually old technology
[00:10:29.370]and it's very powerful.
[00:10:30.690]We can use that to measure full sensor yield to use.
[00:10:33.660]People use that to detect water stress and nitrogen status.
[00:10:37.170]Like a lot of you here probably have done that as we have.
[00:10:40.530]We stand in the greenhouse, have a leaf clip,
[00:10:43.710]put that on the single leaf and press the button beep,
[00:10:49.260]we got the fluorescence right? That's simple.
[00:10:53.130]But that's at leaf scale.
[00:10:57.240]And we actually detect this signal beyond the leaf,
[00:11:00.930]beyond the one single leaf beyond the greenhouse from maybe
[00:11:05.430]a larger scale.
[00:11:08.250]The answer is yes.
[00:11:10.500]So here is one paper published in 2014 and they reported
[00:11:16.770]that we can measure chlorophyll fluorescence from space
[00:11:20.220]using satellite data.
[00:11:22.860]So that's the figure is gen,
[00:11:24.480]they generate GPP for products based on the fluorescence.
[00:11:30.510]And then here I just showed the maximum fluorescence image
[00:11:34.500]and then based on the previous studies
[00:11:37.440]and the European Space Agency,
[00:11:39.210]they started a special mission called
[00:11:41.820]Fluorescence Explorer Mission.
[00:11:43.920]And then this one is to design a satellite to particularly
[00:11:48.220]focus on mirroring chlorophyll fluorescence from the ground
[00:11:52.800]and use that to predict, to monitor global productivity.
[00:11:59.160]So with this in mind,
[00:12:00.990]and we can say that from the previous slide,
[00:12:04.620]we have four reasons married on a single leaf,
[00:12:07.950]but here we have a huge job from a single leaf to this guy.
[00:12:13.920]So the whole earth, the idea is pretty cool,
[00:12:17.040]but can we really do this or these guys or these values,
[00:12:21.240]do they actually mean anything or is there a relationship
[00:12:25.170]between the actual fluorescence and the GPP
[00:12:27.540]at the different scales?
[00:12:29.400]So there's the answers that we need,
[00:12:31.140]there's the questions we need to answer.
[00:12:33.341]We need to work on this to better use this information.
[00:12:37.140]Okay, with this in mind, European Space Agency,
[00:12:39.913]they have a project they call the Photo Proxy Project.
[00:12:44.610]The idea of that whole project is to understand
[00:12:47.400]at different scales,
[00:12:48.600]what is the relationship between peripheral fluorescence
[00:12:52.260]matter using remote sensing and the productivity
[00:12:55.739]or GPP of plants
[00:12:57.240]from single crop fields to large scales. And we,
[00:13:01.740]I mean UNL as the one site sitting in the middle
[00:13:08.252]participating in this project,
[00:13:09.984]but by us see this part that's the maximum fluorescens
[00:13:15.750]they measured derived from the satellite and this map was in
[00:13:19.510]2009, 2010 and they saw our spot here right in the
[00:13:26.550]middle of the North America, that's a corn belt area.
[00:13:29.940]So we have very high fluorescence values and we want to know
[00:13:34.230]why we have this, what happened.
[00:13:37.170]So that's an experiment, that's a PhotoProxy experiment.
[00:13:40.530]The idea is to test the fluorescence.
[00:13:42.690]And then we to do that,
[00:13:44.190]we have a series of different instruments. Put on
[00:13:47.995]the ground, the tower scale, we have airplane,
[00:13:52.737]we have this guy, the Hercules,
[00:13:54.060]we have some sensors on this side can measure reflected
[00:13:58.096]fluorescence.
[00:13:58.929]And then John was on the ground to measure the tops to
[00:14:01.590]calibrate the airborne data.
[00:14:03.420]So that's what we have in 2018 and that's the overall
[00:14:07.500]experimental design.
[00:14:09.030]And they selected Nebraska because they put this sentence in
[00:14:12.390]their project saying we have the patchy,
[00:14:15.390]agriculture landscape parentheses large fields.
[00:14:19.800]So you can imagine they have another series of this
[00:14:22.740]landscape, agricultural landscape in Europe.
[00:14:26.010]They will just change this to small fields.
[00:14:29.137]So that's the importance of our work contributed to this
[00:14:33.120]overall project. Okay cool.
[00:14:36.030]Now we know we can use this to do the work.
[00:14:38.430]We want to test the idea,
[00:14:40.110]see how different fluorescence we measure as a power scale
[00:14:43.560]from the air using this kind of a machine,
[00:14:46.710]how we can use that to monitor photosynthesis.
[00:14:49.680]The first thing we need to solve is how can we measure
[00:14:52.350]fluorescence from the air using remote sensor platform.
[00:14:56.250]This is a figure showing the black line and the red line of
[00:15:01.260]the solar radiation and the planet reflected the radiation.
[00:15:04.710]That's the energy.
[00:15:06.810]All this black and the red line,
[00:15:08.730]they are mapped on the left, y axis.
[00:15:12.960]And then I have a fluorescence spectrum and this value is on
[00:15:17.430]the right y axis.
[00:15:19.407]And you can see the difference of this signal,
[00:15:22.113]just the unit fluorescence is a very, very small signal.
[00:15:28.590]It's only about the less than 3% of the static state
[00:15:33.076]illumination.
[00:15:34.140]That means this signal is very, very big.
[00:15:36.330]How can we measure that? That's small signal.
[00:15:40.680]So if we add another figure,
[00:15:43.260]so instead of mapping the SIF or the fluorescence spectrum,
[00:15:48.120]I change that to to a leaf reflectance or the plant
[00:15:51.510]reflectance. That's a reflectance.
[00:15:53.820]We typical, say the purple one here,
[00:15:57.808]you say under this region the 760,
[00:16:01.080]this area we see a little bit little spike.
[00:16:06.660]So this was actually caused by fluorescence.
[00:16:09.930]So I believe many of us, when they measure the kind of,
[00:16:13.693]the reflections in the field.
[00:16:14.526]Sometimes we do see a spike in the reflectance at this area,
[00:16:18.840]but you may not pay attention to that.
[00:16:21.390]And sometimes this spike will be very, very small.
[00:16:24.420]Not like this big, like this tall.
[00:16:27.390]So why we can measure this fluorescence signal at this
[00:16:30.090]region.
[00:16:31.470]So the idea is, so we zooming here that spike, that's,
[00:16:34.920]that's the spike, the fluorescent signal.
[00:16:37.500]The idea is there are a lot of dark lines in the atmosphere.
[00:16:42.330]So see this black lines,
[00:16:44.040]remember these black lines are the solar radiation.
[00:16:47.040]So during within this wavelength or within this dark lines,
[00:16:52.320]the energy in the solar radiation will be absorbed by the
[00:16:56.526]earth's atmosphere.
[00:16:58.650]So there are very, very minor energy left in this area,
[00:17:03.990]in these dark lines.
[00:17:06.450]Then from these dark lines we can calculate or retrieve the
[00:17:09.720]fluorescence signal.
[00:17:11.580]So the idea behind this is actually very simple.
[00:17:15.060]For example, imagine we are in a very bright room,
[00:17:18.894]then I turn on my cell phone,
[00:17:22.890]you probably didn't see anything because
[00:17:25.800]this guy is very bright.
[00:17:28.320]But if I put my cell phone somewhere in a dark room,
[00:17:32.580]then you can see the light, right?
[00:17:35.790]That's the idea.
[00:17:37.860]If the atmosphere absorb all the energy or most of energy
[00:17:42.650]in this region and then the plants release
[00:17:46.650]a little bit energy to here, that's fluorescence.
[00:17:50.190]Then we can use this to calculate the fluorescence.
[00:17:56.040]But we need a instrument that has a very,
[00:17:59.730]very high spec resolution to detect this minor or this
[00:18:04.620]subtle signal in the reflected energy.
[00:18:09.780]That's why we have this guy here. We call this Ibis.
[00:18:13.260]So that's a imaging spectrometer have a very,
[00:18:15.510]very high spectral resolution.
[00:18:17.040]The spectral resolution is a 0.2.
[00:18:19.620]So you ready for the field spectrometer like ASB or spectral
[00:18:23.670]resolution.
[00:18:24.630]The spectral resolution is about three and but this guy is
[00:18:28.830]only one 10th of that with waste, twice it's is that was
[00:18:34.260]designed to measure this fluorescence signal.
[00:18:39.090]Okay then we need to use it. We have the instrument,
[00:18:41.093]we have our tool, we have our weapon, we need to use it.
[00:18:44.824]So we put that on the airplane and we fly it over Iraq.
[00:18:48.540]Finally look at this image.
[00:18:51.780]We have different crop fields in the Iraq experiment that
[00:18:55.749]CSB 1, 2, 3, 4 and LTAR.
[00:18:58.200]And for each of this we have one flex towers.
[00:19:01.740]Then we can use these fields to build a connection between
[00:19:06.600]ground for the censuses measured by added variants and
[00:19:11.370]airborne remote sensing signal.
[00:19:13.800]But here today I want to focus on this guy here,
[00:19:17.307]the CSP3, that's where we have most of our measurements.
[00:19:24.390]This was a picture I showed on the previous slides.
[00:19:28.380]That's a tower we have that's a flex tower behind this tower
[00:19:32.970]and there we put one instrument,
[00:19:36.000]optical signal has a very fine spectral resolution
[00:19:39.780]on this tower.
[00:19:41.790]They measure this area.
[00:19:44.250]So that's a continuous measurement.
[00:19:46.290]So it generates measurements every minute.
[00:19:50.520]So we put that in, in late May and keep that for the whole
[00:19:54.540]growing season.
[00:19:55.740]That provide us a continuous measurement of the
[00:19:59.790]fluorescence, the red line.
[00:20:01.860]And there's another detector we put on this instrument to
[00:20:05.940]provide us some reference.
[00:20:09.100]Let's say NDVI.
[00:20:10.649]So when we talk about remote sensing of vegetation,
[00:20:12.726]we have to use NDVI. That's our reference.
[00:20:15.060]Okay? So that's a continuous measurement.
[00:20:18.000]Then we fly these crop fields a few times in that year.
[00:20:21.450]In the summer we had a really good year.
[00:20:24.120]So that's,
[00:20:24.953]this starts showing the black points are the NDVI derived
[00:20:29.490]from the airborne data and the red points are the
[00:20:32.340]fluorescence derived,
[00:20:33.870]derived from the airborne data and from this figure we say
[00:20:37.529]it matched pretty well.
[00:20:40.820]But we see here at the beginning of the season,
[00:20:43.110]fluorescence NDVI, they follow the GPP, the blue line,
[00:20:47.070]the added variance.
[00:20:48.090]If you derive the data really well until here in the late
[00:20:52.890]season, NDVI stops changing.
[00:20:55.200]It keeps high for a period of time.
[00:20:57.788]But you say during the senescence,
[00:21:00.480]the GPP actually went down but the fluorescence,
[00:21:03.990]the red line tracked it very well.
[00:21:07.170]But you me ask the question, okay, that's NDVI,
[00:21:08.887]you said that NDVI was,
[00:21:10.620]old so we have to use some other indices like EVI, well,
[00:21:14.790]which is not young too young, too younger than NDVI.
[00:21:18.467]But okay let's compare those.
[00:21:20.867]We have fluorescence, we have NDVI.
[00:21:24.467]So the blue points here is actually the Hercules that
[00:21:28.350]machinery will running in that field.
[00:21:31.770]We run that few time in the summer and provide another point
[00:21:35.160]on the ground we compare instance and NDVI, EVI,
[00:21:39.990]that's these two operate old vegetation indices developed
[00:21:44.480]to study plants.
[00:21:46.080]And then we have the cardinal OID index that was designed to
[00:21:51.240]start a, the subtle changes in the plant pigment.
[00:21:55.890]So we map that and we have some even new young newer
[00:22:00.188]indices.
[00:22:02.520]This one is basically take the reflections
[00:22:05.370]from near infrared band and the time that
[00:22:07.474]by NDVI that's supposed to give us,
[00:22:11.190]well a better index for measuring photosynthesis than NDVI
[00:22:16.920]because that use the,
[00:22:18.900]the information in the near infrared band. So this guy,
[00:22:21.960]instead of using the reflectance in the near, near
[00:22:25.352]infrared band, they use the radius,
[00:22:27.780]that's the actual reflected energy in the near
[00:22:30.813]near infrared band.
[00:22:32.370]So we put a series of this indices together and try to
[00:22:35.801]make this time series figure across the season to see how
[00:22:40.260]they match the black line. That's the GPP we see here.
[00:22:44.430]Fluorescence did a decent job and this guy here,
[00:22:47.793]when you actually consider the reflected energy,
[00:22:52.200]it start to doing pretty well to track GPP at this seasonal
[00:22:57.300]scale.
[00:22:58.710]So what about that finer scales?
[00:23:03.000]So we have one day, so that's a diurnal plot.
[00:23:07.530]We have GPP as a black line and a different vegetation
[00:23:12.330]indices are the remote sensing products like fluorescence,
[00:23:16.522]NDVI, EVI, CCI and NIRv on the right
[00:23:19.770]as the red lines for each of this figure.
[00:23:23.010]The figure on the left side is the data from one single day
[00:23:26.585]that's in July.
[00:23:28.650]And the figure on the right side is the average the data
[00:23:32.610]from three days from July 9th to 11th in 2018.
[00:23:38.400]And we can see that fluorescence and NIRv especially
[00:23:44.580]fluorescence has the diurnal change of that crop
[00:23:48.930]field very well and NDVI, EVI, CCI,
[00:23:52.650]all this index plus this guy we can solve for scale.
[00:23:56.313]So that's at the temporal scale. We have seasonal scale,
[00:24:00.323]we have diurnal scale. But what about in space?
[00:24:03.960]So once we have the airplane,
[00:24:05.790]we have the airborne derived data,
[00:24:08.570]we can get these images of fluorescence on July, in July,
[00:24:13.170]August, September.
[00:24:14.730]So that's kind of during the second half of the growing
[00:24:17.580]season during the senescence.
[00:24:19.380]And then we also put the NDVI and the chlorophyll index
[00:24:22.650]image there as reference.
[00:24:25.260]We see here the fluorescence compared to these guys they,
[00:24:30.390]it is more sensitive to the minor changes within this field.
[00:24:35.250]And the thing I found interesting is this is is a soybean
[00:24:39.510]field, this is corn and this guy is oats.
[00:24:44.460]So here the oats were not planted during July and
[00:24:48.480]August and you see
[00:24:50.253]here the fluorescence is also very sensitive to the new
[00:24:53.580]plantation in September. That's the oldest crop fields.
[00:24:58.800]So from this study we can get like conclusion that the
[00:25:05.867]SIF or the fluorescence provide an excellent proxy of GPP
[00:25:10.800]and also the airborne data for fluorescence and other
[00:25:15.540]airborne products we derived can offer new approaches in
[00:25:20.820]estimating crop photosynthesis as different spatial
[00:25:24.660]and temporal scales.
[00:25:26.490]So here I think the airborne data or the study we do
[00:25:31.590]actually can bridge this whole area of using fluorescent to
[00:25:36.780]study plants at a different scale from very small scale and
[00:25:40.620]bridging that to the whole ecosystem scale.
[00:25:44.940]So the second topic I want talk about is a remote
[00:25:47.053]sensing of biodiversity.
[00:25:49.290]So this is a rock storm planet boundary figure from 2009,
[00:25:54.399]2009 and these guys,
[00:25:56.160]they developed the framework saying we need a SIF space to
[00:26:02.340]op, for human operating with the respect
[00:26:05.630]of the earth system.
[00:26:07.950]They hear that's a climate change,
[00:26:09.810]but we hear this every day,
[00:26:11.370]climate is important but see this guy,
[00:26:14.335]the biodiversity loss that's even worse in terms of this
[00:26:19.080]planetary boundary.
[00:26:22.200]So the, when I see biodiversity,
[00:26:24.300]so I mean also a broad definition of biodiversity.
[00:26:27.960]That means diversity,
[00:26:29.340]especially diversity, ecosystem diversity, and genetic
[00:26:31.703]or phytogenetic diversity.
[00:26:35.160]So biodiversity and the ecosystem functioning, biodiversity
[00:26:38.250]is, is important.
[00:26:39.330]Why do we care about that?
[00:26:40.740]So people make an argument and from this early work,
[00:26:43.890]the pioneer papers,
[00:26:45.090]it gets a conclusion that biodiversity can support
[00:26:49.290]ecosystem productivity, stability,
[00:26:51.570]resilience and so on so forth.
[00:26:54.627]And we have a meeting,
[00:26:56.490]a huge conference last December the UN Biodiversity
[00:27:00.420]Conference that's in Montreal, Canada.
[00:27:02.310]If you have been following this meeting,
[00:27:03.810]I know some of our students and our colleague went to this
[00:27:06.360]meeting and the UN has a news that saying the countries
[00:27:10.620]reach the reach,
[00:27:11.910]the historical agreement of biodiversity. If you are
[00:27:15.661]interested in that COP 15 report is online.
[00:27:20.370]So one more thing I want to say.
[00:27:22.050]If we don't care about biodiversity, what may happen?
[00:27:25.740]Some people think that we are having the very exciting
[00:27:28.898]events that called the sixth extinction.
[00:27:32.400]So if we don't care about this, this thing,
[00:27:35.070]if we do don't do anything,
[00:27:36.750]probably we can remove this guy here and put something else
[00:27:40.350]on that.
[00:27:43.380]Then measuring biodiversity in nature.
[00:27:46.410]So it's a very funny,
[00:27:47.670]very interesting work and we send the people to the field to
[00:27:51.496]gather this plant species,
[00:27:53.580]the prairie species back to the lab and sorting them,
[00:27:57.240]reading them, drying them.
[00:27:59.790]This is fun.
[00:28:00.623]You can sit there for days to do this work or if we want to
[00:28:04.350]count the species in the forest,
[00:28:06.540]like we want to count every single tree we can have this
[00:28:09.690]total station that's one of the forest geo plot developed by
[00:28:13.943]professor Sabrina Russo in Indiana cave.
[00:28:17.280]So we see every single tree from this little hole and write
[00:28:20.490]down the number like the coordinates where the tree is
[00:28:24.232]and write down the species of that tree.
[00:28:26.430]Imagine if you need to count let's say 10,000 trees in
[00:28:30.000]a summer. That's a fun job to do, right?
[00:28:33.600]So it's very expensive,
[00:28:35.130]very time consuming, then people start to think about,
[00:28:37.920]well can remote sensing help?
[00:28:39.240]Very easy guys, we have to find something to do this
[00:28:42.600]tedious work for us. Well the simple answer is yes.
[00:28:48.150]The the idea behind this whole idea is,
[00:28:50.900]whole remote sensing of biodiversity
[00:28:52.170]is now these plants, they measure the reflected energy
[00:28:56.070]and you see here for different species and for different
[00:28:59.640]targets on the ground, if we put, plot the reflections,
[00:29:04.230]the reflected energy we see actually very different
[00:29:08.190]reflections.
[00:29:09.390]So they say, or how's scoring it, and maybe this part is
[00:29:14.460]something in the shadow and they say some soil spectral but
[00:29:19.080]they're very different.
[00:29:20.310]So if we can utilize this information,
[00:29:22.500]the reflected energy to differentiate the different species
[00:29:27.090]and the possibility of some other diversity,
[00:29:29.759]we can use that map bio diversity at the different levels.
[00:29:33.030]Okay, sounds cool but, there're always a but, unfortunately.
[00:29:38.100]So here is Cedar Creek, that's an experiment.
[00:29:42.304]Dave Tillman and his colleagues set up this this experiment
[00:29:46.710]in 1994, 1995.
[00:29:48.930]So each of this plot is nine meter by nine meter and they
[00:29:52.650]plant different prairie species within each plot and then
[00:29:56.490]study, the relationship between biodiversity
[00:29:59.190]and ecosystem function.
[00:30:00.840]Some of the professors at the UNL was were also involved
[00:30:04.429]in this experiment like they they in SNR and that's the
[00:30:10.980]early study to have the foundation for the relationship
[00:30:15.270]between biodiversity and the productivity.
[00:30:18.360]And what we do is we utilize this plus because they set up a
[00:30:23.580]great gradient at a different levels from monocultures to
[00:30:28.800]high diversity plots like 16 different plants in one plot
[00:30:33.000]and we measure reflected energy from ground and from the air
[00:30:36.630]and to see if we can see a relationship between the
[00:30:40.470]reflectance and the ground biodiversity.
[00:30:44.850]So that's what we did. But see here, once we have this guy,
[00:30:49.380]we put the one imaging spectrometer
[00:30:52.320]so that can generate the image
[00:30:54.540]and also the spectral for each pixel at a very fine scale.
[00:31:00.360]The peak size of this image is one millimeter like this big,
[00:31:04.710]very, very small from this plot.
[00:31:07.460]And that's one meter.
[00:31:09.832]If we scan that from the air, that becomes a one pixel,
[00:31:15.600]that's one meter.
[00:31:17.010]And you see the spectral variation in this image here it's
[00:31:20.760]pretty big but it goes to here, it's narrowed down,
[00:31:25.110]the variation is getting smaller but that's from the air.
[00:31:29.460]What about the from satellite,
[00:31:31.650]that's the Sentinel satellite, the European satellite,
[00:31:34.620]the pixel size is 10 meter, it's not that bad.
[00:31:38.130]Well the landside has 30 meter models,
[00:31:40.080]have 250 and even 1K spatial res, spatial resolution.
[00:31:44.940]Then 10 meter is not big but remember the size of this plot,
[00:31:50.490]that's nine meter by nine meter.
[00:31:52.920]One plot becomes the one pixel.
[00:31:56.520]How you still even detect biodiversity here? Probably not.
[00:32:02.910]That means there's a scale dependence between the remotely
[00:32:07.080]sensed biodiversity and the biodiversity on the ground.
[00:32:10.950]So that's a figure showing, that's Ceder Creek work we did.
[00:32:15.503]We have this one millimeter resolution and we degraded that
[00:32:20.340]to different visual resolution like different pixel size.
[00:32:23.730]Then calculate the correlation and build a relationship
[00:32:27.060]between the spectral diversity which is derived from the
[00:32:30.300]remotely sensed data and the biodiversity on the ground.
[00:32:34.110]And we see that with increasing pixel size this relationship
[00:32:39.480]starts to collapse.
[00:32:41.850]So here one millimeter,
[00:32:43.290]one centimeter, pretty good. It goes bigger,
[00:32:48.660]gone, nothing, that's 0.75 meter.
[00:32:53.160]And our colleagues here at the UNL,
[00:32:55.980]they applied another study.
[00:32:57.960]So they actually use the field work with the river,
[00:33:01.590]use the airplane to collect data as a different spatial
[00:33:04.950]scale. That means different pixel size.
[00:33:08.100]We found that at a half, half meter pixel size, they still
[00:33:13.740]say very good correlation of the group.
[00:33:15.930]Very strong relationship between all diversity and remotely
[00:33:20.490]sensed spectral diversity.
[00:33:22.440]But the similar pattern here,
[00:33:25.320]when they increase the pixel size,
[00:33:27.960]this relationship starts to be gone soon after four, five
[00:33:33.060]meter like when you increase the size to six meter, nothing.
[00:33:37.320]And there's another study we start, we did in Canada,
[00:33:41.490]we use the airplane and then we got a very decent
[00:33:44.520]relationship between spectral diversity and biodiversity at
[00:33:47.970]one meter scale that means we are saying something like at a
[00:33:52.440]different ecosystems this relationship may change.
[00:33:58.050]So we still don't know or we still need to study what
[00:34:01.200]actually drive this relationship between the remotely sensed
[00:34:05.670]data and the ground biodiversity.
[00:34:08.070]There's a lot of things to do.
[00:34:10.080]By the way this figure is from John's book chapter in 2020
[00:34:15.415]but it is a condensed almost all my four years PhD in one
[00:34:20.270]chapter, in one data set or in one figure
[00:34:23.300]in this book chapter.
[00:34:26.190]So that's my PhD.
[00:34:30.360]Okay, another thing we need to consider
[00:34:32.640]is that's a spatial scale.
[00:34:34.993]What about the time?
[00:34:36.930]So when we went to the field at the different season in
[00:34:40.035]early spring, summer, late fall we see different plants.
[00:34:44.910]So that's in the early spring, that's grouping,
[00:34:48.180]they start growing very pretty plants.
[00:34:51.660]And then later in the early summer,
[00:34:54.450]I remember this about in early June we start to see this
[00:34:59.220]very beautiful flowers, these purple flowers,
[00:35:02.580]they were everywhere.
[00:35:04.470]So if you measure the reflections of different leafs,
[00:35:08.250]these are the leafs and also the flowers you see very
[00:35:12.720]different reflections.
[00:35:14.910]So let's say you take the reflectance of the remote
[00:35:18.030]sensed environment at one time point here
[00:35:22.320]or here and build a relationship between remote
[00:35:25.860]sensing data and biodiversity.
[00:35:27.840]You will definitely see different relationships,
[00:35:30.780]that's the temporal scale.
[00:35:32.790]But comparing to the spatial scale,
[00:35:35.100]this part is less started.
[00:35:37.620]You can think about why because it's very hard to get multi
[00:35:42.480]temporal remote sensing and field sampling at one time
[00:35:46.140]all the time.
[00:35:47.520]It's just too hard to do that.
[00:35:50.820]So that's from the Cedar Creek experiment.
[00:35:53.820]That's a nine meter by nine meter experiment.
[00:35:56.730]Then let's see something we do here that was involved
[00:35:59.670]in this work and we learned that Dan and Craig
[00:36:04.680]and those folks,
[00:36:05.513]they do very fancy,
[00:36:06.480]very interesting study as a part of broader range.
[00:36:10.065]So say, well, probably we can join them and do something
[00:36:13.265]interesting together. So that's what we did the last year,
[00:36:17.279]last July, 2022.
[00:36:19.855]We fly the hybrid spectral and the thermal image over the
[00:36:23.714]Barta Brother Ranch.
[00:36:25.805]And let's focus on this part particularly N5
[00:36:29.580]and N7, we have different management regimes
[00:36:33.480]for these two plots and see if we can see a difference.
[00:36:38.820]That's a burned plot and five that's an unburned plot
[00:36:43.500]and seven, that's these two guys and five burned
[00:36:46.710]and seven unburned.
[00:36:49.710]That's a picture.
[00:36:50.610]So from this picture we can clearly see the difference,
[00:36:54.420]right?
[00:36:55.253]You can see this old feeders in they are standing in the
[00:36:59.250]unburned field and this is a burned field
[00:37:03.069]and if we went there in June
[00:37:05.190]and collect some ground reflections
[00:37:07.650]from the ground handheld spectrometer.
[00:37:11.734]So we see here, you see the blue line that's the five,
[00:37:17.804]that's the burned plot and the orange line in the middle,
[00:37:20.670]that's the unburned plot of the field.
[00:37:24.450]We see a clear difference of strong arbitral effects of
[00:37:29.045]burning say much higher reflectance of the burned plot.
[00:37:34.170]So if we calculate NDVI and the one industries we call
[00:37:39.120]coefficient of variation, which is an index,
[00:37:42.690]we use to measure spectral diversity.
[00:37:46.208]If we hold the idea that spectral diversity can be used to
[00:37:49.170]estimate real biodiversity or species diversity on the
[00:37:52.470]ground we have these two figures.
[00:37:55.350]This is NDVI and five and seven.
[00:37:58.920]So we can see a clear effects of burn on NDVI and the
[00:38:03.900]diversity, the spectral diversity.
[00:38:06.510]But to better understand this part,
[00:38:08.640]there are a lot of work we need to do.
[00:38:11.250]I'm just showing you some preliminary results here.
[00:38:14.040]We see this management effects but we still don't know
[00:38:18.390]how to interpret all our results.
[00:38:22.410]So we have this fancy figure,
[00:38:25.110]this paper was published the last year.
[00:38:27.660]They have built this fancy idea saying combining remote
[00:38:31.890]sensing and all kinds of ground data enable us to measure
[00:38:36.630]diversity at a different spatial and temporal scales.
[00:38:42.900]That's our goal. But how to get there,
[00:38:46.080]there's a long way to go. That's my idea.
[00:38:49.470]So the summary for this part is from the Barta Brothers
[00:38:53.863]work, the predominant result we see
[00:38:55.680]high productivity lead to
[00:38:59.010]cross abundance to high spectral diversity,
[00:39:01.530]the probably highest diversity on the ground and the low
[00:39:05.190]temperature from the air, the thermal data we measure.
[00:39:08.790]And the second part is we see a clear management schemes
[00:39:12.930]affect both diversity and productivity and all this stuff
[00:39:17.280]can be measured using airborne remote sensing.
[00:39:22.350]So I want the first,
[00:39:24.210]the previous part I want to focus on the crop fields
[00:39:27.840]and the grassland,
[00:39:29.160]but I will say a little bit more besides that because we do
[00:39:32.670]have another land power type here in Nebraska
[00:39:36.780]even it's not best.
[00:39:38.640]So this one is Nebraska Earth Observatory.
[00:39:41.760]So that's the airplane we have and we have two instruments.
[00:39:45.330]Those are the two instruments we use to get the remote
[00:39:48.480]sensing data. One is the IBIS,
[00:39:50.087]this guy was the designed to measure fluorescence.
[00:39:52.898]This one here, the red one and this guy is we call Castrol.
[00:39:58.020]So they offer the very conventional reflectance from the
[00:40:02.070]visible part to the near infrared band.
[00:40:07.350]So we have these guys, Barta Brothers,
[00:40:11.340]we have Nine Mile Prairie, Bobcat Prairie, that's a prairie
[00:40:14.850]site in Nebraska and we have this crop field and see this
[00:40:19.350]red ones, we do have forest.
[00:40:22.860]Don't forget that we have forests in Nebraska and we do
[00:40:26.820]start a forest in Nebraska.
[00:40:30.090]That's what we did.
[00:40:31.920]We fly over the Indian Cave forest geo plot.
[00:40:36.750]So Professor Sabrina Russo set up this plot.
[00:40:40.320]So in this region, this area,
[00:40:43.050]they spent a year to identify every single tree in this
[00:40:47.610]region and they know exact the location of the tree species
[00:40:51.470]of the tree and how big the tree is.
[00:40:54.540]And then we fly over and we calculate two things once we
[00:40:58.980]call the power beetle we use that as an indicator of,
[00:41:03.006]for the incoming solar radiation.
[00:41:05.400]And the second one is PRI.
[00:41:08.040]So that's a photochemical reflection index.
[00:41:10.770]We use that to to represent the photosensitive regulation of
[00:41:14.905]these different plants.
[00:41:15.810]So I don't have enough time to explain this PRI thing here
[00:41:18.990]if you have questions, guys, here.
[00:41:22.800]Okay, so this is at different locations
[00:41:26.220]of this landscape
[00:41:28.272]that's a valley and we go, that's a canyon.
[00:41:30.420]And then we go to top of the hill, you see different pattern
[00:41:34.200]like the higher core radiation on top of the hill,
[00:41:37.620]lower radiation or the lower energy down the canyon and
[00:41:42.930]see here how good PRI correspondence to this part.
[00:41:48.690]So the low values here and then go to high values here.
[00:41:54.330]And if we build this relationship,
[00:41:56.720]we actually measure kind of light and response of PRI and
[00:42:01.320]that can reveal the difference between different species and
[00:42:05.910]the different functional types.
[00:42:07.890]I put three different species here. You see they have very,
[00:42:12.603]very different responses in the PRI, land response,
[00:42:16.500]light response and then zoom in.
[00:42:19.436]These are the three figures for each species.
[00:42:22.680]And this paper just came out a few weeks ago as,
[00:42:27.147]I dunno if the official one is there, okay,
[00:42:32.100]so you can google this title and it's there and it is
[00:42:36.076]accepted version and then the official one probably will be
[00:42:38.940]here this week or next week, I dunno.
[00:42:40.830]So if we, you are interested in this work, please read.
[00:42:46.230]So we can also do the same thing with the Indian Cave.
[00:42:49.650]That's the natural landscape.
[00:42:51.870]We apply the same idea to place here. Our campus,
[00:42:57.270]that's a chlorophyll fluorescence,
[00:42:59.805]the fluorescence we calculated for all the campus trees.
[00:43:02.670]You see this guys the vegetation,
[00:43:04.980]the orange or the reddish color means higher fluorescence.
[00:43:09.450]This data was collected in 2018 August.
[00:43:13.170]That's a pretty hot day. And let's see the flex here,
[00:43:19.590]those are the trees that standing in the middle of the
[00:43:22.740]parking lot. So they definitely,
[00:43:24.900]they were not quite very happy during a very hot day.
[00:43:29.370]That's a fluorescence. So if we apply the idea,
[00:43:32.790]same idea about the PRI and the solar induced
[00:43:37.509]chlorophyll index in this chlorophyll fluorescence
[00:43:42.960]and apply the idea of light response curve
[00:43:46.800]to these two indices,
[00:43:48.887]we got a figure like this one. So for this two species,
[00:43:53.160]one a green and, one a green and one residual species,
[00:43:58.710]we see very different response,
[00:44:01.440]the light response of these two indices.
[00:44:05.280]And in that year we fly our campus, over our campus three
[00:44:09.720]times, August, September and October.
[00:44:13.740]So August that's kind of in the midsummer
[00:44:16.978]and then September, October,
[00:44:18.969]that's at a different time during the senescence
[00:44:20.790]you see this relationships changing
[00:44:23.460]and then we think we are looking at
[00:44:26.310]something really interesting that's like the response we can
[00:44:31.885]use to say and see the difference in photosynthetic
[00:44:35.672]activities between different species and different
[00:44:39.510]functional types.
[00:44:41.250]And that's not a typical work that people do or people use
[00:44:45.240]using remote sensing like the airborne work.
[00:44:48.240]So we are still trying to understand what we are seeing in
[00:44:51.660]this figure,
[00:44:52.830]but we can apply the idea to different landscape,
[00:44:56.430]different gravitation types and the relationship seems
[00:45:00.180]whole.
[00:45:01.740]So some take home messages from this whole presentation.
[00:45:06.390]So I think most sensing does offer a new and very powerful
[00:45:10.950]tool to monitor the earth.
[00:45:12.870]Especially we use the airborne data because very high
[00:45:16.680]quality data but the more efforts are needed to,
[00:45:19.590]interpret this data.
[00:45:22.200]That's why we need to gather everyone work together to help
[00:45:25.620]us interpret what we are seeing with this data set.
[00:45:29.889]So if you want to look at the plants in a different way,
[00:45:33.090]come fly with us.
[00:45:35.910]Thank you.
[00:45:37.410]Thank you Ran.
[00:45:38.640]We do have time for questions. Several minutes here.
[00:45:43.681]I'll go ahead and as I mentioned I have a mic I can pass
[00:45:48.690]around in the room.
[00:45:49.523]We may also have some questions coming in online.
[00:45:52.800]Maybe we can start with the room and give folks online a
[00:45:55.440]chance to type in. Any questions from here in the room?
[00:46:02.730]Noah?
[00:46:05.279]So I'm interested in VPD affects.
[00:46:07.830]So like how does VPD throw off indices?
[00:46:10.620]So in your last slide you showed east campus with SIF.
[00:46:15.750]Yep.
[00:46:17.070]But wouldn't you have to normalize
[00:46:20.130]or like classify out
[00:46:22.650]irrigated areas on east campus be, when you're evaluating
[00:46:26.670]photosynthetic activity
[00:46:27.990]because SIF can actually be thrown off by irrigation itself.
[00:46:35.850]And I do have one more question, just to, how in terms of
[00:46:41.303]resolution curves on going way back, how does that
[00:46:48.142]compare the spatial scale as it compares to canopy?
[00:46:53.220]So if you were to have like a very high resolution product
[00:46:56.070]like a planet super dove around a three meter range or a
[00:46:59.490]Maxar satellite around a one meter range,
[00:47:03.090]one to two meter range, in a non-sennesting timeframe
[00:47:09.840]like in a region a man-,
[00:47:12.300]a heavily forested mangrove region
[00:47:14.280]that has a semi high canopy
[00:47:16.710]but a very high biodiverse region.
[00:47:19.260]But that biodiverse region could actually be lower at
[00:47:21.840]the peat bog range.
[00:47:24.330]You, we could actually find that curve
[00:47:26.900]to be completely thrown off.
[00:47:28.629]So have you ever looked at the canopy as compared to spatial
[00:47:31.440]scale in terms of sensible,
[00:47:35.168]sensible species?
[00:47:39.960]Not framing that question correctly.
[00:47:41.940]Thank you. That's, yeah,
[00:47:43.710]there are very good questions actually.
[00:47:45.810]So let's start from here since we are on this slide.
[00:47:49.350]So I say second question first.
[00:47:51.990]So these relationships are built on prairie plants.
[00:47:55.800]So we'll definitely say different relationships and maybe
[00:47:58.410]apply that the same technology or the same idea to forest.
[00:48:01.710]That's the one work we have been working on using the Indian
[00:48:04.650]Cave data.
[00:48:05.483]We change the size of the like aggregation or whatever you
[00:48:10.530]want call that spectral diversity,
[00:48:12.540]the pixel grain size in remote sensing part.
[00:48:16.320]And also we change the shape of that sampling on the ground
[00:48:19.680]and from very small to very big and we see very different
[00:48:23.190]relationship between the spectral diversity and the species
[00:48:26.400]diversity. Like we,
[00:48:27.680]we count the species, region is for different trees, and at
[00:48:31.170]one scale we probably see one very strong relationship.
[00:48:34.860]But when when we increase that remote sensing in similar
[00:48:39.120]size or also decrease that the relationship goes through
[00:48:43.920]changes very fast.
[00:48:45.780]So we're thinking we are seeing something interesting that
[00:48:49.050]data set and,
[00:48:50.723]but again that's just from one forest site and we are still
[00:48:54.000]trying to figure out is that a magical number or the magical
[00:48:57.660]size that we say is something true or is just some arbitrary
[00:49:02.040]numbers. Yeah,
[00:49:03.090]I think that's a good question and we are working on that.
[00:49:06.390]I cannot give you an answer,
[00:49:07.770]but that's a interesting interaction.
[00:49:10.440]And the second part of this, trees,
[00:49:13.740]yes I agree for this campus trees we actually trying to
[00:49:16.570]cooperate with the people on campus who like watering or
[00:49:20.760]taking care of these trees to give us more information to
[00:49:23.790]help us interpret this data set.
[00:49:26.190]So the initial information we got from them was the trees
[00:49:30.150]probably more or less pretty happy.
[00:49:34.200]Well that's, that's what I got.
[00:49:35.880]So I ask about do you,
[00:49:36.890]do you have like more information about like which trees
[00:49:40.350]were watered and at which time they were watered and what
[00:49:45.420]about other things you did to the tree?
[00:49:49.017]I haven't heard anything about that yet.
[00:49:52.440]So I think that's an interesting question but we need like
[00:49:56.430]more information to interpret this.
[00:49:58.625]Thanks.
[00:50:00.300]My question goes back to one of your first slides
[00:50:03.300]where you showed that if there's excess energy
[00:50:08.430]that the plant can't
[00:50:09.360]use it either goes to heat or to fluorescence.
[00:50:13.350]Yep.
[00:50:14.489]And and plants are really quite specific
[00:50:17.190]in the ways that
[00:50:18.023]they can compensate for this extra radiation.
[00:50:22.020]So which comes first?
[00:50:23.850]It, the plant takes care of the extra heat by
[00:50:28.297]evapotranspiration and then fluorescence follows.
[00:50:33.990]Is that, is that the sequence that that happens?
[00:50:38.790]So, here I think we are talking about this figure.
[00:50:44.430]Almost the first one. There you go.
[00:50:46.980]So when we talk about this guy
[00:50:49.410]we are not talk about as the
[00:50:51.663]canopy or the canopy level, right?
[00:50:54.960]So here this for instance,
[00:50:56.340]they see this has a very fine scale molecular level.
[00:51:00.690]So this heat is not the thermal data we can measure,
[00:51:04.260]let's say at the temperature change. This heat is energy.
[00:51:09.510]Plants dissipate through the xanthophyll cycle.
[00:51:13.620]So that's some people when we use that use fluorescence to
[00:51:16.500]measure that. Some people call that MPQ.
[00:51:19.715]So that's not a sensible heat from the this guy.
[00:51:23.971]Okay. So, so is it true that plants
[00:51:25.680]have to be under a stress to fluoresce?
[00:51:28.770]No.
[00:51:31.200]Okay, a second question, back, you showed some
[00:51:36.420]fluorescence peaks at about 761. Is that,
[00:51:39.660]is that the Ron Heiser, Fran Hoffler band?
[00:51:44.280]Yep.
[00:51:45.450]Okay.
[00:51:47.070]So this value is accurate.
[00:51:48.840]So this is the SIF of the fluorescence data we can retrieve
[00:51:53.640]from this remote sensing platform.
[00:51:57.900]And this part is, this part is for Hoffler,
[00:52:03.870]that's the energy where the earth's atmospheres absorbs the
[00:52:09.180]incoming energy that makes the retrieval easier for us to
[00:52:13.470]gather this value. There are two,
[00:52:15.450]they do have other methods to retrieve fluorescence at
[00:52:18.840]different events like how we get these curves.
[00:52:23.036]But for now our work saying that the most stable part is
[00:52:27.630]just to focus on this, this thing here.
[00:52:30.780]If you pay a little bit pay attention to this range,
[00:52:34.084]this 680 you can can probably still a little spike here
[00:52:39.810]that's also caused by fluorescence this region.
[00:52:45.270]So that's respond to this area but we try some method to try
[00:52:50.970]to calculate this fluorescence as it is from our data is not
[00:52:54.960]stable so we want to focus on this one first.
[00:52:58.560]So that's what I have been doing. That's good question.
[00:53:01.260]Thank you.
[00:53:02.790]I don't know if I fully answer your questions.
[00:53:07.110]All right, well we,
[00:53:07.943]let's pivot to an online question and we have a question
[00:53:11.970]from Chris Helder about measuring biodiversity amidst
[00:53:17.790]grazing. So in a range on site, I'll read it fully here,
[00:53:22.710]exciting about the increasing ability to measure
[00:53:25.020]biodiversity in grasslands with remote sensing.
[00:53:27.420]But one big complication you haven't talked about here yet
[00:53:30.720]is grazing.
[00:53:32.430]So if a pasture is part of a four pasture grazing rotation,
[00:53:39.630]it would give a very different reflectance in different
[00:53:42.510]years as well as seasons.
[00:53:44.970]Do you think we'll ever get to the point where we can smooth
[00:53:48.060]across all that variability and measure biodiversity in an
[00:53:52.740]accurate way within grazed sites?
[00:53:56.430]Well that's a good question.
[00:53:58.706]I think that's one thing we need to start a,
[00:54:00.600]I don't think that there's a easy answer to this question.
[00:54:03.660]For example,
[00:54:04.493]if we say this is three different reflection lines that's
[00:54:07.770]actually from different fields. This blue one,
[00:54:11.130]that's the N5 and this like reddish one,
[00:54:15.270]that's the N7 and these two fields,
[00:54:19.023]they have grazing, have cattles there. But the third one,
[00:54:23.100]that's the yellow one, that's these guys,
[00:54:27.390]that's biocom complex experiment.
[00:54:30.000]So typically we can call it no grazing, no burn,
[00:54:34.680]and we see very different reflections from the N5
[00:54:39.668]and N7.
[00:54:40.980]I think that's a very good question and I just don't have an
[00:54:45.480]answer.
[00:54:46.313]I think there's a long way to go till we have a good answer
[00:54:49.250]to say well we can use reflectance if we get that signal.
[00:54:52.868]That's how much about diversity, that's the relationship
[00:54:55.320]will change according to when we get
[00:54:57.679]to that environment what instrument we have
[00:55:00.450]and what management is there.
[00:55:05.790]Good, thank you.
[00:55:07.410]Yeah, Chris always has great questions.
[00:55:11.010]One more online here from Sahib Amid. How could we use it to
[00:55:16.830]measure plant stress, what you've talked about here today.
[00:55:21.570]So there are a lot of different ways we can use to measure
[00:55:26.085]stress. I think that's kind of a very broad question.
[00:55:29.730]So, and also it's depending on the status of the stress.
[00:55:33.180]If you want to say,
[00:55:34.740]let's say that part has been dropped for a long time,
[00:55:37.110]all the plants change their,
[00:55:39.330]even the canopy structure or that's the early detection of
[00:55:42.750]the stress or if you want to measure stress
[00:55:45.172]for one single plant.
[00:55:46.170]And I see the experiments we do in the greenhouse or if you
[00:55:48.840]want to measure the stress using as
[00:55:51.396]let's say using satellite
[00:55:53.490]as a lot of scales.
[00:55:54.900]So I think there are few ways we can do that.
[00:55:57.960]Be like depending on what's,
[00:56:00.810]what is your goal and for example, we can use this guy,
[00:56:07.710]the PRI, that can be a good index for stress like water
[00:56:12.270]stress, for this instance we we showing here and then we do
[00:56:16.650]discuss some of that in this paper.
[00:56:19.080]But this relation,
[00:56:20.640]this index can is only available from ground or the
[00:56:24.783]airplane, not from the satellite data
[00:56:28.772]but if you want to measure stress
[00:56:29.605]for the crop fields,
[00:56:30.510]we have other methods to measure the predilation
[00:56:34.980]water content for the longer draw stress.
[00:56:37.662]That can be a good indicator.
[00:56:39.330]So I think the answer to this one is there are multiple ways
[00:56:42.540]but depending on how you want to use that and what's your
[00:56:45.000]context.
[00:56:47.640]Excellent, thank you.
[00:56:50.220]We have any more here in the room?
[00:56:52.059]Yeah, we've got time for, I think.
[00:56:54.746]So.
[00:56:57.410]Oh, okay. Gotcha.
[00:56:58.746]Thank you.
[00:56:59.850]I wanna follow up on the stress thing. Ran,
[00:57:01.680]you showed in one of your last slides from east campus a
[00:57:05.280]response of of PRI and fluorescence light curves.
[00:57:10.260]These are light response curves for two different species.
[00:57:14.100]So I would suggest that there's a clue to that or an answer
[00:57:19.710]to that question here as well. So can you,
[00:57:22.980]can you say something about how this information can tell us
[00:57:26.850]something about stress for example, like.
[00:57:31.920]Yeah, so here remember three different lines.
[00:57:36.090]They tell us the data were collected at three different
[00:57:39.000]days. So let's assume the August one here is
[00:57:43.080]like the plants, they are healthy plants,
[00:57:45.900]they are pretty happy.
[00:57:47.669]And then during this timeframe,
[00:57:49.260]from August of September to October,
[00:57:51.960]you see this relationship, the PRI from this blue line,
[00:57:56.759]they are changing for this decision species and for this
[00:58:01.140]algorithm species.
[00:58:03.540]The inter,
[00:58:04.440]the slope of this lines is change the less than this
[00:58:09.690]visual species.
[00:58:11.700]But this intercept changed. That tells us the,
[00:58:16.080]the pigmentation of those plants changing,
[00:58:19.050]especially for this guy,
[00:58:20.430]this green ash, that species have very early senescence.
[00:58:26.430]They see this collapse in early October and the SIF kind of
[00:58:31.958]tell us the same thing in August.
[00:58:33.955]They see very high SIF values but they see here in October
[00:58:40.230]the SIF is very low. That means low for low for censuses.
[00:58:45.570]From August to October this slope change and also the
[00:58:49.530]values change cause that something happened to those plants.
[00:58:53.970]Okay, thanks.
[00:58:58.260]Okay, well I think we will have to wrap it up,
[00:59:00.600]but please join me, and thank you Ran for a great
[00:59:03.300]presentation.
[00:59:04.440]Thank you
[00:59:06.065](applause)

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

Applications of Remote Sensing in Monitoring Ecosystem Function and Biodiversity

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments