Novel Technologies for Monitoring Field and Farm-scale Greenhouse Gas Emissions

EDUARDO SANTOS Author

04/06/2023 Added

17 Plays

Description

Associate Professor of Micrometeorology, Department of Agronomy, Kansas State University

Searchable Transcript

Search:

[00:00:00.780]The following presentation is part
[00:00:02.700]of the Agronomy and Horticulture Seminar Series
[00:00:05.820]at the University of Nebraska-Lincoln.
[00:00:08.400]Hey, good afternoon everyone.
[00:00:10.470]Thanks for being here again
[00:00:12.000]in the Agronomy and Horticulture spring seminars.
[00:00:15.660]We have the professor today to introduce Dr. Eduardo Santos.
[00:00:18.993]He's an associate professor of agricultural micrometeorology
[00:00:22.860]at Kansas State University.
[00:00:24.480]He grew up in Southeastern Brazil.
[00:00:26.760]He did his bachelor's at
[00:00:28.680]Federal University of Espírito Santo,
[00:00:30.810]and his master's at the University of São Paulo.
[00:00:34.020]Lately he moved to do his PhD
[00:00:36.180]at the University of Guelph in Canada,
[00:00:39.390]working with micrometeorological techniques
[00:00:42.390]to study the gas and energy exchange balance
[00:00:45.900]between plant canopies and the atmosphere.
[00:00:48.420]And he joined K-State in 2012.
[00:00:52.110]His research focus on the transport of mass and energy
[00:00:56.040]between the land surfaces and the atmosphere,
[00:00:59.160]applying micrometeorological approaches and other techniques
[00:01:02.610]to measure fluxes of trace gases.
[00:01:05.580]So with that, he's also involved
[00:01:07.362]in teaching graduate courses there.
[00:01:10.320]With that, I will give a welcome to the seminar.
[00:01:14.220]Yeah, for those online, if you have any questions,
[00:01:17.580]you can post it in the chat, and we'll have time
[00:01:20.370]at the end to go through that, okay?
[00:01:24.327]Okay. Floor's yours.
[00:01:25.638]Thank you for the invitation.
[00:01:27.183]It's actually my first time in Nebraska.
[00:01:29.550]I'm living so close, but it's the first time I've come here.
[00:01:33.397]So as O'Brien said, I'm associate professor
[00:01:37.050]of micrometeorology at the
[00:01:38.686]Department of Agronomy at K-State.
[00:01:40.950]Just wondering, how many of you knows
[00:01:42.363]what micrometeorology is?
[00:01:45.930]Don't know much?
[00:01:47.670]But basically what we can do,
[00:01:48.538]is that you divide the atmosphere layers,
[00:01:51.240]and then the boundary layer of the atmosphere
[00:01:54.030]is what's close to the ground,
[00:01:55.440]and it's affected by surface forcings and processes.
[00:02:02.310]So micrometeorology is the branch
[00:02:03.900]of meteorologists studying those processes.
[00:02:06.810]And it's basically meteorology for people
[00:02:09.510]that are afraid of heights,
[00:02:10.343]so they stay close to the ground.
[00:02:18.450]So today, where I'm planning to talk here
[00:02:22.350]about some background information
[00:02:26.040]on greenhouse gas emissions while related to agriculture.
[00:02:31.920]And then I'm gonna present here
[00:02:33.813]the results from, basically, two studies,
[00:02:36.120]one is a measurement of methane emissions from a feedlot,
[00:02:40.800]using a micrometeorological approach.
[00:02:43.590]And then the second study is on a new technique,
[00:02:48.690]a new laser that is being developed
[00:02:51.840]to measure trace gas emissions from agricultural sites.
[00:03:01.410]So here's just a general idea here,
[00:03:06.200]we have this diagram here showing global energy flows.
[00:03:13.110]So you basically can divide the radiation reaching the earth
[00:03:18.750]in longwave and shortwave radiation.
[00:03:20.550]So we have a portion here of this radiation
[00:03:27.750]that's shortwave, it's coming from the sun.
[00:03:30.750]And you see that 50% of this radiation
[00:03:32.790]is actually reaching the earth's surface.
[00:03:36.450]So Earth's surface is above the absolute zero.
[00:03:41.640]So it's also meeting some longwave radiation,
[00:03:46.307]and that's helping to dissipate some of the energy.
[00:03:49.740]So some of the energy is going up in the atmosphere,
[00:03:52.743]it's being absorbed by the gases in the atmosphere
[00:03:56.850]being emitted back.
[00:03:58.290]And I wanna call your attention to this region here,
[00:04:01.620]that's called an atmospheric window.
[00:04:03.780]So that's part of the range of that radiation,
[00:04:10.547]and it is not being absorbed by any gas,
[00:04:12.840]it's just going straight through.
[00:04:15.900]So now if you look here,
[00:04:20.820]I want to call your attention to this graph here.
[00:04:22.710]So we're just showing there how much there,
[00:04:25.170]radiation intensity from the sun and from the earth.
[00:04:28.530]And you see here that most of the energy
[00:04:32.700]here emitted from the sun or the peak (indistinct) here
[00:04:37.130]or emission is occurring here
[00:04:38.750]in this visible range of the spectrum.
[00:04:42.000]So, I mean, that's why probably plants and eyes
[00:04:46.740]are specialized in seeing visible light,
[00:04:49.530]and plants are absorbing visible light.
[00:04:52.380]So if you look here, at the earth,
[00:04:53.790]you are meeting in this radiation here
[00:04:57.720]is in the thermal infrared.
[00:05:01.440]So black-body radiation, Wien's Law,
[00:05:03.927]you can actually come up with these curves
[00:05:09.000]using some radiation loss.
[00:05:11.670]You've probably seen this at home if you cook
[00:05:13.530]in an electric stove and if you turn the stove on,
[00:05:17.250]like, why the elements start glowing
[00:05:19.890]is because you are increasing the temperature,
[00:05:21.720]and then you're shifting where that energy is
[00:05:25.080]immediate to the visible range.
[00:05:27.210]Before, it's just thermal infrared,
[00:05:28.740]and then it shifts to the visible range,
[00:05:30.900]but the reason I'm talking about this
[00:05:32.570]is that if you look here, most of the energy here,
[00:05:36.690]the earth is immediate here around this 8, 10 micrometers.
[00:05:43.140]And here's what's the previous graph has talked about
[00:05:48.000]in that atmospheric window.
[00:05:48.990]You see that there is two windows in this area here,
[00:05:53.460]this region here, that you basically don't have anything,
[00:05:58.560]basically don't have much absorption
[00:06:00.300]occurring from the atmosphere.
[00:06:02.460]But then if you look here, methane and then nitrous oxide,
[00:06:06.418]if you go up and down, you see that
[00:06:09.990]they're absorbed in that region there.
[00:06:12.840]So that's why they're part of the greenhouse gas,
[00:06:15.300]because if you are starting adding more gas in that window,
[00:06:18.780]then you're gonna close this atmospheric window where most
[00:06:21.600]of the energies is being emitted out of the atmosphere.
[00:06:27.270]So then there is different ways you can compare
[00:06:30.690]the greenhouse gas potential of these gases.
[00:06:34.110]So if you look here, carbon you use as a reference,
[00:06:38.550]and then methane use possibly has a greenhouse gas
[00:06:41.400]potential 25 times higher than the methane,
[00:06:45.270]depending on how many years you investigating,
[00:06:51.720]but if you look here, nitrous oxide also
[00:06:53.835]has a much higher greenhouse gas potential about
[00:06:58.020]300 times the global warming potential of carbon dioxide.
[00:07:06.330]Now this compounds here, fluoride gases,
[00:07:12.660]they have a much higher greenhouse gas potential.
[00:07:16.530]You see that here, some of them are 22,000 times the CO2,
[00:07:22.380]this is being made,
[00:07:24.060]it's manufactured for some industrial processes.
[00:07:27.240]And they did some heavy regulations on how much
[00:07:30.313]this can be produced and emitted,
[00:07:31.800]and some places even that some of them are no longer allowed
[00:07:39.025]to be used in some of these industrial processes.
[00:07:42.780]So here, just showing, like, where these greenhouse gases
[00:07:48.990]are coming from, and then by sector.
[00:07:51.480]So you see the energy is responsible for
[00:07:54.060]most of the greenhouse gas production. So about 74%.
[00:08:00.300]And then agriculture here, worldwide,
[00:08:04.440]and forest and land use, about 20%.
[00:08:09.210]It's a little hard to divide the sector,
[00:08:12.570]because think about agriculture,
[00:08:14.490]you also have to move things around,
[00:08:17.280]and you also, depending on energy and transportation sector,
[00:08:21.570]but basically you see that agriculture is not the largest,
[00:08:27.180]but it's still a substantial amount of greenhouse gas,
[00:08:30.720]it's possibly a substantial amount
[00:08:32.240]of greenhouse gas production.
[00:08:36.210]So if you look at main contributors of
[00:08:44.550]producing greenhouse gas in agriculture,
[00:08:46.920]we have enteric fermentation here, 40%.
[00:08:50.730]So enteric fermentation, if you don't know,
[00:08:52.500]is production of methane by ruminants, of the rumen,
[00:08:58.203]part of the byproduct of the digestion is methane.
[00:09:02.700]I have livestock manure, 20%, also methane,
[00:09:07.770]and then you have synthetic fertilizers and N2O emissions
[00:09:11.910]from nitrogen fertilization and others here.
[00:09:18.030]So in the US,
[00:09:22.260]you see enteric fermentation and manure management
[00:09:28.470]considered the largest anthropogenic source of methane.
[00:09:35.850]And then recently, as you see,
[00:09:38.100]there has been increase in the rate
[00:09:41.670]that methane has been added to the atmosphere.
[00:09:47.520]They suspect that it plateau here a little bit
[00:09:50.460]on the early 2000s, and then different theories about that.
[00:09:55.020]One of the theories were that there was more drought
[00:09:56.820]in tropical regions, so less methane produced in wetlands,
[00:10:01.590]but in recent years there has been an increase.
[00:10:06.930]So going back to ruminants, and then I'm gonna talk
[00:10:09.600]quite a bit about methane here in this talk.
[00:10:13.530]So as I mentioned, the methanes produced
[00:10:15.930]is a byproduct of that digestion.
[00:10:18.000]So you have anaerobic condition, the ruminate animals,
[00:10:22.020]and then you have some of that,
[00:10:24.990]the carbon is oxidized and then you get methane,
[00:10:31.050]and then water vapors and water as the byproduct.
[00:10:39.330]So, that's not desirable, like,
[00:10:41.820]you lose some efficiency with that methane production.
[00:10:47.820]There is some incentive in terms of efficiency
[00:10:52.320]to reduce the methane production from cattle.
[00:10:57.090]There is different ways it can be done.
[00:10:59.070]There's some changes in diet,
[00:11:02.640]type of forages affects how much methane is produced.
[00:11:06.090]There is some companies now producing additives
[00:11:08.820]that can help to minimize that production,
[00:11:14.016]and I think even immunizations, like, giving vaccine
[00:11:17.850]that would affect how much the microorganism,
[00:11:25.026]the rumen and reduce methane.
[00:11:28.740]And then, more recently, you develop more carbon markets.
[00:11:32.340]We get farmers asking about this.
[00:11:34.560]And then so most of the processors and markets
[00:11:42.120]in the US, they are planning to reduce
[00:11:46.920]the greenhouse gas emissions by 10 to 30%.
[00:11:51.690]They want to reduce the 10 to 30%
[00:11:55.140]baseline emissions by 2030.
[00:11:58.440]And then over 90% of the beef processed in the US
[00:12:02.307]is processed by a packer that has publicly stated
[00:12:05.940]an emission reduction goal.
[00:12:08.940]So however, like, the measurements here are challenging,
[00:12:13.290]we don't have a lot of ways of verifying those measurements.
[00:12:17.730]And then that's the motivation of some of this research
[00:12:21.120]that I'm gonna present here today.
[00:12:24.540]So this is the first study that I'm gonna show
[00:12:27.060]is that we used a micrometeorological approach,
[00:12:30.570]it's called eddy covariance, to measure methane emissions
[00:12:34.050]from a commercial feedlot in Kansas.
[00:12:39.390]So I don't know if you're familiar with this method,
[00:12:43.650]it's a large scale method, but before I go to that,
[00:12:47.490]I just wanna show like, how traditionally
[00:12:49.200]you do this measurements, some of the techniques here.
[00:12:51.407]So you have the first one there is tracer technique.
[00:12:55.620]So this is a picture from Brazil,
[00:12:58.980]like, they're doing some studies in Embrapa,
[00:13:00.570]that's kinda like the USDA there.
[00:13:02.910]And so basically, you view this PVC canisters,
[00:13:08.130]and then it has a negative pressure,
[00:13:10.050]and then you see there is a little tube here
[00:13:12.090]that goes close to the nose of the animal,
[00:13:15.060]and then the animal swallows a capsule with a gas
[00:13:20.040]or trace gas with a known emission rate.
[00:13:23.430]So based on the concentration of that gas
[00:13:27.300]and then methane in this canisters here,
[00:13:31.110]you can determine or estimate an emission rate.
[00:13:35.430]Yeah, so another way, you can just put the animal
[00:13:39.000]in a chamber and then use a mass balance approach
[00:13:43.200]and know, what's the concentration of methane
[00:13:46.240]and other gas are going in,
[00:13:47.455]and then the concentration going out.
[00:13:49.710]And then based on the flow rate,
[00:13:51.630]you can determine the flux as well.
[00:13:54.780]And then more recently, I think there's the companies
[00:13:58.290]building this GreenFeed system.
[00:14:00.900]So they provide some feed to cattle here,
[00:14:05.550]and then to attract them to stick their head inside.
[00:14:10.800]And then you sampling the gas that's going in
[00:14:13.560]and then going out, and then estimating your flux
[00:14:17.214]based on those concentrations that you're getting
[00:14:22.800]in and out of the GreenFeed system.
[00:14:25.710]So you see that most are here, these techniques,
[00:14:28.500]they're restricted to a single animal,
[00:14:33.285]and some of them are pretty time consuming,
[00:14:36.210]like, you have to retrieve these canisters
[00:14:38.820]and do some lab analysis after that.
[00:14:43.860]So what you thought of doing is that, using more like,
[00:14:47.640]develop a field-scale approach to do these measurements.
[00:14:52.530]So what we did here is that we use what's called,
[00:14:56.130]like, an eddy covariance system.
[00:15:04.567]So these are a network of flux towers around the world.
[00:15:08.933]And you see here that different places,
[00:15:12.480]and in some of these places they have quite towers
[00:15:15.120]that have been up for a very long time.
[00:15:17.130]I think some in Nebraska that are part of this network.
[00:15:23.070]And then they have towers like this,
[00:15:24.690]that basically measure wind and concentration,
[00:15:27.450]and then, through a method called eddy covariance,
[00:15:30.330]they derive fluxes from those measurements.
[00:15:33.780]So we thought, okay, so now the restriction back,
[00:15:40.770]or using the constraint to use this method before,
[00:15:44.190]is that you didn't have sensors to measure methane,
[00:15:46.440]but we recently developed some sensors
[00:15:48.840]that you can measure methane at the rate and accuracy
[00:15:53.820]required for the method.
[00:15:55.650]So then I thought we could use that in a feedlot
[00:15:58.050]to measure methane emissions.
[00:16:01.110]So here's just a little bit of the theoretical principle,
[00:16:05.130]how this method works.
[00:16:06.600]So you can think about the airflow through a field
[00:16:11.550]as a turbulent flow with, like, several eddies,
[00:16:16.710]different sizes. They're superimposing here,
[00:16:19.680]and then mixing things and transporting energy and matter.
[00:16:28.170]So if the field is large enough and then homogeneous enough,
[00:16:38.970]your flux will be a constraint.
[00:16:40.890]You'll be just a one dimensional problem,
[00:16:44.130]like, you're just moving up and down,
[00:16:46.080]so the main wind is not transporting molecules and energy.
[00:16:54.587]So what we did is that we set up a tower like this in
[00:16:58.050]the edge of a feedlot. So it's a relatively large feedlot.
[00:17:02.800]So we have 1200 meters here downwind from the tower.
[00:17:09.210]So the total air is about 59 hectares flat terrain
[00:17:14.610]and surrounded by agricultural fields.
[00:17:16.590]So it's important that you don't have
[00:17:18.030]other methane-producing areas around it.
[00:17:24.150]And then the feedlot had the total capacity
[00:17:26.790]of 60,000 head of cattle.
[00:17:30.150]So here's like, this is a picture of the tower.
[00:17:32.910]So the tower, you basically you have some wind sensor here
[00:17:38.040]that is measuring orthogonal components
[00:17:41.820]of wind at a high rate. So, admission about those eddies.
[00:17:45.450]So to capture the scales relevant for the flux,
[00:17:50.520]you have to sample these really quickly.
[00:17:52.080]So you're getting 10 samples per second here,
[00:17:54.510]and then sometimes more than that,
[00:17:55.800]depending on the height of the tower.
[00:17:58.050]And then if you see here, we also had an open-path analyzer,
[00:18:03.150]that's actually produced by Licor here,
[00:18:05.040]just used as a reference because it's an established sensor.
[00:18:08.700]And then you have here a methane sensor
[00:18:11.160]that was placed inside this enclosure.
[00:18:15.840]So then we follow that, like,
[00:18:17.970]getting this high frequency data to fluxes,
[00:18:21.960]half an hour fluxes, follows several steps,
[00:18:24.242]I'm not gonna go into details here,
[00:18:25.860]but then this is some of the results here
[00:18:30.390]comparing the CO2 obtained by the methane sensor
[00:18:34.170]also measuring CO2, and then the open-path
[00:18:37.110]with the Licor.
[00:18:38.640]So you see that you have a good agreement here
[00:18:41.160]between the sensors, and that gives us some confidence
[00:18:44.430]that the methane, that means you doing this processing
[00:18:47.640]the data and then you're getting good CO2 fluxes.
[00:18:51.390]I mean, the likelihood is that the methane flux
[00:18:52.950]are also gonna be good,
[00:18:54.390]'cause the steps for the calculation are the same.
[00:19:03.060]So here's just some flux data from methane now.
[00:19:07.050]So you see that the wind speed, it would explain
[00:19:10.920]some of the methane emissions.
[00:19:13.128]So if the tower is here, the wind is blowing from the north,
[00:19:17.340]don't get much flux, because you're not getting emissions
[00:19:21.240]from the... But then if the wind shift is to the south,
[00:19:26.340]the wind's blowing from here, now you're getting,
[00:19:28.410]most of the air is moving through that feedlot
[00:19:30.690]and then transporting that methane to your sensor.
[00:19:33.870]Okay.
[00:19:34.740]So, you're getting here flux in
[00:19:37.530]micromoles meters squared per second.
[00:19:40.650]So that's not very useful, okay,
[00:19:43.680]because if I wanna compare this feedlot
[00:19:45.660]with a different feedlot,
[00:19:47.310]what I really want is that a flux per, you know,
[00:19:51.137]pen or a flux per animal.
[00:19:53.430]And that's where you try to scale the flux now,
[00:19:56.313]to get that...
[00:20:00.090]So to do that, you had to do some footprint analysis.
[00:20:06.581]So footprint analysis there, a footprint
[00:20:08.370]is the area that's influenced the fluxes in your tower.
[00:20:12.690]So there is different models out there
[00:20:15.540]that can estimate that area.
[00:20:18.150]But the way to visualize that,
[00:20:20.760]you have at some point downwind of the tower,
[00:20:26.520]you're gonna get a high contribution from fluxes.
[00:20:29.430]And as you go away from the tower,
[00:20:32.820]you see that the contribution decreases,
[00:20:36.060]and that varies depending on the wind direction
[00:20:38.880]and the wind speed. So for instance,
[00:20:41.040]if I have a more troubling flow during the day,
[00:20:45.810]this area here of high influence
[00:20:48.120]is gonna be closer to the tower.
[00:20:49.950]At night when you have more stable conditions,
[00:20:52.980]and then you have light winds.
[00:20:54.990]Now this area is gonna stretch further downwind.
[00:21:02.100]So we did some simulations here, and then
[00:21:04.892]we basically looked at the distance here that you
[00:21:09.102]had the highest contribution to the flux,
[00:21:12.188]and then you plot the fluxes and look at that contribution.
[00:21:16.170]And you see here that that's what you expect, like,
[00:21:19.410]as you move away from the tower you get higher fluxes here.
[00:21:24.480]Now when I move this way here,
[00:21:26.400]that I get some contributions from airs outside the feedlot.
[00:21:29.940]I dilute that flux, and then I have lower values.
[00:21:36.300]So what you did then, is that we ran, like, a 2D model,
[00:21:42.600]and then you're looking at a two-dimensional footprint,
[00:21:46.980]flux footprint. So you can see here that
[00:21:52.189]this is a nighttime footprint, you see they can stretch
[00:21:55.170]a very long area in the feedlot.
[00:21:57.720]And then here you see that during the day,
[00:22:00.270]that sampling area is constrained closer to the tower.
[00:22:06.360]And then you also can see in this area here,
[00:22:08.400]that I'm sampling pens,
[00:22:10.500]but I'm also sampling alleys and roads
[00:22:12.480]that are not emitting methane.
[00:22:14.550]So the challenge there is like, whoa,
[00:22:16.440]how are we gonna scale the flux
[00:22:18.840]so I'm only getting fluxes from pens?
[00:22:22.380]And the way we did this, is that we used that
[00:22:26.430]fraction of errors from pens and alleys to scale the flux.
[00:22:33.030]So that's the corrected flux here in this light blue.
[00:22:40.080]And then that's just the raw flux without any scaling,
[00:22:44.340]like you're getting...
[00:22:46.710]And then for comparison,
[00:22:48.240]you did use some times that you have most of the fluxes
[00:22:51.330]come from a pen, like usually during the daytime,
[00:22:54.870]winds blowing from the south, that then have more certainty
[00:22:57.720]that that those fluxes are coming from the pen.
[00:23:00.150]And you see that the scaled flux
[00:23:01.680]followed the pen flux pretty closely.
[00:23:05.640]And here, like, we just use two different footprint models,
[00:23:09.150]and then this model here turned out to be more robust
[00:23:13.953]than the the second model. This is just a
[00:23:20.670]flux per animal now, and then these are monthly averages.
[00:23:26.760]So you see here that you have some outliers here,
[00:23:32.500]and you think that the outliers, like,
[00:23:34.350]one of the assumptions of the matter that the animals
[00:23:38.010]will be close to even distribution in pens,
[00:23:42.240]but sometimes they go to one corner of the pen
[00:23:45.300]and then you get some outliers that are caught by that.
[00:23:49.890]But overall, you see that there is a small reduction here
[00:23:55.920]in the fluxes here during the the winter
[00:23:58.890]and in spring months, 'cause one of the reasons
[00:24:01.440]is that they took some animals out of the lot,
[00:24:03.840]so they reduced the number of animals and the stockiness.
[00:24:07.740]So the average here ranged from 75 to 125 grams
[00:24:12.810]of methane per animal per day.
[00:24:17.340]Just showing here, like, the difference here,
[00:24:20.310]like, in the dry, we think that some of the flux
[00:24:22.650]could also be caused by the surface being wet,
[00:24:26.160]and then some methane produced by the soil.
[00:24:29.850]But going back here, so here's the comparison
[00:24:32.730]with the previous studies.
[00:24:34.920]I wanna point out this studies here
[00:24:37.320]that are more close to a system that you measure,
[00:24:39.810]more like feedlot in the Great Plains.
[00:24:41.970]So I think this one's done in Texas,
[00:24:45.930]this one was done in Texas.
[00:24:46.980]So you have a good agreement here,
[00:24:49.170]the range of values that you'd get with the method.
[00:24:58.080]Okay, so moving on here,
[00:25:00.840]just going to our second study here.
[00:25:06.720]So a few years ago, maybe like five years ago now,
[00:25:11.820]we got a grant from NSF with some faculty
[00:25:19.050]in physics at K-State, they're laser physicists.
[00:25:22.660]They work with fiber optics and fiber lasers.
[00:25:27.840]So they develop these lasers that you inject some light
[00:25:32.010]in the fiber, and then you get some light,
[00:25:34.710]different colors or different range of the spectrum.
[00:25:37.500]And then we got this grant to build a laser
[00:25:42.810]to measure trace gases, so, important for agriculture.
[00:25:49.890]So I'm gonna show here the results of some of these studies.
[00:25:54.150]So before I start here,
[00:25:55.783]I just wanna explain a little bit about the technique.
[00:26:00.960]So this technique is based what's called mode locked lasers.
[00:26:04.620]So mode locked lasers,
[00:26:08.010]you have a laser, like a pump laser,
[00:26:11.880]you inject some light in the cavity that's gonna
[00:26:16.200]be absorbed by a material,
[00:26:18.390]and then that material is gonna give off
[00:26:20.280]short pulses of light.
[00:26:22.770]So you're talking about trillions of pulses per second,
[00:26:26.160]but you're getting those pulses of light from that.
[00:26:30.840]So then this pulse of light, they are evenly spaced in time,
[00:26:37.299]and then if you go then now from, oh, sorry.
[00:26:40.890]And the cavity is constructing a wave,
[00:26:42.810]so you have a high reflective mirror here,
[00:26:43.643]and then you have a semi-transparent mirror
[00:26:47.627]on the other side.
[00:26:49.530]So some of this light is always escaping,
[00:26:51.900]but then you have a standard wave of light
[00:26:55.350]inside this cavity here.
[00:26:56.700]So you're basically locked in frequency inside that cavity.
[00:27:02.910]So now if you go from a time domain here,
[00:27:06.690]if you move from a time domain to a frequency domain
[00:27:10.920]using Fourier transform, you're gonna see
[00:27:13.170]that the frequency's gonna look like this.
[00:27:15.030]So you have different wavelengths here,
[00:27:17.580]and then those posts are gonna translate
[00:27:21.810]into different teeth here.
[00:27:23.340]So that's why the name "dual-comb",
[00:27:25.530]'cause it looks like a comb, you have a comb of light.
[00:27:28.650]Okay, so how is this can be used?
[00:27:32.310]So this technology has been used for different applications.
[00:27:37.830]One important one is very precise clocks,
[00:27:43.200]so better than atomic clocks.
[00:27:44.760]So some satellites, they have to have these on board,
[00:27:48.600]and then you use this frequency,
[00:27:50.880]these pulses, to actually keep track of time.
[00:27:56.670]So the way you use this is that,
[00:27:59.523]you can think about this comb here as a ruler, okay?
[00:28:04.230]So instead of measuring inches or centimeters,
[00:28:07.020]you're actually measuring frequency with that ruler.
[00:28:10.050]Okay? So let's think about, like,
[00:28:12.150]what to make a ruler useful.
[00:28:15.092]Okay, so what to make a ruler useful.
[00:28:19.890]Basically what you have, evenly spaced marks,
[00:28:26.220]see, in this ruler, and you also have to have a a scale.
[00:28:31.260]So the people that use this and then,
[00:28:34.080]so there is different things, so the laser's not perfect.
[00:28:37.200]So there is some variation here in this time,
[00:28:40.590]but then there is ways you can lock that laser
[00:28:43.320]that time stays pretty constantly.
[00:28:44.850]The way you do this, is that you take
[00:28:48.605]one of this tooths in here of the laser,
[00:28:50.580]and then you lock without the continuous wave laser.
[00:28:53.280]So you keep that frequency stable, and that tooth.
[00:28:58.200]And another way here, so this solves one of the problems,
[00:29:01.140]like they are evenly spaced.
[00:29:03.600]But now you are meeting the frequency
[00:29:06.140]on the original spectrum,
[00:29:07.440]but you don't know what's your frequency equal to zero.
[00:29:10.710]So then you have to do another procedure here,
[00:29:13.590]that you reference the comb against itself.
[00:29:17.940]And this was done in by John Hall and then Theodor Hänsch,
[00:29:22.856]and then they won the Nobel prize in 2005
[00:29:25.950]when they figured out a way of doing this self-reference.
[00:29:29.700]Okay.
[00:29:31.680]So how does it work, then, to do the measurement?
[00:29:35.580]So you basically use two combs
[00:29:37.050]and I'll go back to this and explain why two combs.
[00:29:42.180]Some of the light just shines through a gas
[00:29:44.683]or through the atmosphere,
[00:29:47.640]and some of these lights is absorbed in a very specific
[00:29:56.461]frequencies by the molecules.
[00:29:58.830]And then if you look at the spectrum,
[00:30:00.330]you're gonna see a dip there
[00:30:01.360]where the energy is missing because
[00:30:04.350]that part of the energy was absorbed.
[00:30:09.630]So then what you need to do now is that you may ask,
[00:30:14.040]like, why are you using two combs?
[00:30:16.050]The problem is that the absorption,
[00:30:18.630]that the wiggles in the molecules,
[00:30:20.940]it occurs in optical frequency, terahertz region,
[00:30:25.710]and then you cannot detect that.
[00:30:27.600]So the way you do is you mix the light of two combs,
[00:30:30.000]and you basically look at the difference,
[00:30:32.400]and that's what you're doing here.
[00:30:33.510]You're mixing the light of two combs,
[00:30:35.580]and now you get frequencies here,
[00:30:38.400]instead of terahertz you're getting kilohertz frequencies.
[00:30:41.640]So that's can be detected.
[00:30:43.260]So that's why it's called dual-comb.
[00:30:44.640]You need to heterodyne the light,
[00:30:46.470]or mix the light, to see the absorption.
[00:30:53.760]Okay, so we took that system
[00:30:57.270]to a small feedlot close to Manhattan.
[00:31:01.230]This is a research facility at K-State,
[00:31:04.560]and basically we had two laser beans.
[00:31:07.140]So one here in the north side of the feedlot,
[00:31:12.387]and the other one on the, sorry,
[00:31:14.635]one on the south side and the other one
[00:31:15.900]on the north side of the feedlot.
[00:31:18.032]And one of the nice things
[00:31:22.918]about the frequency that you work on,
[00:31:25.290]the wavelength that you work on,
[00:31:26.670]is that you can use regular telecommunication fiber.
[00:31:30.810]So, like, stuff that's probably in your backyard,
[00:31:34.620]that's the same fiber that you're using for this measure.
[00:31:38.280]So I was actually really surprised when I ordered the fiber
[00:31:43.080]because it's actually cheaper than regular plastic tubing
[00:31:45.960]that you buy for the lab (indistinct).
[00:31:47.760]You know, it's a little cheap. But the things is that
[00:31:50.517]the telecommunication industry made, like,
[00:31:54.000]several hundred kilometers of this fiber,
[00:31:56.730]there's a high supply of this fiber out there.
[00:32:00.060]So you can think back on this component
[00:32:02.520]is that it's relatively cheap now to do this measurement.
[00:32:06.900]So then the way it works that the light is generated here.
[00:32:12.510]So the comb is actually,
[00:32:14.340]the system is inside a trailer, temperature controlled.
[00:32:19.140]Then we take that light through fiber optics,
[00:32:22.710]and then it goes through a transceiver.
[00:32:24.900]Like a transceiver looks like a telescope,
[00:32:30.030]and then you're just sending that light
[00:32:31.800]in an open path there.
[00:32:33.960]Then it hits a corner reflector, basically a mirror,
[00:32:38.340]and then it comes back, that light is focusing again,
[00:32:41.940]and then goes into the fiber,
[00:32:44.910]and then we analyze inside the trailer.
[00:32:48.242]So the nice thing here is that with a single system,
[00:32:51.690]you can measure several paths. So one of the disadvantages
[00:32:56.340]of fusion micrometeorological methods is that
[00:32:58.220]cementation is expensive, and then you can only measure
[00:33:02.100]one treatment at a time usually.
[00:33:04.080]But with this, you can split different paths
[00:33:06.810]and then measure different areas in the field.
[00:33:12.390]So then to validate the method, to check the method,
[00:33:16.170]we had a commercial closed-path gas analyzer,
[00:33:21.810]the same one that we used for the feedlot study.
[00:33:24.780]And then the way we did this
[00:33:26.160]is we had some air intakes here,
[00:33:29.100]those yellow dots are air intakes,
[00:33:33.360]and then they would basically draw air
[00:33:37.174]to those intakes and go to the trailer,
[00:33:38.610]and then you switch off the manifold
[00:33:40.860]and then measure the south and the north path
[00:33:43.080]for the concentration.
[00:33:48.300]And then here, we had some wind sensor here,
[00:33:52.770]so any kind of moment just so you could have
[00:33:55.290]some wind data to do the flux estimation.
[00:34:00.210]So here is just data showing the spectrum.
[00:34:03.150]So the spectrum covered by the sensors is relatively broad.
[00:34:09.090]So you actually get absorption lines for different gases,
[00:34:14.700]including isotopologues of those gases.
[00:34:17.100]So you are able to see different water isotopologues.
[00:34:21.690]And then here, just showing here
[00:34:23.820]some of the absorption lines,
[00:34:25.500]but I just wanna point out that's broad.
[00:34:27.690]And then you are also able to measure ammonia
[00:34:30.930]with the sensor.
[00:34:32.730]So here is just showing a time series of concentration
[00:34:35.820]of methane on the north and the south path.
[00:34:39.750]So this day here, the wind is blowing from the south.
[00:34:43.830]So you see in the enhancement here on the north path,
[00:34:47.490]because you have methane being produced in the feedlot.
[00:34:50.940]And then this is, we just calculate Allan variations.
[00:34:56.100]So basically how did you average the concentration time,
[00:35:01.410]and then you look at the variation in time,
[00:35:03.990]and then you see that these for the closed-path sensor,
[00:35:09.480]you have an increase in time,
[00:35:10.620]and then there is a little decrease here with the laser.
[00:35:13.620]And then what I just wanna point out here,
[00:35:16.200]that there is a slight improvement in hour variation here
[00:35:20.990]in the variability, using these open-path measurements,
[00:35:25.500]because using the laser, you can sample that plume
[00:35:29.340]better than having intakes there during just
[00:35:32.130]bringing the air in.
[00:35:35.160]So then, like, we're not really interested in concentration.
[00:35:38.730]I mean, you need a concentration, but then, like,
[00:35:40.680]how would you go from a concentration to flux?
[00:35:43.320]We use what's called, like, inverse dispersion model.
[00:35:47.160]So basically what you are looking at,
[00:35:49.950]is that the enhancement concentration
[00:35:51.377]caused by a source area,
[00:35:53.730]and then you do inverse because in the forward model,
[00:35:57.804]you infer the concentration downwind
[00:36:00.510]based on the emission rate by a source area.
[00:36:03.600]Now we can go the other way around and you can infer
[00:36:08.458]the emission rate by a source area
[00:36:10.590]based on the concentration enhancement.
[00:36:13.140]And that's what we did in this case.
[00:36:16.350]So this is just some flux data. So this was calculated
[00:36:19.890]with the commercial sensor, the closed-path sensor,
[00:36:23.910]and this was calculated with the laser, the dual-comb laser.
[00:36:28.860]And then you see that these are a good agreement.
[00:36:30.870]So these blue dots, see, is when you have cattle
[00:36:33.810]in the site, and then we extend the measurement
[00:36:37.350]when they remove the cattle and see the drastic reduction
[00:36:40.830]in methane emissions.
[00:36:44.580]We also had ammonia fluxes.
[00:36:47.370]We didn't have a direct way of comparing ammonia
[00:36:51.600]with another sensor.
[00:36:53.670]Ammonia is very sticky, it's really hard
[00:36:55.440]to do measurements of ammonia with a closed-path system.
[00:36:58.920]So we just couldn't do it, but we just related
[00:37:03.480]that flux there with temperature,
[00:37:06.420]and you see there is some correlation
[00:37:08.640]with ammonia flux and temperature, and then wind speed.
[00:37:14.700]Okay, so now, this is an ongoing study,
[00:37:17.130]so what you wanna do now is let's just take this laser.
[00:37:20.790]I mean, measuring the emissions from a feedlot
[00:37:25.830]is, like, relatively easy. You have animals concentrated,
[00:37:28.350]higher concentration enhancements.
[00:37:30.480]So what you wanna do now is, let's see,
[00:37:35.091]can we do this measurement in a pasture?
[00:37:39.360]And then that's what we're doing,
[00:37:40.320]this is a research facility near Manhattan.
[00:37:44.790]So now you're measuring a much longer path.
[00:37:47.640]So in the feedlot, you're going, like,
[00:37:49.740]a one-way path, 50 meters.
[00:37:52.950]Now you're doing 400-meter paths.
[00:37:56.010]That helps, because if you have more absorption,
[00:38:00.120]your signal-to-noise ratio improves.
[00:38:03.900]So what we're doing here now at this point,
[00:38:06.450]is that we don't have any cattle in the pasture right now.
[00:38:09.540]So what you do is then a controlled release study.
[00:38:12.540]So you have here the tank of methane,
[00:38:17.250]and then you have this tubing,
[00:38:20.250]and then you have a manIfold here that controls the flow,
[00:38:23.760]and then that tubing and then you release methane
[00:38:25.640]at a null rate, and then you measure the two paths
[00:38:33.570]downwind and upwind from that manifold.
[00:38:36.300]And what are you trying to do,
[00:38:37.320]is that drew the reverse model,
[00:38:38.263]and see if you can see the same rates.
[00:38:42.180]So you see here that the telescope is here,
[00:38:44.640]and it's hard to see but there is a mirror over here,
[00:38:48.630]like 400 meters, that is refocusing the light
[00:38:52.530]and then getting it back.
[00:38:57.480]This is just preliminary data on the the study.
[00:39:01.679]So you have the manifold here,
[00:39:05.457]and this day I think the wind is blowing from the south,
[00:39:08.280]and then you get some gas going through the north path.
[00:39:13.290]I think this one's located over 50 meters from the manifold.
[00:39:19.620]And you simulating here a herd of, I think,
[00:39:23.490]15 head of cattle. So we're trying like,
[00:39:26.730]really small herds, and see if you can see it.
[00:39:30.150]And then you see,
[00:39:31.410]this is the methane concentration, both paths.
[00:39:36.390]And then you can see here, this is the difference
[00:39:38.220]which you would call the enhancement in concentration.
[00:39:40.890]So here's when I turn the gas on,
[00:39:43.380]and then you see that there is a slight increase here,
[00:39:47.130]and then you gas off,
[00:39:48.150]and then you decrease the concentration.
[00:39:52.560]The challenge here that you're seeing a lot of noise,
[00:39:54.837]and the challenge here, main source of noise
[00:39:57.990]is actually keeping the alignment.
[00:40:02.220]So when you go down 500 meters,
[00:40:05.760]any tiny vibration in the the the mirror or the telescope
[00:40:11.160]will cause this misalignment.
[00:40:13.350]And then the power you get back is slow,
[00:40:16.110]and then you increase the noise in that measurement.
[00:40:20.400]But I mean, we start doing some flux calculation.
[00:40:23.610]So, I didn't bring it here, but the flux is getting close.
[00:40:27.930]So if you average over a long time, even with that noise,
[00:40:30.723]it's getting close to what you're releasing
[00:40:32.820]with the manifold.
[00:40:34.410]So the idea is that you can use this
[00:40:38.670]maybe to do some grazing measurement,
[00:40:41.010]but then you're gonna have to track where the cattle are,
[00:40:44.610]I dunno, like, you do with cameras or maybe GPS,
[00:40:47.610]and that's gonna be tough as well.
[00:40:49.590]But at least I think you can detect
[00:40:52.560]the enhancement concentration.
[00:40:56.430]So right now, you have a lot of focus on methane,
[00:41:00.180]but you can see that you can measure
[00:41:02.280]other trace gases with this system.
[00:41:05.640]And then the idea is that you have a system like this,
[00:41:09.180]and then you can have reflectors in the field
[00:41:11.430]that will show a lot of horizontal gradient measurements,
[00:41:15.510]but you can also do a vertical gradient measurement
[00:41:17.670]and estimate fluxes.
[00:41:20.460]So people from the east
[00:41:22.587]are looking at a different wavelength,
[00:41:27.810]a different range of the spectrum,
[00:41:29.910]and they can detect nitrous oxide, VOCs,
[00:41:35.730]and other compounds, other gases, and then the idea is that
[00:41:40.920]you can use this and some very (indistinct)
[00:41:43.350]to estimate emissions from agricultural fields.
[00:41:51.270]So yeah. So I just wanna finish here
[00:41:53.400]by thanking collaborators from NIST and K-State
[00:41:56.940]that are involved in this project,
[00:41:58.620]and NSF and K-State for the funding.
[00:42:03.990]Thank you.
[00:42:05.730]My name's John Hay from
[00:42:06.870]Department of Biological Systems Engineering.
[00:42:09.060]We don't have much natural gas production
[00:42:11.910]in this part of the US, but other parts do,
[00:42:16.380]particularly south and and east and north of us.
[00:42:19.080]But is this a technology that might be used
[00:42:22.710]to detect leaks in that system?
[00:42:25.170]'Cause I feel like leaks in that system
[00:42:27.180]are a big thing we need to know more about.
[00:42:32.580]How are they doing that now,
[00:42:33.540]do you have any idea how they're testing those now?
[00:42:35.010]There is a company in Colorado that's called LongPath.
[00:42:39.690]They use this technology, and I think their business model
[00:42:44.130]is that they make this stuff, and some of the components
[00:42:48.030]you bought from them. So they make the laser,
[00:42:52.170]and then they come and then they identify where the leak is,
[00:42:55.860]and then they sell the data to these people from Hawaii.
[00:42:58.950]So they don't sell the instrument, but they sell the...
[00:43:04.710]Because, I mean, it's still, like,
[00:43:09.540]feeding the data, and it still requires a laser physicist
[00:43:13.680]to do some of this (indistinct). And then, yeah, so.
[00:43:17.670]But they use this technology, what they do
[00:43:21.332]is that they have a telescope and transceiver
[00:43:25.860]in the middle of a field,
[00:43:27.930]and then they have retroreflectors around,
[00:43:30.990]and then they scan it, and then, I mean,
[00:43:33.540]you're gonna see some plumes, right?
[00:43:34.950]And then they can tells you which sector that it was in.
[00:43:44.160]Just to comment, another way
[00:43:45.330]that they search these gas lines or check 'em for leaks.
[00:43:49.800]They fly in with an airplane.
[00:43:51.120]Yeah.
[00:43:51.953]And just have somebody, a spectator,
[00:43:55.500]an obvious person looking for dead plants.
[00:43:59.760]Oh.
[00:44:07.050]Would you have any way to describe
[00:44:09.570]or guide cattle feeders if, for example,
[00:44:15.900]with wet conditions, do you have less or more methane
[00:44:19.950]in dry conditions? Less or more, in other words,
[00:44:23.010]trying to come up with a way
[00:44:24.900]to enhance their management of their location.
[00:44:33.780]So I guess if you have-
[00:44:34.613]So to minimize that.
[00:44:38.190]Yeah, so if you have a lot of manure located in one,
[00:44:43.020]I guess the way you can see is that if you have
[00:44:44.460]a high concentration of manure under wet conditions,
[00:44:47.552]you're gonna have more methane production, right?
[00:44:53.160]Now, ground emissions for our air there,
[00:44:56.850]which is like, west, it gets relatively dry.
[00:45:00.150]They're not very large.
[00:45:01.650]So I think most of the emissions
[00:45:03.150]are actually coming from the animals, more emissions.
[00:45:06.330]But, I mean, one way is that,
[00:45:08.940]I don't know if that's feasible,
[00:45:10.050]but remove some of the manure,
[00:45:12.480]and then get the manure off of that.
[00:45:20.730]Have any questions online?
[00:45:29.850]What is the global warming potential
[00:45:31.740]of water vapor? Kind of...
[00:45:37.200]So yeah, so water vapors is actually very absorbed.
[00:45:43.500]Like, if you look at this spectrum,
[00:45:45.180]if you go back to that graph, it showed the particular.
[00:46:03.780]I see here water vapor you have absorbed
[00:46:07.650]in several regions there, and so it's pretty-
[00:46:09.983]and you can see that, right?
[00:46:12.073]I mean, during the night, like, cloudy night,
[00:46:16.350]you don't see the temperature dropping as much. So you...
[00:46:20.040]But the question is this, like,
[00:46:20.873]what you gonna do with water vapor?
[00:46:23.020]Water vapor's already there,
[00:46:24.857]and there isn't much way that you can...
[00:46:26.490]There's water everywhere, right? So yeah, but it's large.
[00:46:30.540]I don't think you include that in the count,
[00:46:32.280]because you can't do much about it.
[00:46:36.690]But I mean, if you think about clouds
[00:46:42.090]and the role of clouds,
[00:46:43.440]that's this big unknown in this climate model,
[00:46:46.980]is how much energy is being absorbed by clouds,
[00:46:50.190]and also reflected back to the atmosphere.
[00:46:56.130]I have a question. I mean,
[00:46:57.330]maybe your vision of the future...
[00:46:59.490]How do you see this type of sensors, of technology,
[00:47:02.760]20 years from now? Which type of applications,
[00:47:05.730]like, do you think that...
[00:47:09.510]I think you need to make them simpler.
[00:47:11.790]So people from different fields can,
[00:47:18.330]and it has become easier, like a guy like me
[00:47:21.195]can do a measurement like this.
[00:47:22.800]I can take it to the field and do a measurement,
[00:47:25.080]but I don't know if you're gonna get to the level
[00:47:28.080]that you're gonna get these on the farm,
[00:47:31.710]and maybe a service provider that could go there
[00:47:34.607]and do a measurement and then certify, okay,
[00:47:37.560]maybe this is an X, and then you can get
[00:47:39.450]some credits for that or something like that.
[00:47:44.510]Interesting.
[00:47:49.470]You showed the example
[00:47:50.940]with the release of the methane in the pasture system.
[00:47:55.050]Do you have any plans for experiments with animals
[00:47:59.580]in the pasture system?
[00:48:00.903]Yes. So the plan is to do a animal measurement this May,
[00:48:06.060]when the animals come back, you know,
[00:48:08.580]so that's still fixing some emissions, but (indistinct).
[00:48:20.941]So this system is not commercially available,
[00:48:24.480]but if we add the cost of all the equipment,
[00:48:28.980]it's probably around like 250, 300K.
[00:48:34.860]And then the reason why it's expensive
[00:48:37.101]is that it's not commercially available,
[00:48:39.360]if you have more units being built
[00:48:41.640]to probably lower the cost.
[00:48:44.940]Just have figure out a way to put a dual-comb on a TV set,
[00:48:48.930]so then everybody would have one at home
[00:48:50.880]and that would lower the cost.
[00:48:58.080]Got any other questions?
[00:49:03.660]Well. With that, I would like to thank you again.
[00:49:07.109]Thank you Santos. Thank you.
[00:49:09.149](audience claps)

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

Novel Technologies for Monitoring Field and Farm-scale Greenhouse Gas Emissions

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments