Dr. Helen Worden - 03/05/19

[00:00:00]thank you all for coming I'd like to welcome you to the Fat Tuesday edition of the eol seminar series I'm really pleased to introduce Helen worden as our speaker today Helen holds a bachelor's degree in physics from CU and a PhD in elementary particle physics from Cornell University following graduate school she took a sharp left turn into earth remote sensing science section at JPL and she was there for a number of years presently she's a scientist in a calm and she is the US principal investigator for the moppet instrument on the Terra satellite her primary research interests are satellite remote sensing of air pollution and climate change today she's going to tell us about two decades of global carbon monoxide observations with moppet Helen thanks Mike so a quick history of moppet moppets a Canadian built instrument and it was also funded by Canada and it's on NASA's Terra satellite and it would not exist without Jim Drummond so I want to acknowledge Jim Drummond who's the Canadian PI and then on the US side the the first u.s. P I was John Gilly followed by Merritt dieter and now myself so I'm privileged to follow that history and I want to acknowledge the entire and Karmapa team for the results in this talk so a quick outline mop it I'll talk about the instrument and how it measures carbon monoxide I'll talk about global pollution transport that we can see with Moffitt the emission trends of carbon monoxide and then what we can say about Co emitted species when we have these emissions such as aerosols covering dioxide and methane and finally I'll hope to get to the observations of CO from space in the future so so why Co I always show something like this because people might think it doesn't matter that much unless you need a new carbon monoxide detector in your house or you're stuck in a tunnel with really old cars you probably don't care that much about what Co is like if you walk outside but it actually tells us a lot about atmospheric chemistry and climate the main sources are incomplete combustion so from both fires and fossil fuel and then biogenic emissions and hydro carbon oxidation are the other big half of that the primary sync is oxidation by O H so that means that more CO would mean a longer methane lifetime because there's not as much o H to destroy methane it's a precursor to carbon dioxide and tropospheric ozone so it makes those greenhouse gases and because of that has an indirect radiative forcing of 0.2 2 watts per meter squared so it's not as high as say ozone but it's it's on the same order of magnitude the lifetime is weeks to months so that means it's transported globally but it's never evenly mixed so you can always see it above a background level so you can measure plumes from fires and cities and then like I was saying the the direct emissions of cor about half of what is in the atmosphere and those are from anthropogenic emissions which are relatively stable year-to-year but a large part is from biomass burning which has a lot of inter annual variability that I'll talk about and then there's a couple other about 20 tera grams per year from the ocean and then this is the direct emission from plants but plants also emit quite a bit of VOCs that turn into CO and I'll talk about that as well and then another big chunk of it is from methane so some quick facts we are getting to 20 years here it'll be the Terra at 20 we'll have the celebration at age you and in a special issue and some other things will be going on but it's pretty amazing because the design life with six years for all the instruments there's five instruments on Terra Astor series modus miser and moppet and as I said before the moppet instrument itself is a us-canadian collaboration so University of Toronto is responsible for the instrument building the instrument designing it and the instrument operations that they command through the Goddard folks here at NCAR were responsible for algorithm development and data processing so I'll talk more about this type of instrument that we have it's a it's a unique space instrument at this point halo is sort of a precursor but if that was looking in the limb and so it's fairly different so I'll get into that Terra is in a low Earth orbit so at 705 kilometers with a 10:30 equator crossing time so it's I don't know if anyone knows the history of AOS it used to be and before it was Terra aqua aura it was am/pm chem and so Terra was in the morning orbit hora was in the PM orbit and then so was Kevin which is I'm sorry aqua was in the PM orbit but at any rate we're in the morning orbit and we're in a constellation with a Landsat I think so we get 16 orbit today and that gives us about three days for global core cut takes about three days to get global coverage with moppet and I'll show you the swath in just a moment our pixel size is 22 by 22 so not very fine resolution on the ground but not bad either and then we measure vertical profiles of carbon monoxide in both the thermal infrared and the shortwave infrared and then we have a really high a really fine spectral resolution but it's an effective special resolution that you can see when I talk about the the GFC our measurements and right now we only retrieve cloud free sign cloud free scenes and you can see that in a moment and then finally but I just like to put this up Tara costs 1.3 billion dollars which sounds like a lot but considering it's almost 20 years old is probably a pretty good investment especially since a lot of the data from Tara are operational and a lot of systems and oh and then for perspective moppet cost about 75 million just the instrument and that again that was from Canada so so this shows moppet near real-time data in world view that's a world view so you can see it well I'd like to show this one because the near real-time products are operational for NASA Gao and also ECMWF but I'd like to show this because it shows the Nath the moppets swath and you can see we don't have anything where we have a big cloud cloud feature and and you can also see where Co is really correlated with with aerosols from motifs so in the orange yellow to orange colors we have motifs aerosols we have motifs clouds and then you can see the moppet data going through that big plume of aerosols and you can see where we had also high carbon monoxide and you can see some other correlated spots in the fires in South America so but I really like to show this because this this day seventh of September was just like Armageddon you had this big plume of smoke coming from Canada and then you had three hurricanes Harvey Irma and Jose I'm not sure what's the order up there so everything was just getting pounded that day so how Moffitt makes measurements we use again thermal infrared and shortwave infrared and you have the band the co band at 4.6 is the fundamental and then you have a weaker over term overtone ban at 2.3 microns and what gives these two different bands really unique structure or sensitivity in the atmosphere is the fact that one is solar reflection so the 2.3 band is basically measuring reflected sunlight which gives it sensitivity all through the entire column but it doesn't it doesn't give you a lot of structure vertical structure in that but you do see all the way to the surface whereas in the thermal infrared you're relying on thermal contrast which for most of the earth does not have a lot of difference between the you know the bottom layer of the atmosphere and the earth itself and so your peak looks like this because you have this is where you have a maximum of both thermal contrast in abundance of CEO so using that information together is what I will talk about in a second so moppet again is the gas filter correlation radiometer and so it measures in the nadir in like other infrared instruments it has to do a calibration using a blackbody and then a cold space view to get the cold temperature side and then it measures this radiance from the earth which has these spectral features but instead of going through say a grating spectrometer or an interferometer it goes through a gas cell and what the gas cell does by modulating either a long path or a short path with either the length of the cell or the pressure in the cell you can get two different kinds of transmission spectra one for low optical depth and one for high optical depth and by taking a difference in an average of those two types of transmissions you get a signal that looks like this an average signal and a different signal the average signal gives you information about the surface and the general radiance of the atmosphere whereas the different signal has information only in the spectral lines so that's where you're getting your information about the the gas just that gas in the atmosphere and then this goes to a detector arrays that are separate for the thermal infrared in the shortwave infrared so now on to the data products this is fairly common terminology for all of the NASA EOS instruments level 0 through level 3 where level 0 is the raw data level one is calibrated radiances level 2 is the retrieved product in this case we use optimal estimation to get vertical profiles and total column and I'm not going to go through this in detail except to to show what we actually produces products all of this is done by minimizing a cost function where you have a priori information this is the opry profile this is the operator covariance and then why are your radiances and this is a radiative transfer model with your measurement covariance and then we output the averaging kernel and error covariance matrices as a diagnostic then level 3 our gridded daily and monthly averages and validation is done using in situ data where we really rely on NOAA flights and then more recently hippo and a Tom qcls Co observations to give us latitude information and there's a series of papers by merit dieter at all our latest version is v8 so to illustrate the level 2 products I want to show a comparison to a Tom and so this is taking all of the Moffett profiles within 500 kilometers 6 hours of the an a Tom Pacific profile this was on the first campaign and this is the lat/long of that so what you can see is all these pink fine lines are the separate profiles this is the opry ori that was used for these retrievals and it does vary by latitude but for this scene it was about the same this black line is the actual a Tom observations and then the Green Line is what you get when you apply that the moppet averaging kernel to that and then also the prior so you can see that the pink lines are in this is the standard deviation in those is really matching pretty well to the smooth a tom data and this is what an averaging kernel looks like so I'm going to show some more of these so it's good to understand what this is so this is the pressure and then these are the averaging kernel rows with color coded by the pressure so since this is an ocean scene a pretty clean ocean scene you can see that we don't have a lot of sensitivity near the surface and that even the everything in from like 200 to 500 is sort of peaking around this 300 hectopascal level so that's really where most of our sensitivity is but we still have some in the lower taupe as well so now looking at how that compares to a Tom this is the the Moffett a Tom difference and you can see that compared to the zero line with what we think are the error bars for all of our profiles that we're doing pretty well okay so now the benefit of making multispectral measurements so this is just our product nomenclature in this case it would be seven and we have these seventy products are for the thermal infrared b7n we used to sort of mix up near infrared and shortwave infrared so we've used n to begin with and then v7j means joint so that's using both of these together and you can see from the averaging kernel that if you just use the thermal infrared that you you still have thump this is all all over land this is like the average you still have some sensitivity at the surface but not a lot compared to the middle troposphere and then like I was saying before for the shortwave infrared or the near infrared that information really looks pretty straight up and down like a column and then when you combine these two now you get something where you're really taking the fact that you are sensitive at the surface here and you know a lot about the middle troposphere from here and combining those two gives you a lot more information about the surface and then you can see this in an example of what level three data is so you have the thermal infrared this is level three data at the surface the near surface level so you have the thermal infrared you get a little bit more information at that level from the near curette and then finally you can really see a lot of the sources that you expect in that combined product so going on to just our our total record from space and I like this because you can really see pretty much everything in just one shot so this is the total column Co so this is just taking everything in the atmosphere adding it up and producing the molecules per centimeter squared and it's the zonal average by latitude and so I've just averaged all the longitudes so this is just a function of latitude and time and then the colors indicate higher values so what is really interesting in this part is you can see what you expect in terms of the southern hemisphere just has a lot less CO than the Northern Hemisphere because of those anthropogenic sources are just a lot higher in the northern hemisphere and then when you take the subtract off the months average month record averages you can see the percent anomalies so that's just taking out the seasonality of this and looking just at the inner annual variability now you can really see the big events that we we know about these are boreal fires from 2002 to 2003 this is the 2007 or 2006 El Nino which went into some a little bit in 2007 in terms of burning and those were some fires in Russia and finally this is the really big event we had from 2015 the 2015 El Nino and you can see this really persisted a long time in terms of high elevated Co but the main thing you can see in this is this overall decline over the the moppet record so that's what I'm going to really talk about when I talk about the emission trends but first I want to say a bit about what you can detect with carbon monoxide for a global pollution transport so this is just one month of an average month from 2010 October 2010 and it's nice that you can see what happens with fires so when you have the the biomass burning events in September October November or well October is a big month when you have the burning in this part of Africa the transport tends to go out here and then go back around there and then eventually it all just goes south and towards Australia so what I want to talk about is what we really saw for transport from the fires in Indonesia in 2015 so as you all know 2015 was a really big El Nino year the biggest one since 1997 and Indonesia was very very dry so this is the just the average monthly rainfall for September 2015 and it's just like nothing for the whole month of September in this part of Indonesia and this is the corresponding amount of Co that we observed at near the surface for for that month and this was at record levels for moppet so the highest we've ever seen and because we can get a pretty good vertical profile we can look at what happens to all of that Co from the Indonesian fires so this is these are all cross-sections looking at this longitude range and this is in August September and November and so for for August before the fires really got started they were a little bit going on still or starting but what you really see is this large amount of Co this was actually some fires in Russia and it follows in this Asian monsoon pattern that is well documented and then when you get to the Indonesian fires that really got going in September and October you can see this interesting rabbit-ear structure but basically it's right at the ITCZ so it gets pumped way up high right away and then goes north and south and this actually would look more like an atomic bomb pattern except for that moppet doesn't measure with where there are clouds and so right in the ITCZ we have less data so what another thing that we've looked at and then including that 2015 only no data is the fact that we can actually do a pretty good job of predicting inter-annual variability of CEO with a statistical model that uses climate modes so in this case we're using ENSO Indian Ocean dipole the Antarctic oscillation and then the tropical southern Atlantic climate indices and what this shows is the model fits not only overall coefficients but also a lag time so in the case of maritime Southeast Asia you you you don't have any leg time it's basically when ENSO is there everything gets dry and if you burn anything it's going to have a lot of CO emitted and then these other other indices had less influence and longer leg times and then in this case it was explaining 75% of the inter-annual variability and then in northern Australasia and South Australia we did pretty well as with the explaining the variance but the leg times are pretty different as well and then this just shows the anomaly and Co and then the black dots are the model and the red or blue dots are the measured anomalies so you can see it's a pretty good method for predicting when you're going to have high CO and that could be important for for example for air quality in Indonesia so now I'm going to get into Co emissions when we do a top-down estimate of Co emissions it's very similar to just estimating the retrieval or the the Co profile except now we have a cost function where X is the emission state vector with some prior yr the co retrievals with the error covariance and then f is a chemical transport model but it some also includes the averaging kernel because it has to match to y so it's basically the model but what the spacecraft would measure from that model and then we look at three sources of co in the in the the next thing that I'm going to be talking about so biomass burning fossil fuels and then biogenic where this biogenic term is not just the direct emissions of co4 plants but everything that turns into co from safe formaldehyde and well isoprene which goes to formaldehyde and then to co so you basically each grid box in the model in which in the case I'm going to talk about next was four by five degrees so pretty big each grid box has its separate prior for each of these three things so in this case we're looking at results from assimilating moppet in the geo scam model and these are the anthropogenic emissions and you can see the the trend from 2002 to 2016 so it was 15 years this is anjana at all and the three different lines here are experiments that he did to look at the result of when you try to assimilate either the column which is black the profile which is blue or just the lower profile and the lower profile alone is is was problematic so in general the column is probably the most robust result but you can see that they they do converge to about the same thing so in general you have a downward trend in the United States Europe and then you have a downward trend in China starting about 2007 whereas it's hard to say what the trend might be in the rest of these areas and then globally it's it's a downward trend so this was different than what for China than what is predicted by most bottom-up emission inventories basically everything is extrapolated and you know China was just going up and up and up so which is real and the only one that's really showing a negative trend is an inventory called Mayock from ting so I'm not gonna say it right things how the University in Beijing so a more recent paper looked at this in more detail and this is also using moppet data and it's using a French model that I'm not going to describe and also Bayesian inversion so here you have the Opera re emissions that went into this study and then the emission trends which you can see in blue are mostly declining over China and then this really looks at the whole bottom up in this inventory from a ik which really includes all the technology changes and the use changes by sector so blue is residential this is construction of materials yellows gasoline vehicles iron and steel industry industrial boilers and then power plants and I want to point out that power plants do not emit a lot of co2 have a power plant that emits a lot of CO you've built a bad power plant because it's not it's you really want complete combustion in your power plant so you can see that for example in residential sources this went down a lot and part of that is because China has been converting to electricity from say burning uh you know braziers in coal or or wood in the in the residential areas so converting to electricity is going to mean a lot less CO but a lot more potentially other things from power plants like no.2 but I'm not going to get into that so so again this is compared now to some other bottom-up inventories and in this case this is the result for all of East Asia the red line and it it's not really agreeing to much with the other inventories whereas if you look at just in China there's a couple of them that I think are trying to account for these technology changes and then this blue one is the mayor one that it agrees pretty well with so that was anthropogenic emissions now I'm gonna go to fire emissions and fire missions to look at that I want to point out a paper from 2017 by Adela at all that just looking at burned area you can see a real decline in global fires since from 1998 to 2015 so it's been a it's about a 25 almost 25 percent decline over that period the top part here shows just where you have these big fires so in Africa and South America and Australia some in Indonesia those are more sporadic though and then this shows the trend so you can see that for the most part it's going down everywhere except for this part of Africa and I'm the top down emissions also show a lot of also show this trend in Co emissions from fires going down but I thought this plot really shows what you could see just looking at Co trends so this is just the moppet trends in total Co column and I've subtracted the global mean trend of 0.43% per year and that lets you see one that you know a lot of the globe is just declining at that rate but this area in Africa is actually going up and you can see where those emissions come out here and you see that increase or relative increase there whereas from South America you see that whole transported Co going down so that really matches well with what this paper is shine for burned area with the emissions going up here and down here and then also down in here oops okay finally on to biogenic emissions so biogenic emissions play a very important role as a lot of people know one of the main emissions is isoprene but there's there's other ones that we need to worry about as well and these oxidize and then turn into secondary organic aerosol and then on the the gas phase they turn into ozone formaldehyde which is also measured from space and used to derive isoprene co methane and then I should point out that formaldehyde also produces co so that's included in our total budget for Co from biogenic emissions so these have been controversial lately and I like pointing this out because there's they've just been a few questions about what do we do for for climate change mitigation do we plant trees which might emit more biogenic emissions or do we cut down the forest it's a good an important thing to answer so unfortunately for Nadine she did not get to say what the title of this op-ed was and so to save the planet don't plant trees actually got her death threats so and that really wasn't the point of her op-ed or her paper as as I'm gonna show it was just mostly that you really have to include biogenic emissions in your climate models but there are more recently there have been some other things as well that are basically showing that you have to be careful what what trees you plant so this is the the climate effects of crop land expansion so this is the paper by Nadine Unger it's basically the effects from pre-industrial to present of changes in human land use and what they've done to biogenic SOA ozone and methane as compared to these other effects like if you cut down the forest then you have less co2 uptake and so you have a delta and co2 you have a change in surface albedo which might be a cooling effect and then you have the net effect of the Climate Pollutants that she sees as cooling so she got point 1 1 watts per meter squared cooling and that's by saying that you have this much SOA coming from it and for this much less SOA and this much less ozone and methane so she found a net cooling effect but in this paper by Scott at all from looking at what the projected change is in radiative forcing due to total deforestation so that might be an extreme case but they wanted to do something definitive I guess and so in their model they also included the short-lived climate forcers from aerosol cooling and also ozone and methane greenhouse warming and I'm not going to go into the separate terms here but for the global impact they did find a positive radiative forcing from cutting down all the trees and this is because in their model they found that the biogenic emissions were positive or warming because the decrease in aerosol cooling was not completely offset by decreases in ozone and methane which we're warming so the emissions of biogenic well the biogenic emissions for these things are very critical for understanding what path we should take going forward so what can we say about it from moppet and this is going to be sort of a a bit hand wavy but but basically we're going to apply this thing called them Markov chain Monte Carlo flux partitioner which I've depicted here as a meat grinder that was sort of my initial impression of it but when we look at the total Co flux that was constrained by Moffitt in this paper this paper was not able to partition the different biomass burning biogenic and fossil fuel terms so it just scaled basically or partitioned it according to the prior and because of that you can't really estimate an uncertainty so using this flux partitioner we basically did another Bayesian estimate and with that using more updated uncertainties in each of these inventories so the G fed biomass burning inventory Megan 2.0 biogenic and fossil fuels from Edgar we now can come up with a pretty good estimate of each term with uncertainties and you can see that the biogenic term is actually bigger for this in terms of tera grams per year for 2005 to 2012 then either of the anthropogenic or biomass burning so I need to do a quick explanation of what Megan does the Megan model for biogenic emissions this is from a paper by Eloise Marais who looked at Megan emissions of isoprene compared to Oh me top-down estimates using the ohmy formaldehyde observations and she found that Megan really significantly overestimated a lot of the isoprene emissions and this is in the term for the basic inventory so this is the emission flux under standard conditions but these other terms could matter as well so you have leaf area index the sensitivity to the above canopy radiance sensitivity to temperature leaf age distribution and soil moisture so all of those things have to be in there to get the Megan emissions right and I'm not going to go too much more into that but there's seasonality that we can see with the Moffett Co but I just want to show the basic comparison for for the spatial pattern between CO and isoprene so this is again estimating all the co not just the direct emissions of co2 plants but what plants contribute that turns into CO say from formaldehyde or isoprene and then formaldehyde and you can see that basically you have the highest values in the same places where you have the high values of isoprene and we see that a similar pattern for top-down I should sorry I should say that this is estimated from a belgian group bureau and we find that in this case so this is isoprene estimated from ECMWF using Megan and this is the top down and so you can see that the top down is getting a lot smaller values for the tropics mid latitudes and the global value and then we see a similar pattern in the co so this is using geo skem and Megan for the tropics we see a lower value about the same for mid latitudes and then a lower value globally and you can compare this to these other terms from biomass burning and fossil fuels so now we have an additional independent constraint on the Megan model using Co observations as well as the isoprene that's estimated from formaldehyde space observations so I'm - Co emitted species is everyone still with me so what can we do knowing these emissions are an app being able to estimate Co emissions or assimilate Co data so I'm going to talk about now what we can say about aerosols carbon dioxide and methane so first aerosols this is a study done from fires that happened in August 2015 in the Pacific Northwest you can see the motors fire count was really high during that period and this is the aerosol map that you get from the deep blue motifs a OD product and so we're going to look at what we see from the aerosols and the carbon monoxide for the carbon monoxide in the free troposphere or the middle troposphere at 500 hectopascals you really see this pattern of transport across this was like I think across five days but at the surface for those same days you see a map that really matches the fire map especially up in here so so from Moffat data you can see where the sources are and where the transport is and using that you can correct what you know for a OD because total a OD doesn't tell you anything about where it is vertically and so you can actually correct the vertical distributions in in this case the camp can model doing a doing an assimilation of moppet Co all right so now on to our favorite greenhouse gasses so these are the NOAA trends in co2 and methane and obviously we want to know what's going to happen there are they going up or down well we know co2 is just going up but how can we figure out elements of the growth rate and other changes that depend on emissions as well as these feedbacks and interactions between climate energy and water cycles so this is work done at JPL using the carbon monetary carbon monitoring system flux estimation and so it's another another meat-grinder but this time it's taking in co2 from oco-2 solar induced fluorescence from gose at and co from moppet and it's using all those things to try to infer net co2 fluxes gross primary production which is from photosynthesis biomass burning and residual respiration so one of the main results from this was published in 2017 and what Jun Jie did was to look at what happened between 2011 when when we had in La Nina to 2015 where we had an El Nino and relative to 2011 the 2015 only no reason released 2.5 gigatonnes of carbon more than than this La Nina but where it did this is or that the diversity in this is really striking so obviously in Indonesia this carbon was released by fires but what happened in South America and Africa is quite different and they found that the real change in South America was actually because GPP went down so there was less uptake and so the net flux was positive and then in Africa there was an increased respiration so you really need a model like that to try to separate out the different processes and you need different kinds of data to inform that and in this case moppet really is telling you where the fires were so on Tim methane this is a result actually from my brother at JPL and what he found also using the MCMC flux partitioning to figure out the actually in this case methane from fires with that since Pirates have been decreasing this actually decreased methane by about four teragrams a year since the 2000s and in tropical fires I should say meanwhile you have other estimates of methane that were increasing from wetlands about twelve tera grams per year and increasing from fossil fuels about seventeen tera grams per year and if you add those up you get 29 tera grams per year but the observed global increase is about 25 tera grams per year so this really was just a an important piece of the puzzle to figure out that these estimates probably are okay if you include this decrease of for tera grams per year from fire all right now on to what we have done here at a calm in terms of assimilating moppet data to come up with a Moffett reanalysis or a CoV analysis this is work by Ben gal bear and so he's been applying the cam-cam model with dart and moppet observations and I don't think I want to read each of these things but you're basically taking moppet observations at this sort of spatial distribution combining it with the cam-cam model in a forecast mode and producing a reanalysis this is these are the increments from the data simulation so in this we analysis there are four basic runs that we need for looking at what's actually happening because you've assimilated Co so this is the mapa tree analysis that that is reported from a comm and that assimilates moppet every six hours it assimilates updates of co concentrations and it uses an ensemble of thirty cam-cam simulations then there's a control run which is the same setup but it's not assimilating moppet so it's assimilating the meteorology but not mop it then another control run that was cam-cam nudge to the Meribah analysis and that's just to look at different effects of meteorology and then finally a control run that is the same as this one except that now it's using the updated co fields from assimilating moppet and so the only difference with this control run now is co so it has the same meteorology which is mera but you can look at just the difference from Co and not meteorology so looking at all of these this is now the mapa tree analysis the control run with the moppet specified CO and then the regular control run that just assimilated meteorology so this is methane lifetime in years as a result of this tree analysis and this is Oh H versus time as well so what you can see from this is that if you look at the change in methane lifetime from here to here it's about a nine-month decrease because of CO decreasing so again what we said is that because the O H has gone up it's going to decrease the methane lifetime because Co has gone down and what is good about this is that you can see in the control one with just the specified Co observations that it also just it has a different overall lifetime but it follows the same pattern and has the same decrease the net decrease in lifetime so it's not due to a change in meteorology all right finally finally onto the future of carbon monoxide from space so this shows all of the low-earth orbit satellite instruments that measure carbon monoxide starting with 2000 moppet and then everything that's come along since in the infrared and everything that's come along in the shortwave infrared so skia maki was on NB set and that measured till about 2012 I think choke-o me just launched and I'll show that in a moment and then we have a whole variety of measurements from from the infrared starting with airs tests which came and went yahzee a Yahtzee B geodesy which just launched last November and then to Chris instruments on NOAA satellites s NPP and NOAA 20 which also just launched recently so moppet is still the only one that measures both of these and so we'd like to be able to use these in some way to extend the moppet record so I'm not going to go through all of this this is basically a comparison of moppet s NPP Chris and s5p troppo me you can see when they all launched they're all in somewhat different orbits but Chris and troppo me are in pretty close formation so they're following each other or Chris is right ahead of troppo me and then the pixel size for Co and troppo mia 7 kilometers it's better for no.2 it's like three and a half by seven and then now we have global coverage in one day and and then really good effective resolution for moppet good it's not it used to be only 2.5 inverse centimeters but you couldn't do a lot with from the Chris instrument but now they're doing what they call full resolution mode and that's point six to five so that's really pretty good and then for Teahupoo me again since it's a shortwave infrared they're doing just a full column so it's really hard to compare the any delta T so so what troppo me measures now is every day a really nice map of CO and no2 which I'm not showing so this is one of their first results it was in a GRL paper recently and so you can really see this was from October so this is when you have a lot of burning in both North and both Africa and South America you can really see India a lot of the expected sources and then Asia so one of the first things they did was to compare with to ECMWF cams model so this is troppo me and this is camp and the nice thing about cams is that this this already assimilates moppet and Yazzie data so you're really getting the benefit of comparing this was not available when Moffitt was launched so but for Teahupoo me they can really look right away and see if they're getting something similar to a reasonably accurate model and you can see there are some differences but for the most part they see a lot of the same patterns so here's how Moffitt compares to troppo me this was just a very preliminary result and we're working on more detailed comparisons now but but basically we're almost on a one-to-one line here which was very gratifying to see so so we really think that it's going to be pretty easy to use troppo me data along with Chris data to extend the moppet record and so finally what what we expect to do this is a simulated result for both Chris and troppo me but what we expect to see is that this is an example of the averaging kernel multispectral averaging kernel for Moffitt using the thermal infrared in The New Yorker red this is Chris alone troppo me alone and then Chris and troppo me together and what you can expect to see from this simulation this is the Geo skem model for a CO near the surface if you just use troppo me you do pretty well but you don't really see much over the ocean because it's looking at the shortwave infrared and you need reflected sunlight which is only good over land in a couple cases over clouds you can do it and that's what troppo me does for ocean observations is they use the reflected sunlight off of clouds but near the surface you're not going to see anything and then Chris alone is sort of like the moppet thermal infrared result and then troppo me alone I'm sorry this was the joint product and troppo me alone is fairly similar with no observations over the water so we were hopeful that we can use Chris and troppo me together to get a similar product from the joint thermal infrared shortwave infrared moppet data finally in conclusion we see global Co concentrations decreasing but still have a significant internal variability from fires we understand changes in CO emissions and these are critical for chemistry climate model modeling the top-down constraints from satellite Co observations could be could be used to reduce uncertainties and biogenic emissions and satellite Co observations are essential for assessing the global carbon cycle so we have promising results from troppo me and these could be used with Chris to extend the Moffett Co record of multispectral Co so thank you we have a little time for questions I really need to take the whole hour hi I have a question of the moppet reports a surface value for CEO is that true correct have you actually how is that constraint how you did you get it actually represents the near surface layer so it's the average from the surface to whichever pressure level on this ten pressure level grid it is so let's say that the surface is that ten thousand and you're it's it's actually going to be at the pressure in between ten thousand and nine hundred hectopascals what one thousand so so yeah that that vertical profile is reported on ten levels but they're actually representative of the full layer average because you're right you can't measure something with with no layer depth anybody else so I'm interested in that the estimation of Oh H from Co it's probably just a direct effect of how much Co there is but I'm wondering if the actual seasonal cycle or spatial distribution of Co say hi southern latitudes would actually tell you something about the Oh H concentration I think C alone would not that I mean we're one of the things like that Kazu Miyazaki has done with trying to estimate do an assimilation of multiple species to estimate Oh H I think it's more promising this is just the result it's not it's it's not it's like how Oh H changed basically because you assimilated Co but it's not the same or it's not the absolute Oh H value so and it doesn't necessarily comment on other evidence of internal variations in oag knows more no it's only the change because of assimilating Co any other questions well let's thank Helen [Applause]