NASA ARSET Case Studies in Trace Gas Monitoring with North American Geostationary Sensors

0:01 Speaker 1: Welcome back everyone to our RSEET training series Geostationary Remote Sensing of Trace Gases for Air Quality Applications in North America. Today is part two of our training series, Case Studies in Trace Gas Monitoring with North American Geostationary Sensors. My name is Christina Pistone and i am a research scientist with the Bay Area Environmental Research Institute and NASA AMES Research Center in the San Francisco Bay Area, and i'm the r set lead for this training. To remind everyone of our schedule, today, we are on Part two and we are going to focus on case studies in trace gas monitoring with North American geostationary sensors.

0:39 Speaker 1: There will be one homework assignment after this training. It will be available today after Part two is complete, and it will be due two weeks from now on February fifth. A certificate of completion will be awarded to those who attend all live sessions and complete the homework assignment before the given due date. Once again, my name is doctor Christina Pistone. I'm a research scientist at Bay NASA AIMS and I'm happy to be joined today by doctor Aaron Nager, who is the TEMPO Mission's application lead at NASA Marshall Space Flight Center to recap what we presented in Part one.

1:12 Speaker 1: TEMPO is a geostationary hyperspectral instrument with coverage over North America. TEMPO provides data on atmospheric trace gas is relevant to air quality, including nitrogen dioxide, formaldehyde, ozone as a total column product and as a beta ozone profile product, and the TEMPO Level two in Level three data can both be visualized in the NASA Worldview tool. The quality assurance settings that are applied in Worldview are suitable for qualitative applications.

1:46 Speaker 1: The objectives for today's parts. By the end of today, attendees should be able to evaluate TEMPO trace gas products in Worldview to anticipate short term air quality risks such as high concentrations of ozone precursors, and should be able to determine in the facial patterns, temporal trends, and likely sources of trace gases related to wildfire, smoke and urban area air pollution events using the TEMPO data available in Worldview and today, we're going to present two case studies, one on fires in the Colorado Front Range from July of twenty twenty four and one on urban air quality conditions in Eastern Texas.

2:22 Speaker 1: Some bookkeeping for how to ask questions. Please feel free to put your questions in the question box and we will try to address them at the end of the webinar. You can put them in the Q and A box within WebEx at any time during the session and at the end of the session. We will try to address as many questions as we can and we will post these questions and written answers to any remaining questions on the training page within a week after the training And now I will turn it over to aerin.

2:58 Speaker 2: Thank you Christina.

2:59 Speaker 3: Now I'll be sharing the ozone case event that occurred over the Coloroffront Range in July of twenty twenty four.

3:07 Speaker 2: So to kick off this use case.

3:08 Speaker 3: Analysis, we're going to look at some true color imagery. Here to the left we have a true color image from Modus Saqua. This is from Worldview and we're zoomed in on Denver. Here they were centered on Denver and this is the first day on July twenty nine, twenty twenty four, when the initial wildfire erupted across the front range. You can see the grayish color intensity here tied to that wildfire smoke plume. Then on July thirty, the following day, this is again a true color of modus image and the fires were becoming more active.

3:44 Speaker 3: On July thirty, we are seeing a larger area here of grayish color intensities tied to that smoke plume. And this is a nice picture taken by Dan Welsh at the colorad Department of Public Health and Environment during his drive to work, where he had this nice view of the Stone Canyon fire and the Alexander Mountain fire.

4:05 Speaker 2: During this event and overall.

4:07 Speaker 3: There were numerous ozone exceedences that were observed across the Front Range during this month. During the first part, in the middle part of the month, those remote mainly the ozone issues were mainly from wildfires across the US West region from regional sources. Then by the end of July twenty twenty four, those local lies wildfires started to pop up across the Front Range, which impacted air quality in the area. So looking at that time period at the very end of July twenty twenty four, when those localies wildfires started to erupt, we are looking at the ozone the air now ozone monitor maps here, there were numerous monitors that did measure unhealthy ozone levels during this time period.

4:55 Speaker 3: July twenty eight was the day prior before the day before the localized wild wildfires started to erupt July twenty nine. Now we're seeing these healthy for a sensitive group ozone levels being observed being measured across the area, and then by July thirty one, we are actually seeing unhealthy for all groups ozone levels that were unhealthy for all groups and those were nearby the Denver Boulder area there. So what happened here, Well, let's take a look at the tempo data. This is two different animations here, with the one on the left is a troposphere two animation the hourly IN two data from morning to evening across the area.

5:41 Speaker 2: Zoomed in on the front range.

5:42 Speaker 3: Here, and we're seeing is this large n O two plume on July twenty nine that was associated with that wildfire smoke event, which you can also see the transport of that No two plume to the east during the afternoon and also kind of your typical urban two tied to the urban area of Denver Boulder in the morning as well. But that in O two plume from the wildfire was really distinct in the region. And looking over to the right at the total column from Adehyde, we also see a large form Adehyde plume that was tied to that wildfire smoke event.

6:21 Speaker 3: So we're going to see that here in a minute that started to pop up and then being transported to the east as during the later afternoon hours. So we definitely saw large increases in both n O two and from Autohyde due to the Alexander Mountain wildfire. Now looking at tempo total calm ozone same day on July twenty nine, and we it was really interesting that we saw this ozone production within the wildfire smoke plume in the total calumn ozone animation here to the top right. Those play this again here and you're going to see in the later afternoon hours this increase in ozone that occurs within the vicinity of that smoke plume as is being transported down wind during the afternoon, and the ozone monitors in the region did measure ozone concentrations of seventy five and seventy one ppb across the area.

7:24 Speaker 3: We're kind of zooming in on those monitors right here, and a zoomed in map, and this is also a zoom in to the modus aqua true color where we can clearly see the grayish color intensities tied to that smoke plume location that were impacted the ozone levels vanilla neat product that can be useful for tracking wildfire smoke is a tempo UV aerosol index. We did mention this in Day one, the Day one training, and now we're showing it for this wildfire smoke event based across the CEO four and range, and the temple UV aerosol index helps to distinguish and track absorbing aerosol particles in the atmosphere like smoke and dust for example, which are generally tied or associated with positive UV aerosol index values.

8:16 Speaker 2: And you can.

8:17 Speaker 3: Clearly see those high UV aerosol index values in this animation here on July twenty nine. Again, we'll see this pop up here in the midday and the later afternoon hours where we have these high values of UV aerosol index tied to that wildifier smoke plume and also the transport that we see down wind during the later afternoon hours.

8:44 Speaker 3: Another tool that can be useful when looking at Temple data is to do averaging a tempo data throughout the day that we did here, we did a daily average of the troposphere in two data from Tempo for July twenty eight, so for all the morning to evening scans on July twenty eight, we average these onto a point zero two degree grid and then same kind of operation on July twenty nine and July thirty. So now we're clearly seeing a few things here. One, this increase or this two column increase tied to the oil and gas well region to the northeast of the Denver Boulderate region.

9:26 Speaker 2: That's pretty distinct on July twenty eight.

9:29 Speaker 3: But then as those wildfires erupted on July twenty nine, we're now seeing this strong n O two signal tied to that wildfire, the Alexander Mountain wildfire near Fort Collins. And then by July thirty there was another wildfire, the Stone Canyon one became more active to the south, and now we're seeing that n O two hot spot tied to the Stone Canyon fire and along with the two associated with the you know, the urban emissions from Denver Boulder.

10:02 Speaker 3: So now looking at formaldehyde, same kind of operation for No. Two, but we applied it to formata Hide doing the daily averaging of fromata Hyde for July twenty eight, July twenty nine, and July thirty, and we're not seeing much much increase in from alle Hide on the pre fire day here maybe a little bit across the oil oil and gaswell region. And then on July twenty nine, July thirty, when those wildfires became active, now we're seeing those large wild those large from aunto Hide plumes that are very distinct in the from auto Hyde maps with the Alexander Mountain wildfire and also a from all to Hyde hot frode Hyde hotspot tied to the Stone Canyon wildfire location.

10:51 Speaker 3: Again, I would make note here that these maps that I showed here we're using a cloud fraction filter less than fifty percent and silver thing angle lesson eighty.

11:03 Speaker 2: Okay, now I'm sharing.

11:04 Speaker 3: Here just the full picture in terms of the two maps at the top and the fromto Hyde maps at the bottom, again using that same cloud for action filter that we applied on the previous slides.

11:16 Speaker 2: Now if we tagle back and.

11:17 Speaker 3: Forth here between a using a cloud fraction of twenty percent on this slide here versus fifty percent on the prior slide. Now we're seeing how using this stricter cloud for action filtering of twenty percent and actually obscure the fire the wildfire smoke plume.

11:38 Speaker 2: So looking around going tiling them back and forth here.

11:42 Speaker 3: Between those two these two slides, we're seeing much of that two plume associated with the Alaxander Mountain wildfire disappear when using the stricter cloud fraction and also less of a signal from the Stone Canyon fire when applying that stricter cloud fraction threshold. And similar depiction there at the bottom with framadehyde. So we tackle back and forth, we're seeing how that from auto hyde plume becomes much less apparent when using that stricter cloud fraction filtering.

12:17 Speaker 2: And the main point here.

12:18 Speaker 3: Is to caution users that a certain cloud fraction filtering may not apply at all kinds of scenarios, so the application is important when applying a cloud fraction filter.

12:33 Speaker 2: Okay, now we're in.

12:33 Speaker 3: Look at the ratio of fromadohide to INDO two, which can be a tool to indicate the likelihood of near surface ozone production. And you can see the two different maps here during the pre wildfire period from July twenty seven to July twenty eight, and then the active fire period from July twenty nine to July thirty. So are two day average maps of mona high to two ratio here, and this color bar here is showing how the scale goes from these nock saturated conditions to NOx sensitive conditions, which are indicative of the ratio from an high to n O two.

13:15 Speaker 3: And we see much different regimes pre fire to active wildfire. So the wildfire smoke in general promoted more efficient ozone production in the region during the active fire period. And we're seeing here our large two increases in the NOx sensitive regime near the wildfire sources. So we have where this Stone Canyon and Alexander Mountain fire locations were here, we're seeing that being converted over to a nock saturate condition in the active fire period due to the large increase in two. But then farther downwind where we had that large transport of Forrmadehyde, we see this oil and gas will region turning more not sensitive with that transport of from aunt of high downwind in that region, but less.

14:05 Speaker 2: N two overall.

14:08 Speaker 3: Also we see that the transport of smoke also likely led to additional increases in ozone production in this NOx saturated regime in the Denver urban area by July thirty one. So looking at our July thirty one ozone modern map again to the right, this is also when those unhealthy ozone levels for all groups were measured and around the Denver Boulder area, as this smoke was transported across this region by July thirty one. All right, so quick validation check in here Tempo versus Pandora two.

14:44 Speaker 2: This is for the Boulder.

14:45 Speaker 3: InCAR site, and I's want to quickly mention here how there's overall good agreement in terms of the monthly and seasonal pattern of No. Two between Pandora and Tempo, and also a general low bias into two compared to this Pandora Boulder in Car site. So overall Tempo seems to perform fairly well. This is all based on the version three data, of course, so okay, now look in more detail here to the monthly correlations between Tempo and Pandora No. Two to the right here, and we see strong correlations throughout much of the year.

15:25 Speaker 3: However, there are lower correlations in the late spring and mid summer time period here have shown in this figure to the right. Again, this is based on the version three n O two data. This is an active area and we have actual validation with our current version four n O two data, and we believe much of this will improve when we do the validation for our version four data, So stay tuned for those results in validation and I'll just then pass it over to Christina for our case one exercise work.

16:03 Speaker 1: Thanks Erin. So now we are going to explore this case in a little more detail in Worldview. So you will remember from the part one, we're just going to go to Worldview dot Earth data dot NASA dot gov. I've already preloaded this here at the bottom left you can see this is the date that Aaron was just describing. So we have twenty twenty four July twenty ninth, And because we already have the tempo data loaded in this bar, we have hour and minutes which is two pm UTC, which is eight am mountain daylight time in this area.

16:41 Speaker 1: So in contrast to the case from yesterday we were little on the coast, we are going to be sure to turn on the reference layers of Worldview. So we're going to turn on some place labels and then the coastlines, borders and roads here. So this gives us a little bit more of an orientation and to where we are in the middle of the rocky mountains, so we can situate ourselves a little bit better. Now we can see that the labels are coming up. We have Denver and Boulder right there in the middle, and once again we are looking at some basse layers from Tarra Modus is the default Terra is a morning overpassed, so it's a time when we haven't yet had a the fire has not yet started.

17:30 Speaker 1: We can toggle off the terror layer, turn on the Aquamotus layer, and now we're starting to see. Just to orient ourselves, we can see it in the true color imagery that here is one of the wildfires that we're going to be exploring in more detail. We can see that by the afternoon overpast time, we're seeing the visible smoke from this event. So we'll just leave that on to orient ourselves. So as yesterday we are going to start with nitrogen dioxide. We are going to do level three vertical column and tropospheric values.

18:09 Speaker 1: So when we turn on this layer two pm UTC is again eight am local time, and we're looking for the event sort of in this area. Nothing really shows up. Again, We've set the increment to be one hourly, so when I do this little increment date button, we're seeing that we're stepping through the Tempo two data one hour at a time. And as we do that, we see some sort of signal from the cities until we get to eighteen UTC, which corresponds to about noon local time, and now we're seeing the beginnings of this wildfire event that's happening in this area.

18:53 Speaker 1: And again as I mouse over it, we can see at the bar on the left, we're seeing what the values are as corresponding to where my cursor is here compared to something a little bit more background in other areas, and we're also seeing some missing data here. These are likely clouds. So the fire event is first being seen by Tempo at eighteen hundred UTC, and as we step through a little bit we can see the evolution of this event through time, so it's not totally consistent throughout the day. It varies.

19:35 Speaker 1: As we get later in the day, we can see an interesting feature which goes to what Aaron was talking about in terms of the data filtering, particularly right here, which is at midnight UTC, which corresponds to six pm local. The plume seems to be quite strong here based on the values sort of at the edge of the plume, and again in scenes with very heavy smoke, we may be seeing those scenes be misidentified as cloud and it is reasonable to infer that that's what's happening right here because we have an area where there are no two values visible but on the edge, and based on the previous time series as we go back in time, we know that there were significant no two values associated with this event.

20:33 Speaker 1: So that's just something to keep in mind as we're looking at these data. So that was nitrogen dioxide. We can also toggle off the nitrogen dioxide. We can look at the formaldehyde values here. So we're just going to go turn on formaldehyde level three vertical column and let's return to our beginning of the day and again earlier in the day, there is no event to be seen. As we step through, we're again seeing in the formaldehyde as we did in the nitrogen dioxide. The plume starts to be visible around eighteen hundred UTC nineteen UTC twenty The plume is more visible later in the day, and again we have some filtering here which is indicative that likely the thick smoke is being flagged as cloud in the routine.

21:46 Speaker 1: But nonetheless we're still able to mouse over and see that their elevated values. Again, formaldehyde is a more noisy product, but we're able to get a sense for what the values are in the downwind plume of formaldehyde associated with this fire, So that is formaldehyde. We can also again look at the ozone product, and so we are going to look at ozone level three and calumn amount right here. Let's turn on this layer. Let's again go to the earlier part of the day. And as you'll remember from the case study that we showed yesterday, the ozone values tend to be much more uniform.

22:37 Speaker 1: They're dominated by the stratospheric component. But we can do some things to get a better sense that the ozone production that Aaron was talking about. So as we did in session one, we're going to go into the view options and play with it a little bit. So here for thresholds, this is showing basically the whole range of the ozone products or the ozone values that we have for this product. And in order to better be able to visualize the ozone signal that might be associated with this particular event, what we're going to do is take the thresholds and we're going to limit the values that it's viewing.

23:22 Speaker 1: That limit the values that are being displayed. Basically, so we're going to sort of exclude the lower values. Let's go to about two eighty five Dobson units. It can be a little fiddly, but I don't think we got it. And then let's also take the upper limit and squish it down a little bit further, you like, right there,

24:07 Speaker 1: and then we're going to tick this squash palette box.

24:14 Speaker 1: So now we've taken the ozone values, we've constrained it to only be showing a small range of the values to sort of be better able to look for features that we suspect might be happening in other parts of the atmosphere. Even though want to emphasize the ozone product that is in worldview is a total calm value, so it's going to be dominated by the stratosphere of component. Since we've done this squash palette, we've constrained all of the colors to only be visualizing as a smaller range of the values, so it allows us to see if we look in the map, we can see a little bit more of the variation which we were not able to see when we were looking at the full range of total colum ozone values.

25:04 Speaker 1: As we saw previously at two pm UTC, their fire had not happened yet, so we're going to just go forward a little bit. We can see that there's some variation that's happening of these values, and by the time we get to eighteen hundred, this is again where we saw the onset of the fire in the two and the formaldehyde data and the true color imagery for that matter. As we step through the times after that, we are able to see some variation in the values here, which because of the context and what we know, and again this is a qualitative analysis of the scene, but because of what we know, because of what we saw from the two data and the formaldehyde and the production that we expect to be happening photochemically in the trumposphere, we may be able to infer that perhaps at this time, in this location, there is enhanced ozone production happening in the troposphere, although again to emphasize this is the colon value.

26:16 Speaker 1: But as we step forward we're at twenty one UTC now twenty two. We are seeing that there is some values here which are elevated and which correspond to where we know which correspond to where we know the smoke plume associated with that fire was happening. So this is a more qualitative analysis, but it can be used to diagnose areas where we might want to look into more detail, more quantitative, either from ground sensors or from other products. So that is ozone. Let's toggle off that layer, and the final thing that we want to look at for this particular event, we are going to look at the aerosol index.

27:09 Speaker 1: Actually before we do aerosol index, because we mentioned the nitrogen dioxide patterns here later in the day where we saw very strong smoke plumes, particularly here around midnight UTC. We're seeing

27:38 Speaker 1: what we expect is as a very thick smoke plume which is masking out the No two data. Another tool that we can use in order to be able to further contextualize these data. There is also a cloud fraction layer which we haven't really looked at in detail previously, but we can turn on this cloud layer and just for clarity. Let's turn off nitrogen dioxide for a minute. The cloud layer again as similar to the other products. As you mouse over it, you can see here these dark values are cloud free conditions, and the whiter yellow white values are high cloud fraction, and above fifty percent is where we've cut the cloud fraction threshold for cloudy versus not cloudy, similar to what we did for ozone.

28:31 Speaker 1: Just now we can go into view options. We can make it a little bit more easy for our human brains to process. We're going to take these thresholds and we're just going to mask out everything about fifty percent.

28:59 Speaker 1: So now we've done is we've masked out all of the cloud free conditions essentially, so the colors that remained here more the yellows and the whites. This gives us a sense of where clouds have been identified as greater than fifty percent, which is our threshold. So if we then turn the nitrogen dioxide layer back on, we're able to see that this corresponds pretty closely well. This corresponds to where the missing data in No. Two are As we talkle this on and off, we can see that the missing values here, most of them are due to the cloud mask masking out values which have been flagged as clouds.

29:44 Speaker 1: Even though for this particular area right here with the plume, this is from our context from what we know of the scene, we can infer that this is quite likely very thick smoke instead of clouds. So one way again to better deal with this these types of situations where the retrievals might be challenging. Let's turn off n O two and now we're going to go to the aerosol index. So we're going to do aerosol index ozone level three. And here we are at the end of the day, midnight u GC, and we can see at this late time in the day it's really a quite significant plume that has been identified as absorbing aerosol in the atmosphere.

30:40 Speaker 1: So this is has been identified as smoke. So looking at this aerosol index in conjunction with the nitrogen dioxide products, this can give us more context of the scene to be able to determine that this is likely absorbing aerosols rather than clouds for this particular scene. Just for completeness of observing the scene, let's go back to the beginning of the day after the fire onset. So if we go before where we know that it, if we go before the onset of the fire, if we're at fifteen hundred zulu nothing eighteen, nineteen twenty, we can get a better extent of the We can get a better sense of the extent of the fire and the extent of the smoke distribution associated with this event here as it progresses through the day, in addition to being able to identify this as smoke as opposed to a thick cloud for this particular event.

32:00 Speaker 3: Thank you for that demo, Christina. Now we're going to dig into another use case looking at urban pollution and emission sources in the East Texas area and for two different time periods June twenty twenty four looking at the version three tempo data and another event in October twenty twenty five after the release of the version four tempo data. So zooming in here on the East Texas do main to the left. Here, I'm showing the Global power Plant database, and we are indicating or denoting these locations of the black circles here, which are locations of coal fired power plants in the region.

32:43 Speaker 3: And the larger the circle, the larger the capacity of the power plant. They looking at our Tempo tropospheric two animation here for this one day on June six, twenty twenty four, from morning to evening. One thing we noticed right away is a lot of n O two data, so this is a very cloud free day, a great day for TEMPO to gather data across the region. But also looking at the high n O two column data and evolution of the in two columns across the Dallas and Houston area as we normally see, and also some higher in two columns in the locations of the traffic corridors in the region.

33:30 Speaker 3: However, going back to the coal fired power plant locations, we are are also denoting those locations on the n O two map. Here we are seeing that very large capacity power plant to the southwest of Houston, and we're seeing a large inn O two plume in the vicinity of that power plant, which is you know, we see that being transport that plume being transported south in the later afternoon hours. In addition to that, we are seeing other areas is of n O two hot spots tied to those coal fired power plant locations.

34:05 Speaker 3: One in particular right here to the west of that large coal fire power plant we're seeing this evolving n O two hotspot tied to that power plant. So it's kind of showing the ability for TEMPO to monitor, to capture and monitor in two hot spots related to coal and also gas fired power plants in certain locations. So doing more of a quantitative look at the actual two column values tied to these power plants on this slide here.

34:39 Speaker 2: But first off, I'll note that.

34:43 Speaker 3: You know two column values and diurnal evolution of those n O two column values in and Dowtown urban location outside of the power power plant, and we're seeing here is also we have Pandora total n O two on this plot as well in black and the Tempo two total two data and red. The main point here I wanted to touch on is the increase in total two column in the morning late morning time period and also this peak that we see in the later afternoon time period, and that's kind of when we often see traffic starting to peak in the area.

35:23 Speaker 3: So those two peaks could very likely be tied to the increase in traffic and the increase in emissions related traffic. And looking at the coal fired power plant locations, though we'll start with the one and here to the north, and the evolving trice was the smaller capacity power plant, and we're seeing this diurnal evolution of O two, a peak happening here in the midday time period on June sixth, twenty point four, and then kind of a decreasing trend in the later after and afternoon. A much different look at the location of the lar the largest coal fire power plant in the region, where we're seeing these relatively low values throughout the morning and mid morning, but then a large increase around mid day and that continues to increase in the later afternoon hours as we kind of saw in the animation on the previous slide as well, so really showing the ability for Tempo to track these inn O two columns and associate with both the urban areas but also these hot spot areas of N two like coal fired power plants.

36:32 Speaker 3: This is for another day after the release of the version four Tempo data on September seventeenth, twenty twenty five. This particular day is October fifteenth, twenty twenty five. Name zoom in across East Texas, and once again we see this as a day where there is a lot of cloud free the very cloud free day temple is able to perform its retrievals and more kind of picture as we saw on the previous animation where we have these large n O two column values tied to the urban areas of Dallas and Houston, and that evolving trend of two throughout the day across those large urban areas, but also this large n O two plume that is once again in the vicinity of that very large coal fire power plant to the southwest of Houston, and we see that being transported to the southwest in the later afternoon hours there so, and also we see hot spots tied to those other areas of coal fire power plants in this domain as well.

37:40 Speaker 3: Now taking a look at formaldehyde. Now we have a total column from Audehyde on October fifteenth, twenty twenty five to the left here and the same tempo tripospric you know two map to the right as we saw on the previous slide. And the main take home point here on this slide is that we see these increases in from at a Hyde in and around the urban Houston area and in the vicinity of that large coal fire power plant that we saw in the two data. So this is kind of a common theme that we see in this very industrialized area of Houston where we have large, you know, enhancements in from aute hyde in and around that city.

38:26 Speaker 3: So one of the really big products that were released as part of our version four data product release was our ozone profile. So this is the same day on October fifteenth, twenty twenty five. But looking at the un animation of our ozone profile data to the left, I'm just sharing a screen capture of the air now ozone monitors, So the ozone concentrations observed by our ground monitors on this day, and what we're seeing here is on a healthy level of ozone being measured at the at the ground level and the area of interest here that we're going to be looking at more closely here to the southwest of Houston, where we're seeing these uh you know, numerous monitors measuring unhealthy, unhealthy ozone levels.

39:15 Speaker 3: Now looking at our tropospheric ozone animation here in the middle and our zero two kilometer ozone map here to the right, we're seeing here is now the ability of TEMPO to track and monitor propospheric ozone levels that do correspond to what we're seeing at the ground level. So one area of interests in that vicinity of the in and around Houston and nearby that coal fired power plant, where we are seeing these increases in tropospheric ozone column values and also an even more distinct increase in our zero two kilometer ozone column values.

39:57 Speaker 3: So and these really do correlate well with what we're seeing at the ground level. So the main take home here is that tempo, spatial and temporal information on ozone can help supplement the ground based monitors in these type of regions.

40:16 Speaker 3: And we'll move on to our case exercise too, and now I hand it off to Christina for that.

40:25 Speaker 1: Okay, So now we're going to explore some of the visualizations of the tempo data that we can do in worldview of these urban air quality conditions that Aaron just talked about. So here we have loaded a view on southeastern Texas for June sixth, twenty twenty four, which was a Thursday. We have our place labels on. We can see Dallas up here, Houston, San Antonio. Some things to note in the visible in a little bit of a contrast to the conditions that we were looking at in the first two case studies for these cases, when we look at the true color imagery, we're not necessarily seeing any any visible signals in that true color imagery as we would associated with wildfires.

41:21 Speaker 1: But we're here to look at tempo, so we are going to get into the tempo data. We're going to again focus on the nitrogen dioxide. We have level three data vertical column troposphere. So we'll just turn on this layer and what we can see because we're at twelve o'clock UTC, that is seven am local time at this time of year, so we're seeing just the beginning of the day. This is where the beginning of the sunlight extends. So we can increment at thirteen UTC. We can start to see the full scene again.

42:08 Speaker 1: As we mouse over the Dallas area, the urban area around Dallas Fort Worth, we can see the higher values associated with the urban emissions, the same thing with Houston, and then some other signals throughout. So we can just step through the day, step through the tempo observations which are available for this day. The time step has been set to one hour, so if I do increment date, we're seeing that it's stepping one hour at a time. And as we do that, we can start to see some of the diurnal, some of the time varying signals that are happening during this particular day.

42:50 Speaker 1: Again, this was a Thursday, around fifteen hundred three pm UTC is around ten am local. This is about rush hour. We can sort of see some elevated nitrogen dioxide between San Antonio and Austin. If we zoom in a little bit, we can see that this corresponds to the highways that go between the two metropolitan areas. So it's quite possible that this is a signal associated with some of the traffic emissions during a typical weekday. We can step through. We're seeing that there's some variability, there's a pretty consistent signal of nitrogen dioxide associated with the urban areas some of the traffic emissions.

43:52 Speaker 1: And if we go back to the beginning of the day, we can also look at the feature look for the features that Aaron had highlighted, specifically those cold fire power plants do that again, So if we go back to the beginning of the day, we can change the view a little bit to draw out some of the more strong features of higher nitrogen dioxide. So similar to what we did, we can go to this view options little slider similar to what we did for ozone previously. We're going to take the thresholds and we're just going to essentially mask out the lower values.

44:49 Speaker 2: Of NO.

44:49 Speaker 1: Two. So we've just masked out the lower values to be a to better highlight the areas of higher nitrogen dioxide in this scene. So if we step through the day again, we can see the highlights of the traffic that we noted before. It stands out a little bit better when we're only focused on the elevated values. And one other thing to note, as Aaron had pointed out in his slides, there are some coal fired power plants in the area. If you keep an eye on this area of High N two right here, and there was another one just to the southwest of Houston right there.

45:43 Speaker 1: Keep an eye on those. You can see that they are associated with elevated two in a way that the surrounding areas are not necessarily showing. So you can pull out some areas of focus for potential sources in this way.

46:10 Speaker 1: So that's the day of one example of a weekday. If we go back to the beginning, now, we're going to do some comparisons with across multiple days. So we're going to go to the comparison feature in Worldview. So this start comparison button at the bottom left. When the comparison panel first opens, it has the layer that we were just on, which is the June sixth example. It's added a different layer for a week previous. That's just the default that it chooses, and we can see that this is a particularly cloudy case, so not necessarily great for looking at tempo data.

46:54 Speaker 1: Anyway, We're just going to go a couple days later. We're going to go to June eighth at the same time, and the time incrementor has defaulted back to six minutes. Let's just change it back to an hour to be associated with the tempo resolution there. So now we have two days. We have on the left June sixth, twenty twenty four, on the right we have June eighth, twenty twenty four, and we can move this slider back and forth to see the differences between the two. And so this is a weekend day, a weekend day, and a weekday day, and some things jump out.

47:41 Speaker 1: Let's step forward a little bit. Unfortunately, we have to do it one step at a time, each one individually. But let's compare some apples to apples. So if we go to fifteen UTC, that's ten am local time on June sixth, we see that there's a pretty visible association with commuter traffic. Likely that's not necessarily the case, or that's not the case for the two days later June eighth. And of course this is a sample size of two. This is a qualitative examination of the conditions over two separate days.

48:32 Speaker 1: But this allows us to identify locations and times and potential areas of interest for further quantitative analysis, hopefully using some of the other tools that are available to an analyzed tempo data. So let's go a little bit later in the day just to check a different timestep. This is going to nine pm UTC,

49:09 Speaker 1: which should be four pm local time. Okay, here we are at nine pm UTC, four pm local time. We can again see traffic is present at this time, that's not necessarily present, and we don't see the stronger signal from these power plants that were identified either on the weekend versus the weekend or the weekday versus the weekend days. So we can explore the spatial patterns a little bit and we can identify locations and places for further examination or further analysis. One more example, just from these snapshots of two days, Let's go to twenty three UTC about six pm.

50:09 Speaker 4: Let's go to the same time on June eighth.

50:17 Speaker 1: This is an interesting example because it also highlights the differences in spatial distribution of urban air pollutions. So if you look at Houston on June sixth here and June eighth here, Admittedly the magnitudes are different, but it seems quite likely that there's something meteorologically going on different because the highest concentrations tend to be in the south, and then they tend to be in the north for both the Dallas and the Houston metropolitan areas for these particular dates. So again a tool to examine, a tool to identify areas and times for further examination.

51:08 Speaker 1: Okay, so another thing that Aaron showed was a second snapshot in time from October of twenty twenty five. Unfortunately, the version four data are not available in Worldview at the time of this taping, but we are going to go and explore some of the seasonal differences by going to October of twenty twenty four.

51:35 Speaker 1: October tenth was also a Thursday, as was June sixth, twenty twenty four, and the differences between the two. So now we have June sixth on the left and October tenth on the right, and the differences in magnitude of two values that we can see when the scales are the same are really quite significant here. So there's much higher concentrations of No. Two in this particular day several months later compared to the signals which are in June. So here we have October and let's just step through to see what the daily cycles look like.

52:32 Speaker 1: Remember last time we started at twelve UTC, which was seven am, it's actually six am because of the time change that happens in the US, So if we go to twelve o'clock at six am, there's no data because there hasn't been sunlight yet in this particular time of year. So let's just go a little bit later, and here we're starting to see at thirteen hundred we're starting to see some of the data to the east. By the time we get to fourteen hundred, we can see that there's a pretty significant we're seeing a view of the Houston and Dallas areas, which are within Tempo's view at this time of day, and so fourteen hundred is eight am local time at this time of year.

53:25 Speaker 1: So if we go and look at the nitrogen dioxide, we can see that the values are much higher than they were for a similar day in June. If we move a slider back so we can look at it a little bit better, expand that let's go to the same time of day

53:48 Speaker 1: at fourteen hundred if we'll remember what that looked like.

53:58 Speaker 1: So here is October and here is June. So these data really allow us to examine potential seasonal patterns as well. And of course, as we know, these are only snapshots in time. This is one particular day, one particular time of day for this location. It lets us identify a place to start for a more quantitative analysis looking at different patterns, either between seasons or at different parts of at different times throughout the day. But we can quite easily see the magnitude of these values that are happening in October, so a late fall.

54:46 Speaker 1: Let's go so we can see the scale right here. We can see that these the maximum values here are much higher in Dallas and across Houston compared to what they were in June, where we had lower values overall, and masking out some of the lower values can sort of highlight that to the human eye a little bit better. So we can see that the values are significant much higher at this time. So we're going to step through the time the diurnal cycle for this particular date in October. So starting two pm UTC is eight am Central Standard time the local time.

55:39 Speaker 1: Here we can step through similar to what we did in June. We can see there is variation throughout the course of the day, still seeing some of the sources associated with transport likely but the No. Two values very early in the day and towards the end of the day,

56:25 Speaker 1: so it was twenty three twenty three hundred would be five pm local. There's still some significant differences that are visible at each of these timesteps. Comparing these two particular cases,

57:12 Speaker 1: so that can give you a sense of how to use this world view tool to identify times and places further analysis to compare particular cases particular days where you may be interested. We can also let's exit comparison now,

57:36 Speaker 1: so we can also set up animations to visualize the full daily cycle for example of a particular parameter. Notice I've slimmed down the variables that we're just that we're visualizing right now, just because this can be a little computationally intensive depending on your network connection and your computer system. But we go down to set up animation in this little video icon, and it'll bring up a tool here, and we have our start time here June sixth at one pm UTC, two June sixth at eleven pm UTC.

58:23 Speaker 1: So let's yeah, let's do a twelve hour window. It's going to animate the map in one hour increments. You could, for example, pick one the same timestep every day, and then you can also alter the speed of the animation so how quickly it displays them. But three is three is a decent speed. And we've turned on the animation loop. We can enable or disable the animation loop, which just means it's going to play it over and over again.

58:55 Speaker 1: As we do that, we can see down at the bottom. It's giving us an indication of what timestep is being displayed at a given time, and then we can just sort of watch the daily evolution of the two patterns here based on how we've chosen to visualize it. If we want to save this for later viewing, it's easy to create an animated GIF, click on this little save icon with a video on it. It may give you a notification based on whichever layers you've had. It won't save the place labels, but you can just say okay, and then it will give you oh and it also will only it won't preserve those thresholdings that we've done to only display the higher values of n O two if we want to export it as a variable, but we can are exported as an animation mother.

1:00:02 Speaker 1: But we can drag this box to encompass the area that we'd like to save. There's some parameters here based on the file. You can change the resolution. It'll give you a sense of how big it is. And including date stamps is nice so you know exactly what time and place you're looking at. We can say create gift again. This may think about it for a little bit, and then we can see that it is saved this animation over the daily time cycle for this June sixth, twenty four day. We can click download and we can see that that just has saved this to our computer and make a bigger view it in whichever of the gift visualization rules that you would like.

1:01:01 Speaker 1: So let's keep that off to the side for a little bit longer. Another way to compare different cases, we can go back to our October case,

1:01:16 Speaker 1: and we were on the tenth Thursday. Let's go to the beginning of the daytimes. It needs to be manually be done. It's mad in the red there because the ending time is earlier than the starting time, but we will fix that October tenth. October tenth. There we go, and we can see that the blue here is showing where we're going to display it, which is going to twelve hour period. Some of these, as we can remember, will not be showing data because of the lack of sunlight and as we get into the winter, the Northern Hemisphere winter season.

1:02:13 Speaker 1: So now we can see that it is animating as we hope it will over the course of this particular day. If we want to save this animation,

1:02:30 Speaker 1: and it doesn't want to deal with the change thresholds, but that's okay, and it doesn't want to have the place labels, but that's okay. This is about the same area that we had viewed last time. So if we create our gift, you can see that there are a couple of time steps of missing data that will incorporate when you support it in this way. But overall, it can give you a sense of what the diurnal cycle is for a particular case or a particular time open that perhaps pare the two to get a sense for how the diurnal cycles may compare and how the magnitude of the values may compare one to the other.

1:03:43 Speaker 1: So it's again a qualitative visualization, but it can be useful for some purposes to be able to visualize the evolution in that particular way. So one final thing that we haven't yet looked at is other parameters aside from nitrogen dioxide. So if we go back to June sixth, beginning of the day again, let's see if we can do a comparison back to our case in the fall, so summer versus fall. Let's go back here and change our eight

1:04:35 Speaker 1: Let's speak.

1:04:40 Speaker 4: Sure we're looking at hourly, and let's turn off our nitrogen dioxide layer. Let's not do fifteen hundred, let's do sixteen, and let's do from aldehyde here. So similar to what we saw with nitrogen dioxide, formaldehyde as a parameter tends to be, as we know, more noisy. But we can set up a comparison between two different days or two different timesteps within a day to see how those variables compare with one another. So similar to what we saw in the nitrogen dioxide, there are higher values here in this is October on the right hand side, So the main part of the figure right now compared with the values that we see on the day in June, and there is some variability, so we can see that that variation as well.

1:06:10 Speaker 4: So this is again another thing to explore the different variables. Hopefully the tools that we've shown here can give you a sense of how to look.

1:06:23 Speaker 1: At events that you may be interested in or that you want to learn more about or see how they evolve at a broader spatial area, and give a starting place for where you can get into maybe more quantitative analysis or otherwise look at the effects of these kinds of air quality events in your in your area of interest. Okay, and with that, we're coming to the end of the presentation portion of this training to summarize what we've looked at today. Over two case studies, we've examined the spatial patterns, temporal evolution, and likely sources of trace gases related to cases of wildfire, smoke and urban air pollution events using the Tempo data visualization via Worldview.

1:07:21 Speaker 1: The Temple trace gas products in Worldview can be used to anticipate short term air quality risks such as eye concentrations of ozone precursors. The hourly tempo data allows for tracking of diurnal cycles of pollutants and identifying time varying emission sources. The Temple products is good correlation with ground based column measurements, indicating they may be useful to complement ground based air quality monitoring to improve the situational awareness of air quality risks. And finally, the new temp ozone profile products will add more insight to the air quality impacts of near surface ozone and those products should be within should be available within world View shortly, so keep an eye out for those.

1:08:05 Speaker 1: Now, I just want to present a few other options for accessing the Tempo data for more quantitative purposes. NASA data are stored at NASA's Distributed Active Archives Centers or dacks, and the Tempo data live at the Atmospheric Science Data Center highlighted here in yellow. And here are some links which will be provided on the slides associated with this training for some further tools and services which allow for TEMPO data distribution and access and understanding. And here's a link to the TEMPO Operations page, which provides a record of the nominal operations the scans some particular special ops such as we saw with the California wildfires or any unplanned outages.

1:09:05 Speaker 1: You can follow that link and check for particular cases. We also have the Earth Data Forum, which is a place where scientists and data users can connect to answer any questions that may come up regarding utilization of the TEMPO data. And that's a resource available to the community. And now we're at the ending of our training, so to summarize what we've gone over over the past two sessions. Tempo's hyperspectrol capability across ultraviolet to visible wavelengths enables retrieval of criteria trace gas pollutants in the troposphere at unprecedented spatial and temporal resolution across Greater North America.

1:09:57 Speaker 1: TEMPO observes small scale emission sources and fine scale pollutantingradients that have not been adequately resolved by previous satellite missions, and its geostationary hourly and daylight observations and special observation modes allow for monitoring of rapidly evolving pollutants supporting air quality analysis and forecasting. The newly available ozone Profile product which we talked about, offers enhanced capabilities to monitor and characterize ozone concentrations within the troposphere, and as a Worldview provides visualization capabilities of the hourly Level two and Level three trace gas data products with basic quality assurance screenings.

1:10:36 Speaker 1: This can be used to determine spatial patterns, temporal trends, and likely sources of trace gases related to wildfire, smoke and urban air pollution events, so Worldview can be useful in qualitative analysis. For more quantitative analysis, you just will need to apply quality assurance variables to remove tempo data associated with larger errors or uncertainty for their analysis. And hopefully we've given some resources to help those interested in that next step. Regarding homework and certificates, there will be one homework assignment after this training available later today.

1:11:16 Speaker 1: You can access it from the training web page. Answers will be submitted via Google Forms and it will be due two weeks from today on February fifth, twenty twenty six. A certificate of completion will be awarded to those who attend all live sessions and complete the homework assignment before the given due date. You will receive a certificate via email approximately two months after the completion of the course.

1:11:38 Speaker 1: Thank you everyone who's attended today. And now we're going to transition to our question and answer session. Okay, we have a number of questions in the Q and A that are coming in. Any other questions, please feel free to keep typing those in the Q and A box. It should be at the bottom right of your web x screen. There's a little Q and A right there. Okay, So yeah, so question one was it came in a little earlier. It was asking about how to tell the cloud fraction threshold that is used in Worldview hopefully a little bit later in the demo we showed how to do that.

1:12:20 Speaker 1: By toggling the thresholds, tuggling on the cloud layer and then changing the threshold, you can isolate particular values of cloud fraction to better be able to identify where it is. So everything in Worldview is masked out at greater than fifty percent, and by toggling those thresholds you can have a better sense of what the cloud fraction is for different values. There, So and Aaron Please feel free to chime in for anything, but I'll just go through as many of these as I can. So question two, Aaron, do you want to speak to this one?

1:13:03 Speaker 1: This is asking about recommended cloud fraction and solar zenithangle quality control recommendations. Given that we showed in the case study that there are clouds smoke smoke plumes being identified as clouds. Uh, do you have any recommendations for how we can address that or how people can address that.

1:13:27 Speaker 3: Yeah, So, as was mentioned here or is mentioned here, the kind of stricter cloud fraction thrusholds that are within the tempo user guides, for example, are trying to kind of remove those those higher uncertain areas where tempo might be you know, not performing as well. So that is a complication, you know, when you're looking at the wildfire smoke. What can be a complication when looking at wildfire smoke plumes. We don't have a specific quality ish quality control recommendations for using tempo data in those smoke wildfire scenes.

1:14:09 Speaker 3: As we kind of highlighted it in the presentation material, you know that a cloud fraction of less than fifty percent appears to retain the majority of the tempo data within those smoke looms, but in terms of the understanding of the actual uncertainties and errors within those scenes have not been fully evaluated yet. So I think it's an area of ongoing work right now that we have, you know, to better understand kind of the best cloud fraction and also the uncertainty within those smoke looms.

1:14:43 Speaker 1: And retrieval of trace gases and aerosols where they coincide as always a that's a scientific challenge, I think beyond the scope before we're.

1:14:51 Speaker 2: Talking about exactly.

1:14:52 Speaker 1: Yeah, question three, if I wanted to graph the pollution the pollutant changes over period of time, can you download that data in Worldview? We didn't show it in the demo. Worldview does have a charting feature that is in beta. It's available for some of the layers, but it's not available for Tempo and it again is in beta. But if someone is interested in graphing the changes over a period of time on some of the ending slides, specifically tool specifically slide thirty five, there were a number of links that you can explore to download the data and to do a more quantitative analysis of changes over time.

1:15:40 Speaker 1: And hopefully we did show later after this question came in. We showed how to export an animation as a gift of a particular scene. So depending on how quantitative versus qualitative you want the changes over time the visualization to be, that might also sovere purposes. Question for.

1:16:07 Speaker 3: Aaron, did you have some of the ned Oh no, that was great, okay.

1:16:13 Speaker 1: Question for the aerosol product didn't overlap the true color smoke plume neatly? Yes, that is a that is due to the true color imagery being again a snapshot from either Modus or Veers. I believe this one may have been Veers, so those have a single overpass. They are polar orbiting satellites, so it only captured one snapshot at one particular time of day. So that is the base layer. For some scenes in Worldview, you can get the Goes imagery. I believe it only goes back about ninety days, and then they only save the Goes imagery for particular scenes that have been already flagged in advance.

1:17:07 Speaker 1: So the short answer is the true color imagery is showing a single overpast time, and one of the advantages to tempo is it allows us to see the evolution as opposed to just having the one snapshot that we have previously gotten from polar Orders. So question five came in as the demo was whereas we were going through the demo, it wasn't totally clear what we meant by the resolution. The demo focused on the level three products, so those are at point oh two by point oh two degree resolution throughout the field of regard.

1:17:48 Speaker 1: But there is a link. There's a link here in the Q and A documents and it's also on the slides in slide in the slide deck on slide twenty one one. So if you have any particular questions regarding the resolution or particular visualizations, you can go and explore that yourself and hopefully answer all of your burning questions there. And if anything else comes up, please feel free to add it into the Q and A box while we're still here. Question six is regarding converting molecules per unit area to concentration units.

1:18:26 Speaker 1: Arin, do you want to speak.

1:18:27 Speaker 3: To this a little bit, yeah, you know, this maybe a longer response here, but there are techniques or methods to as from the standard molecule for unit area, you know unit that's measured by tempo. Of course, this would involve inclusion of auxiliary data to convert those units, which can lead to additional uncertainties as well even after that, but even after that can version. Just keep in mind that the n O two ppm will still be valid for that column rather than the purpose level of values that you at those AQS data are providing.

1:19:14 Speaker 3: So there's not a direct, you know, comparison in terms of your AQS data versus the converted even if you did the conversion to n O two ppm. So, but there's some more information here though that people can check out. There's a Jupiter notebook that is shown here where folks can go through there and estimate surface concentrations from tempo data using the GEOCF model as additional source of information. So yeah, feel free to go check that link out there if you're intrigued by driving or if you are interested in driving a surface level concentration from tempo data.

1:19:56 Speaker 1: And I'll also note in in the slides there's the Earth Data forums, So any particular questions I think there, I think there's an active community on there for if anybody has any particular questions you can ask ask the community. Internet forums can sometimes be useful. So question seven, how do you create those figures in Houston. I think this was regarding talking to Aaron slides hopefully some of the figure in terms of the gift that was exported from Worldview. This is recorded. You can go back and you can follow through for that, Aaron, Do you want to talk about how you did it?

1:20:40 Speaker 1: It's python code, is the short answer.

1:20:43 Speaker 3: Yeah, Python code is a short answer. Again, there's a good link here that you can check out for the data access scripts that's provided by the ASDC center. So can I recommend clicking on that link there or using that link there to check out more information in regards of that, But yeah, those are just you know, what we did there was download the TEMPO data from the Atmospheric Science Data Center and then run Python code on our system here to produce those maps. So it's kind of the long or the short answer there.

1:21:24 Speaker 1: Our question eight is the question about improving temporal and spatial resolution using combined satellite products and other tools. Is it possible to check trace gas variability over small spatial skills in regions without ground based stations such as Pandora. I think that is exactly the advantage of TEMPO is that it has a very high spatial resolution that it's described in this answer here two kilometers by four point seventy five kilometers at the center of the field of regard. So it does have a lot higher spatial ReSm compared to some other sensors.

1:22:02 Speaker 1: Aaron, I don't know if you want to answer and add anything.

1:22:06 Speaker 2: To that or I think that's great.

1:22:09 Speaker 3: I was a little bit confused by using the using combined satellite products or other tools. But yeah, essentially TEMPO will provide this directly right because of a high spatial resolution alongside the high temporal resolution, so we can resolve those kind of variability and trace gas plutants in those you know, and those highly complex urban areas where you might expect those to occur.

1:22:34 Speaker 1: Sounds good. Question nine is asking when whether Tempo will be added to the Noah Aerosol Watch site. Yes, that is something that has come up that we've all been eagerly waiting for. The We can see that the answer is being written right now. As far as this is in line with my understanding as well, there are plans to include the aerosol products into the Noah Aerosol Watch site. We are hoping mid twenty twenty six, and we are planning a aerosol specific training. So this is the tempo trace gas specific training, and we're planning a tempo aerosol training for after those data are available, so that we can fully demonstrate all the capabilities there.

1:23:26 Speaker 1: And yeah, there's a link to the in person training that is mentioned in this question as well. It was a much longer version of what everyone here has gotten as sort of a crash course of about three hours. It was instead three days. So the materials are available on the website and feel free to look into those a little bit more as well. Anything else, Aaron, That was great. Yeah, you always seem to have the insider infota when the temple airsol.

1:24:01 Speaker 3: Oh, looking forward, I know there is a discussion. Yeah, and then mid twenty twenty six timeframe may or thereafter, so hopefully by may's kind of the UH as early as May of twenty twenty six, I would say, in terms of the timeline, but it could be.

1:24:17 Speaker 1: After of course, but we're hoping I think we're hoping for around yeah, mid mid twenty twenty six. UH. Question ten is Question six was so how to convert molecules per unit area to concentration units? But for ground level ozone an Aaron, I see your I know you can't write and talk at the same time.

1:24:40 Speaker 3: Do you want to speak to that and I'll just speak to it as I just saw this come across right now. But yeah, So again the tempo we showed tempo ozone data, right, we have two different types of tempo ozone products. We have the Tempo Total column ozone and then the Tempo ozone pro file, which provides that additional information on ozone concentration in the troposphere. So that being said, that product does not provide information on ozone concentrations at the ground.

1:25:14 Speaker 2: Or the surface level.

1:25:15 Speaker 3: So we do provide that lower tropospheric ozone concentration which was kind of referred to as a zero f two kilometer ozone concentration that was shown in the presentation, but there is not an actual surface level ozone that one can get at from that product. That would require additional additional techniques to derive a tempo surface ozone value that would be directly comparable to the AQS minors.

1:26:01 Speaker 1: Question eleven is regarding formaldehyde available on Google Earth Engine. It seems like there might I haven't got into that personally myself, but there's some links here in this answer, and again this document will be posted on the website probably in the next couple of days, if not sooner. So yeah, formaldehyde, I guess should be available on Google earth Engine.

1:26:34 Speaker 3: They should what you're saying, they should already be available right now. I thought you right this one, but yes, let me.

1:26:45 Speaker 2: I can verify that.

1:26:46 Speaker 3: I'm pretty sure they're supposed to have from out of hyde data already available in Google earth Engine.

1:26:51 Speaker 2: But I can verify.

1:26:56 Speaker 1: And our emails are also in these slides as well, So if there are any particular questions that can't be answered on again, peruse the earth Data forum as well to see maybe somebody else has the same question. That's always a good place to start for quick answers. But you also have our emails for anything particular.

1:27:19 Speaker 3: Yeah, I'll just say that when I'm kind of googled here when on Earth Engine data catalog for Tempo, and I'm looking at five different products in here right now, including from Autohyde, the O two and also the Ozona total column should be available within Google earth Engine.

1:27:41 Speaker 2: I'm seeing on their main page here.

1:27:44 Speaker 1: Sounds good, And the link to Tempo and Google earth Engine is also in the slides on slide thirty five on the link of Tempo Data Distribution Tools, so you don't need to have I think exactly that link. That might be the same link that are in the slides, but again this will be posted as well. That might be the end of our questions. Give it. I mean, we're, we're. We got two minutes left, so I think we That was actually quite on time, wasn't It filled the time with the questions and hopefully we answered everyone what everybody was wondering about.

1:28:32 Speaker 1: Yeah, I think we can just end it there. Thank you everyone who joined. Thanks again for having so many good questions for us. We hope this was useful and have a good rest of your day and thanks for joining us.

1:28:53 Speaker 2: Thank you every much.