NASA ARSET North American Geostationary Trace Gas Data Products for Air Quality

0:02 Speaker 1: Welcome everyone to our RSET training series Geostationary Remote Sensing

0:06 Speaker 1: of Trace Gases for Air Quality Applications in North America.

0:10 Speaker 1: Today is part one of our training series North American

0:13 Speaker 1: Geostationary Trace Gas Data Products for Air Quality. My name

0:17 Speaker 1: is Christina Pistone. I am a research scientist at the

0:20 Speaker 1: Bay Area Environmental Research Institute and NASA AIMS in the

0:24 Speaker 1: California Bay Area, and I am the RSET lead for

0:28 Speaker 1: this training. Before we begin today's training, I'm just going

0:33 Speaker 1: to say a quick few words about the ARSET program.

0:38 Speaker 1: The NASA Applied Remote Sensing Training or ARSET program, provides

0:42 Speaker 1: cost free training on the use of remote sensing observations,

0:45 Speaker 1: analysis methods, and tools. We provide training in several thematic

0:49 Speaker 1: areas including agriculture, climate and resilience, disasters, ecological conservation, health

0:55 Speaker 1: and air quality, water resources, and wildland fires. Our SET

1:04 Speaker 1: provides trainings both online or in person. Our online trainings

1:08 Speaker 1: are delivered in two formats, live and instructor led like

1:11 Speaker 1: today's training, or asynchronous and self paced like NASA's free

1:15 Speaker 1: and Accessible data. All of our trainings are offered at

1:18 Speaker 1: no cost, and we only use no cost and open

1:20 Speaker 1: source software and data. We try to offer trainings in

1:23 Speaker 1: more than one language whenever we can, including a planned

1:26 Speaker 1: Spanish version of this training later this year. We offer

1:30 Speaker 1: our trainings at a range of levels, so you can

1:32 Speaker 1: find a training series that will fit your level of

1:34 Speaker 1: experience and need. Please visit us at our website to

1:37 Speaker 1: learn more. We'll start today by giving an overview of

1:41 Speaker 1: this training series. Geostationary a remote sensing of trace gases

1:45 Speaker 1: for air quality applications in North America. The Tropospheric Emissions

1:51 Speaker 1: Monitoring of Pollution or TEMPO mission represents a new capability

1:55 Speaker 1: of hourly monitoring of trace gases across North America. It

1:59 Speaker 1: is the first space base based hyperspectral instrument to continuously

2:02 Speaker 1: measure daytime air quality over North America from the Atlantic

2:06 Speaker 1: Ocean to the Pacific and from Central Canada to southern Mexico.

2:10 Speaker 1: These data are essential for understanding diurnal changes in air

2:14 Speaker 1: quality and monitoring real time movement of pollutant plumes such

2:18 Speaker 1: as wildfire and urban emissions. This training will provide an

2:22 Speaker 1: overview of the TEMPLE capabilities and available trace gas data products,

2:26 Speaker 1: and will illustrate how TEMPO data can be visualized using

2:29 Speaker 1: the NASA Worldview online tool. Temple's capabilities will be demonstrated

2:33 Speaker 1: through hands on case studies using the Temple data, and

2:36 Speaker 1: participants will learn how to interpret the TEMPO observations for

2:40 Speaker 1: understanding of the scope and potential air quality impact of

2:43 Speaker 1: some common scenarios. By the end of this training, attendees

2:52 Speaker 1: will be able to describe the basic characteristics of TEMPO

2:55 Speaker 1: and its benefits and limitations compared to other air quality

2:57 Speaker 1: relevant missions and instruments. Identify available trace gas data products

3:02 Speaker 1: from TEMPO and their associated characteristics. Visualize hourly Tempo trace

3:06 Speaker 1: gas data products using NASA Worldview for current and historical events.

3:11 Speaker 1: Evaluate Tempo trace gas products in Worldview to anticipate short

3:15 Speaker 1: term air quality risks such as high concentrations of ozone precursors.

3:19 Speaker 1: Distinguish uses for TEMPO trace gas data products given cloud

3:23 Speaker 1: and solar zenith angle thresholds, and determine the spatial patterns,

3:27 Speaker 1: temporal trends, and likely sources of trace gases related to wildfire, smoke,

3:31 Speaker 1: and urban area air pollution events using Tempo data in Worldview.

3:38 Speaker 1: As a perrequisite to this course, we've recommended that you

3:40 Speaker 1: take our Fundamentals of Remote Sensing training to familiarize yourself

3:44 Speaker 1: with some terminology and concepts that you might hear in

3:46 Speaker 1: today's training. Some examples that are important in today's training

3:51 Speaker 1: include what is meant by spatial and temporal resolution of

3:54 Speaker 1: a sensor, the differences between level one, level two, and

3:57 Speaker 1: level three of satellite data products, and the characteristics of

4:01 Speaker 1: satellites in geostationary orbit. This training will take place over

4:06 Speaker 1: two days, with Part one, which is today we are

4:09 Speaker 1: covering North American geostationary trace gas data products for air quality.

4:15 Speaker 1: It's offered in two sessions at two different times as

4:17 Speaker 1: you can see here. Part two will be two days

4:22 Speaker 1: from now, covering case studies and trace gas monitoring with

4:25 Speaker 1: North American geostationary sensors at the same times. On that day,

4:30 Speaker 1: after Part two, a homework assignment will be posted on

4:33 Speaker 1: the training webpage and a certificate of completion will be

4:36 Speaker 1: awarded to participants who attend all live sessions and complete

4:39 Speaker 1: the homework assignment before the given due date. And with

4:45 Speaker 1: that we're going to get into part one of this training. First,

4:49 Speaker 1: a brief introduction to who we are. As I said earlier,

4:52 Speaker 1: my name is doctor Christina Fistone. I am a research

4:55 Speaker 1: scientist at Barry and with me today is doctor Aaron Nager,

4:59 Speaker 1: the Tempo Mission Applications Lead at NASA Marshall Spaceplight Center.

5:06 Speaker 1: The objectives of Part one of this training. By the

5:08 Speaker 1: end of this part, attendees will be able to describe

5:11 Speaker 1: the basic characteristics of TEMPO and its benefits and limitation

5:14 Speaker 1: compared to other air quality relevant missions and instruments, Identify

5:18 Speaker 1: available trace gas data products from TEMPO and the associated characteristics,

5:22 Speaker 1: distinguish uses for Tempo trace gas data products given cloud

5:25 Speaker 1: and solar zenith angle thresholds, and be able to visualize

5:29 Speaker 1: hourly Tempo trace gas data products using NASA Worldview for

5:33 Speaker 1: current and historical events. So brief logistics for how to

5:38 Speaker 1: ask questions. Please feel free to put your questions in

5:41 Speaker 1: the questions box and we will address them at the

5:43 Speaker 1: end of the webinar. You can enter them as we go.

5:46 Speaker 1: We will try to get to all the questions that

5:48 Speaker 1: we can during the Q and A session, but the

5:50 Speaker 1: remainder of the questions will be answered in the Q

5:53 Speaker 1: and A document which we posted to the training website

5:56 Speaker 1: about a week after the training, and with that I

5:59 Speaker 1: will turn it over to Erin.

6:04 Speaker 2: All right, well, Thank you Christina for that introduction. Now

6:07 Speaker 2: we're in dig into an overview of the Temple Mission

6:10 Speaker 2: and the instrument and its capabilities. A few slides here

6:15 Speaker 2: the Tempo Mission. The major goal here is to provide

6:18 Speaker 2: high resolution observations or measurements on air pollutants across grittered

6:22 Speaker 2: North America every hour during the daytime, and some of

6:26 Speaker 2: the key objectives here are to deliver revolutionary air quality

6:31 Speaker 2: data to the community and the public, improve air quality

6:34 Speaker 2: forecast and alerts, better understand pollution sources, better inform policy

6:41 Speaker 2: and regulation, and enhance health studies from pollution exposure estimates.

6:47 Speaker 2: To the right, I'm sharing a few of the key

6:50 Speaker 2: products or observations from TEMPO, including nitrogen dioxide for modehyde

6:56 Speaker 2: and total ozone and those matters. Showing also the Tempo

7:01 Speaker 2: Field over guard as we call it as well there

7:03 Speaker 2: in red. And at the bottom left, I'm showing a

7:07 Speaker 2: really nice image of a nice picture of our Tempo

7:10 Speaker 2: instrument when it was completely integrated on our satellite IS

7:14 Speaker 2: forty E prior to launch a few quick facts on

7:19 Speaker 2: the Tempo Mission. It was NASA's first Earth Dventure instrument,

7:23 Speaker 2: selected in twenty twelve and it's a joint project with

7:26 Speaker 2: the Smithsonian Astrophysical Observatory. It is a hyper spectral ultraviolet

7:32 Speaker 2: visible spectrometer that is sensitive to policy relevant trace gases

7:37 Speaker 2: and aerosols. And do the bottom right here is a

7:42 Speaker 2: figure showing the observation area that TEMPO performs in. So

7:48 Speaker 2: it performs on two different detectors in the ultra violet

7:51 Speaker 2: and visible, so we're seeing these two ninety four ninety

7:55 Speaker 2: nanometer wavelength spectrum and five forty seven forty nanometer spectrum

8:00 Speaker 2: where Temple provides these very high spectral resolution observations to

8:06 Speaker 2: get at key information on these air pllutants shown here,

8:10 Speaker 2: and the Temple fieldver guard covers greater North America. The

8:14 Speaker 2: standard coverage of TEMPO is shown in the black area here,

8:19 Speaker 2: and however Tempo is able to go outside those bounds.

8:23 Speaker 2: There's a max field reguard coverage shown here in purple.

8:26 Speaker 2: We've done some test with moving Tempo outside the standard domain,

8:31 Speaker 2: but of course our standard normal observations were done within

8:35 Speaker 2: the black area shown there. Temple was launched into geostationary

8:40 Speaker 2: Earth orbit on a SpaceX rocket to its current longitude

8:44 Speaker 2: at ninety one degrees west on April seventh, twenty twenty three.

8:48 Speaker 2: Tempo's first light or Earth observations were done on August second,

8:53 Speaker 2: twenty twenty three, and we had a law fairly six

8:57 Speaker 2: month about commissioning phase, which was successful and then led

9:01 Speaker 2: into our nominal operations phase of the mission which we're

9:04 Speaker 2: currently in, and that occurred on October nineteenth, twenty twenty three,

9:09 Speaker 2: and we are currently Our mission is currently funded, funded

9:14 Speaker 2: and extended through September twenty twenty six. Of course, TEMPO

9:17 Speaker 2: can be up there ten plus years depending on funding

9:21 Speaker 2: down the road. So here's a nice picture too of

9:25 Speaker 2: the SpaceX launch the Tempo launch on the SpaceX rocket

9:30 Speaker 2: on April seventh, twenty twenty three. So during the pre

9:36 Speaker 2: launch phase of the mission, we definitely did discuss the

9:42 Speaker 2: apple core applications of TEMPO with our early adopter community

9:45 Speaker 2: of the mission, and this was a diagram that the

9:49 Speaker 2: science and the early adopter community kind of built during

9:52 Speaker 2: the pre launch phase of the mission. You can see

9:54 Speaker 2: that the core applications definitely center around air quality, air

9:59 Speaker 2: qual modeling and forecasting air pollution emissions and monitoring.

10:03 Speaker 3: We have a large.

10:05 Speaker 2: Number of stakeholders in the air quality management community and

10:08 Speaker 2: also a pretty long list of early adopters involved in

10:12 Speaker 2: public health applications of TEMPO. Additional application areas include vegetation

10:17 Speaker 2: and ocean monitoring and weather analysis and forecasting. All Right,

10:23 Speaker 2: this slide here is showing the tempo's footprint size across

10:27 Speaker 2: the Temple field guard. The top figure here is showing

10:31 Speaker 2: the east to west pixel size and kilometers and the

10:35 Speaker 2: north south pixel size in kilometers. And as we're seeing here,

10:40 Speaker 2: we have very fine tempo footprints or high spatial resolution

10:45 Speaker 2: from Tempo across much of the field of our guard.

10:48 Speaker 2: We lose some of that information content as we go

10:50 Speaker 2: farther north in the field of our guard, especially in

10:53 Speaker 2: the northwest and northeast corners, but altogether very high spatial resolution.

10:59 Speaker 2: This table to the right shows different locations across the

11:02 Speaker 2: field regard and the Tempo footprint size tied to those locations.

11:08 Speaker 2: So key take home message here is again when you

11:11 Speaker 2: go further north in the Temple field reguard, like Juneo,

11:14 Speaker 2: Alaska Canadian oil sands, you have larger footprint sizes. If

11:19 Speaker 2: you go farther south, for example, over in Mexico City,

11:22 Speaker 2: we have fine scale pixel sizes. So there are some

11:25 Speaker 2: variations throughout the TEMPO field re GUARD. And here's a

11:29 Speaker 2: map zoomed in showing the TEMPO footprints in by these

11:33 Speaker 2: red boxes here red polygons, and showing the spatial coverage

11:39 Speaker 2: of TEMPO across the LA area. So you can see

11:42 Speaker 2: we're getting a lot of information content spatial information on

11:46 Speaker 2: air pollutants across the LA area by having this type

11:50 Speaker 2: of spatial resolution or footprint size across these urban areas.

11:55 Speaker 2: Now zoom into New York City or the New York

11:58 Speaker 2: broader New York area and again getting a lot of

12:01 Speaker 2: different Temple footprints that can retrieve information on air pollutants

12:05 Speaker 2: across the New York area and showing how Temple can

12:08 Speaker 2: observe the urban interurban air pollutant gradients. And same picture

12:15 Speaker 2: here over a similar picture here over in Mexico City.

12:18 Speaker 2: Of course, we have even higher spatial resolution across Mexico City.

12:23 Speaker 2: So we're really zoomed in here and showing the large

12:26 Speaker 2: number of Temple footprints that we're able to that we

12:30 Speaker 2: have over this region of interest to the south of

12:32 Speaker 2: the field for GUARD.

12:35 Speaker 3: All right, now we're.

12:36 Speaker 2: Going to dig into some actually showcase some TEMPO data

12:41 Speaker 2: here and look at our how temple performed it scans

12:45 Speaker 2: across the field for guard. This is an example on

12:50 Speaker 2: July twenty six, twenty four, where we had our nominal

12:54 Speaker 2: operations performing, and this is Tempo tropospheric. You know two

12:59 Speaker 2: data shown in this animation where we're seeing how these

13:02 Speaker 2: nominal operations consist of both these standard hourly daytime scans

13:07 Speaker 2: where we're gathering data across an entire field reguard and

13:11 Speaker 2: also these optimized shorter scans that we're seeing here later

13:15 Speaker 2: in the afternoon across the west. So we see the

13:18 Speaker 2: same picture here across the east in the morning where

13:22 Speaker 2: we have only daylight in that region, so we're performing

13:25 Speaker 2: the tempo operations in that region the optimized scans. Then

13:28 Speaker 2: we moved to these standard operations and then back to

13:31 Speaker 2: the optimized operations in the west.

13:34 Speaker 3: In the afternoon.

13:35 Speaker 2: So that's kind of the day in the life of

13:37 Speaker 2: Tempo in terms of making us observations.

13:41 Speaker 3: And we also had.

13:42 Speaker 2: A special component of Tempo where we can zoom in

13:47 Speaker 2: to areas of interest and perform Tempo special operations. So

13:53 Speaker 2: this was a four day time period from January sixteenth

13:57 Speaker 2: and nineteenth, twenty twenty five where we reformed these very

14:01 Speaker 2: high resolution time stands across the west coast of the

14:05 Speaker 2: US and then went back and forth between our nominal

14:09 Speaker 2: hourly operations and our special operations throughout the day, and

14:14 Speaker 2: this for a four day time period, so this is

14:16 Speaker 2: another really unique component of tempo as well. This was

14:20 Speaker 2: done right after the large La wildfires that occurred in

14:24 Speaker 2: January twenty twenty five, So if we zoom in to

14:29 Speaker 2: that La Basin area and this is kind of when

14:33 Speaker 2: the wild wildfires kind of died down, but they're still

14:35 Speaker 2: smoldering from the wildfires. And this animation to the left

14:41 Speaker 2: is showing though the propostrict two that was collected during

14:46 Speaker 2: those special operations and the high variability in two columns

14:50 Speaker 2: that were preserved across the La Basin region. And when

14:54 Speaker 2: taking a look at the time series information here for

14:57 Speaker 2: two days from January through seventeenth and taking a location

15:03 Speaker 2: of interest within the location of the Palisades wildfire, we're

15:07 Speaker 2: seeing how during this moldering time period of the wildfire event,

15:12 Speaker 2: we're capturing this two diurnal variability which is not quite

15:16 Speaker 2: captured by the standard normal observations shown here in black.

15:21 Speaker 2: So these the blue line here blue dots are showing

15:25 Speaker 2: the addition of the special operations over the region compared

15:29 Speaker 2: to the normal standard observations. Now down here to the

15:34 Speaker 2: bottom bright we're showing an area in downtown LA where

15:38 Speaker 2: again we observe the stronger variability and two variability in

15:43 Speaker 2: downtown Los Angeles during the morning of January seventeenth, which

15:48 Speaker 2: was not captured by our Tempo standard scans. So again

15:52 Speaker 2: the addition of the special scans helped collect more information

15:56 Speaker 2: on two. And you can actually go to our Tempo

16:01 Speaker 2: operations log through this QR code here and find the

16:05 Speaker 2: day is when these special operations occurred. And here's a

16:09 Speaker 2: screen capture of part of that operations log, and where

16:14 Speaker 2: the special operations that were done for the LA region

16:18 Speaker 2: from January sixteenth through nineteenth we're done. They can kind

16:21 Speaker 2: of find that in the operations log along with the

16:25 Speaker 2: other special operations that we've done to date. All Right,

16:31 Speaker 2: So now we're going to discuss a little more detail

16:33 Speaker 2: on the Tempo trace gas products in the following slides,

16:39 Speaker 2: and first showcase how Tempo observes these trace gas plutants

16:43 Speaker 2: that we provide from the mission. And the key variable

16:47 Speaker 2: of the key input here is the altra violet to

16:49 Speaker 2: visible radiation from the sun. TEMPO requires sunlight to make

16:54 Speaker 2: observations of these trace gas plllutants and which then we

17:00 Speaker 2: observe these top of atmosphere radiance shown here in this

17:04 Speaker 2: diagram and along with our retrieved slant calmn density information,

17:10 Speaker 2: and we combine that with sun and satellite viewing condition

17:13 Speaker 2: information to report or retrieve information on vertical calm densities,

17:19 Speaker 2: which really the VCDs here provide information on the target

17:24 Speaker 2: trace gas of interest within a standard verdized vertical column

17:29 Speaker 2: of the atmosphere. So we're saying this is our key

17:32 Speaker 2: output and our key parameter that we provide in our

17:36 Speaker 2: tempo trace gas products.

17:40 Speaker 3: And how we do.

17:40 Speaker 2: That going from our slant column to our vertical column

17:44 Speaker 2: information or data in the tempo retrieval process is done

17:49 Speaker 2: by first driving our slant column densities from our measured

17:55 Speaker 2: top of atmosphere radiances that were shown in that previous diagram,

18:00 Speaker 2: and then we calculate vertical calm densities using air mass factors.

18:05 Speaker 2: And these air mass factors are calculated offline using a

18:10 Speaker 2: ray of transfer model and input from a global three

18:14 Speaker 2: D atmospheric composition model and Tempo in particular uses the

18:18 Speaker 2: Gotter Earth Observing System composition forecasting system to put.

18:23 Speaker 3: Together these air mass factors.

18:25 Speaker 2: And this equation shown to the left is showing the

18:29 Speaker 2: equation where we have slant colmn densities in the numerator

18:33 Speaker 2: and this air mass factor variable which actually combines a

18:38 Speaker 2: large number of information in the denominator which then derives

18:43 Speaker 2: or retrieves information on our vertical calm densities. And shown

18:47 Speaker 2: at the bottom of the slide here are the in

18:50 Speaker 2: O two slant colmn densities zoomed into the Pacific northwest

18:55 Speaker 2: here and our INNO two air mass factors in the

19:00 Speaker 2: middle and our total n O two Erkele Colm densities

19:03 Speaker 2: on the right. The key message here is that you

19:06 Speaker 2: can see how our slant colmn densities from off in

19:10 Speaker 2: O two are higher than our total n O two VCDs,

19:14 Speaker 2: and that shows how this process from going from our

19:17 Speaker 2: slant Colm density to our verical calm density leads to

19:21 Speaker 2: lower VCDs. As we're looking at the vircal column directly

19:24 Speaker 2: above the temple footprint compared to the the slant column

19:29 Speaker 2: information that's shown here. And we also go a bit

19:33 Speaker 2: further with our temple retrieval to derive propospheric n O

19:38 Speaker 2: two VCDs. So here is a comparison between our tropospheric

19:43 Speaker 2: n O two rcle calm densities and our total n

19:47 Speaker 2: O two And as expected, our tripospheric n O two

19:51 Speaker 2: VCDs are lower than our total column in O two

19:54 Speaker 2: the VCDs. And there was a wildfire that occurred during

19:57 Speaker 2: this time period, so it's large and O two plume

20:00 Speaker 2: in this map is tied to a large wildfire smoke

20:03 Speaker 2: event that occurred in North California. Now we go a

20:07 Speaker 2: bit to do this whole operation where we derive tropospheric

20:11 Speaker 2: n O twovcds. We apply a stratosphere troposphere separation technique

20:17 Speaker 2: in our temple retrieval process, and this method estimates the

20:21 Speaker 2: stratospheric n O two BCDs based on the slant colmn densities,

20:26 Speaker 2: the air mass factors, and then O two columns from

20:30 Speaker 2: the GEOCF forecast. The stratospheric n O two BCDs are

20:35 Speaker 2: then subtracted from the total column in two VCDs to

20:39 Speaker 2: arrive at our tropospheric n O two. So, and note

20:43 Speaker 2: the different scale on these figures here but to the

20:46 Speaker 2: left here again I'm sharing the total N O two

20:49 Speaker 2: BCD map. Here's our stratospheric O two BCDs can lower

20:55 Speaker 2: scale here much lower two in the stratosphere in general.

21:00 Speaker 2: And then to the right, I'm showing the tropospheric you

21:03 Speaker 2: know to BCD, so you can see here how we

21:07 Speaker 2: have these lower tropospheric colum amounts compared to her total,

21:11 Speaker 2: and how we go from our total to our tropospheric.

21:14 Speaker 2: And there's also quality assurance variables within the data files

21:18 Speaker 2: of TEMPO, which I'll go over here in a minute

21:20 Speaker 2: to where we can filter lower.

21:22 Speaker 3: Quality or higher uncertain data.

21:24 Speaker 2: And here is where I applied a filter where we

21:28 Speaker 2: only kept data with the cloud fraction less than fifty

21:31 Speaker 2: percent and a Slursian thing go less than eighty degrees,

21:34 Speaker 2: and going back and forth, you can see how we

21:37 Speaker 2: end up having these areas where clouds are removed, but

21:41 Speaker 2: also areas where even the smoke plume were removed, which

21:44 Speaker 2: I'll go into a bit more detail of another use

21:47 Speaker 2: case of when that happens. All right, So for the tip,

21:52 Speaker 2: this is a table showing the TEMPO data products the baseline,

21:57 Speaker 2: the near real time and also the upcome NOAH aerosol products.

22:01 Speaker 2: And we start with the level one data, the level

22:05 Speaker 2: one radiance data which is at this tempo footprint size

22:09 Speaker 2: of two by four point seventy five kilometer squared, which

22:12 Speaker 2: is the tempo footprint size at the center of the

22:15 Speaker 2: field regard. But as we saw earlier that footprint size

22:19 Speaker 2: varies throughout the tempo field regard. Then we drive or

22:23 Speaker 2: retrieve these various products shown here, including cloud products, ozone

22:28 Speaker 2: total column, nitrogen dioxide, fermadehyde, and ozone profile. Some of

22:36 Speaker 2: these are also near real time products, so we provide

22:39 Speaker 2: we are currently providing near real time products with data

22:41 Speaker 2: latency of less than two hours, and that includes our

22:46 Speaker 2: cloud nitrogen dioxide and FRAMATOHYDE data. And a key point

22:52 Speaker 2: with our ozone profile is the ozone profile provides information

22:56 Speaker 2: on triposphereic ozone column amount, so we get at information

23:01 Speaker 2: in the troposphere and lower troposphere, and we can actually

23:04 Speaker 2: track ozone concentrations really well in the troposphere through the

23:08 Speaker 2: ozone profile. Upcoming products in the near future are going

23:12 Speaker 2: to be our aerosol products including aerosol optical depth and

23:16 Speaker 2: aerosol layer height, and an aerosol detection product. Those level

23:21 Speaker 2: two products shown here. The baseline products also have a

23:25 Speaker 2: Level three product tied to them, which are pretty much

23:30 Speaker 2: very similar to our level two except they are put

23:33 Speaker 2: on a regular standard point zero two degree grid across

23:37 Speaker 2: the Tempo field regard, so those those grid point sizes

23:41 Speaker 2: don't change like the footprint sizes do.

23:43 Speaker 3: Up here.

23:44 Speaker 2: I do want to note that the Ozone profile Level

23:47 Speaker 2: three is a big courser at point zero four degrees

23:50 Speaker 2: compared to the other Level three products, and the reason

23:54 Speaker 2: for that is, as you can see here the resolution

23:58 Speaker 2: of our Ozone profile for level two courser because there's

24:01 Speaker 2: some co adding or averaging done of pixels to make

24:04 Speaker 2: that product of higher accuracy precision for the community, which

24:09 Speaker 2: is required. And following we have level four products which

24:13 Speaker 2: will include our surface PM two point five product with

24:16 Speaker 2: an hourly PM to point five estimation. Okay, now we're

24:21 Speaker 2: going to do a little more discussion on level two

24:24 Speaker 2: versus Level three products from Tempo. Picking off here is

24:30 Speaker 2: the general methodology where we go from our tempo level

24:34 Speaker 2: two products to a level three and the level three

24:38 Speaker 2: products are essentially created from re gridding level two data

24:44 Speaker 2: onto a regular grid using an area weighted averaging approach.

24:49 Speaker 2: So we do that at the point zero two degree

24:51 Speaker 2: grid for n O two from Autohyde and total ozone

24:55 Speaker 2: and at that point zero four degree grid for ozone profile.

25:00 Speaker 3: The equation that.

25:02 Speaker 2: Is used here is shown on the slide, and it's

25:05 Speaker 2: really just geomemetric regridding where we take contributions from nearby

25:11 Speaker 2: grid points that overlap that level two pixel a footprint

25:16 Speaker 2: and then apply this equation to get at our level

25:19 Speaker 2: three data for each grid point. And here's an example

25:24 Speaker 2: of the level two verst level three data zoomed into

25:28 Speaker 2: the Elliot Basin area on June eighteenth, twenty twenty five,

25:33 Speaker 2: zoomed in on Eli Basin and showing in the comparison

25:36 Speaker 2: between level two and level three. So we're seeing those

25:40 Speaker 2: kind of polygon shapes when we look at our level

25:42 Speaker 2: two which was shown on that map earlier, the Temple footprints,

25:46 Speaker 2: and now we have this like standard regular grid that

25:49 Speaker 2: we're seeing this Temple data appear in the level three product.

25:53 Speaker 2: So this is in the morning hours where we have

25:56 Speaker 2: High two likely tied to the traffic in the morning

26:00 Speaker 2: across the La Basin area.

26:04 Speaker 3: Before our level two data.

26:07 Speaker 2: We provide these in different data granules across the tempo

26:12 Speaker 2: field regard. There's nine different Level two data granules that

26:16 Speaker 2: compose each full tempo field regard scan during our hourly

26:22 Speaker 2: standard operations. So we're seeing us for one, this is

26:25 Speaker 2: simply an animation showing one hour of those nine different

26:30 Speaker 2: six of sevenment data granules that are provided across the

26:34 Speaker 2: temple field regard and notice did change before starting September

26:39 Speaker 2: twenty ninth, twenty twenty three, we went from ten to nine,

26:42 Speaker 2: so during the commissioning phase we had a little larger

26:45 Speaker 2: number before September twenty ninth, all right, So then we

26:51 Speaker 2: take those nine different tempo data granules for each full

26:56 Speaker 2: tempo field regard scan and compose a level three rescan

27:00 Speaker 2: data file. This is similar to the previous or same

27:04 Speaker 2: time frame as a previous animation, but just taking all

27:08 Speaker 2: those level two data granules, stitching them together and doing

27:13 Speaker 2: that remapping regrinding process to put that two data on

27:18 Speaker 2: a point zero two degree grid across a temple field reguard.

27:23 Speaker 2: This is how those level three scan data files are

27:26 Speaker 2: provided for tempo and zooming in Here to the LA

27:32 Speaker 2: basin again looking at the level two verst level three

27:37 Speaker 2: in two comparison for a morning, early afternoon and late

27:42 Speaker 2: afternoon evening time frame during this one day on June eighteenth,

27:47 Speaker 2: twenty twenty five, and we're seeing similar information as before

27:51 Speaker 2: in terms of you can really see the polygon shapes

27:55 Speaker 2: where the two column data are shown and then kind

27:58 Speaker 2: of the regular grid shown here. Uh. Kind of one

28:02 Speaker 2: key take home here is in general level three O

28:05 Speaker 2: two maps or data will have a smoother appearance compared

28:09 Speaker 2: to the level two n O two maps after doing

28:12 Speaker 2: that re guitting averaging process, and this can also lead

28:16 Speaker 2: to some significant differences in the like fine scale in

28:22 Speaker 2: O two hot spots and large in O two gradients

28:24 Speaker 2: when you're comparing the level two in level three data

28:28 Speaker 2: and maps. This is the same kind of comparison here

28:34 Speaker 2: except for for modehyde, so we're going from level two

28:37 Speaker 2: to level three for modehide comparison, Moldehyde is a noisier,

28:42 Speaker 2: more difficult retrieval compared to n O two, So in

28:46 Speaker 2: general that just makes this comparison a bit harder in

28:49 Speaker 2: terms of seeing where all these gradients or differences are occurring.

28:53 Speaker 2: But again similar thing here where we have a smoother

28:56 Speaker 2: appearance to our level three from moldehyde map due to

28:59 Speaker 2: that remap being recruiting process compared to our level two.

29:06 Speaker 3: All right, so.

29:06 Speaker 2: Now we want to talk a little bit more in

29:08 Speaker 2: detail about the quality assurance and data filtering methods which

29:14 Speaker 2: are really needed by the user community. And our recommendations

29:19 Speaker 2: here are based on the latest recommendations for the version

29:22 Speaker 2: four product release. It's an important to note here that

29:27 Speaker 2: we had a we went from version three to version

29:29 Speaker 2: four data. Version four data started being produced on September seventeenth,

29:34 Speaker 2: twenty twenty five. We haven't reprocessed the data prior to

29:39 Speaker 2: September seventeenth, twenty twenty five yet, but for now all

29:42 Speaker 2: our version four data are available starting September seventeenth, twenty

29:47 Speaker 2: twenty five. So these recommendations here are focused on that

29:51 Speaker 2: version four data and this table is showing the key

29:55 Speaker 2: quality assurance parameters that are recommended for use and this

30:01 Speaker 2: includes effective cloud fraction, solar ZMP angle and our main

30:05 Speaker 2: data quality flag.

30:08 Speaker 3: This quality flag is a bit.

30:10 Speaker 2: More complicated in terms of how you know what these

30:14 Speaker 2: flags refer to in the product in the variable and

30:18 Speaker 2: for example, zero of flag of zero means high quality

30:22 Speaker 2: temple retrieval.

30:24 Speaker 3: A flag of one.

30:27 Speaker 2: Means the suspect retrieval due to air mass factor or

30:30 Speaker 2: viewing geometry issues, and a flag of two means outlier

30:35 Speaker 2: retrieval or there is no successful air mass factor calculation.

30:40 Speaker 2: So for looking at the different ranges of these variables,

30:43 Speaker 2: we have the effective cloud fraction ranging from zero to one,

30:47 Speaker 2: solar zap angle as you'd expect from zero to ninety degrees,

30:51 Speaker 2: and our main data quality flag going from zero to

30:53 Speaker 2: two currently in Worldview for all the available Version three

30:58 Speaker 2: data that are available to view. These are the qualitative

31:02 Speaker 2: use thresholds that are applied, so they apply a cloud

31:06 Speaker 2: fraction lesson point five sols, the youth angle less than eighty,

31:10 Speaker 2: and a mandated quality flag less than two or quantitative

31:14 Speaker 2: use and research use. There are different recommendations, and our

31:18 Speaker 2: Tempo user guides that they recommend a much lower cloud

31:22 Speaker 2: fraction threshold of less than point one to retain high

31:27 Speaker 2: quality data that can kind of vary, you know, point

31:30 Speaker 2: two point one five depending on your application. Of course,

31:34 Speaker 2: solar nf angle is less than seven degrees for quantitative

31:38 Speaker 2: use and they recommend only retaining the high quality tempo

31:43 Speaker 2: retrievals or quantitative.

31:45 Speaker 3: Research use as well.

31:48 Speaker 2: All right, so this is an example of comparing those

31:52 Speaker 2: quality a currance thresholds and the different maps that can

31:57 Speaker 2: be shown based on those different assurance thresholds. So this

32:00 Speaker 2: is for one day on September twentieth, twenty twenty five,

32:05 Speaker 2: zoomed into the Pacific northwest here of tempo troposphrict n

32:09 Speaker 2: O two, and we're seeing here at the bottom are

32:13 Speaker 2: our most stricter quality assurance recommendations shown on the previous table.

32:18 Speaker 2: These are the thresholds at the top here are those

32:20 Speaker 2: currently used for the worldview in two visualizations. And the

32:25 Speaker 2: key message here is that there were some cloud cover

32:29 Speaker 2: during this time period and we're seeing more white masked

32:33 Speaker 2: out areas in the lower panels here due to cloud cover.

32:39 Speaker 2: But also we're seeing a wildfire event that occurred during

32:42 Speaker 2: this time period, so there is a pretty large wildfire

32:45 Speaker 2: and we're seeing very high n O two plume tied

32:49 Speaker 2: to that smoke. But when we apply that cloud fraction,

32:53 Speaker 2: the strict cloud fraction quality assurance threshold, we remove that

32:57 Speaker 2: smoke plume, the n O two plume from the map

33:00 Speaker 2: and analysis. So you know, carriit needs we take in

33:03 Speaker 2: when applying these cloud fraction thresholds based on your application

33:08 Speaker 2: and needs and looking at the similar thing here. But

33:13 Speaker 2: looking at solar zeth angle quality assurance, and this is

33:18 Speaker 2: using the strict cloud fraction of ten percent, but then

33:22 Speaker 2: looking at the solar z youngth angle of eighty degrees

33:25 Speaker 2: versus seventy degrees.

33:26 Speaker 3: So when we apply that.

33:29 Speaker 2: Stricter solarcanth angle threshold during the morning and late afternoon

33:35 Speaker 2: here we can see that we lose information content on

33:38 Speaker 2: two based on that updated or strict solars youngth angle.

33:44 Speaker 2: In these maps here and to the right, we're looking

33:47 Speaker 2: at the whole entire temple field reguard here applying the

33:51 Speaker 2: less strict cloud fraction. But the main change here was

33:54 Speaker 2: looking at a flag of equal to zero versus a

33:57 Speaker 2: flag of less than two. And by using a flag

34:02 Speaker 2: equal to zero we lose this triposphere you know two

34:05 Speaker 2: information in the far northwest and far northeast due to

34:09 Speaker 2: the viewing angle restriction flag within the product. So it's

34:14 Speaker 2: another thing to keep in mind when using the flag

34:17 Speaker 2: variables within the tempo n O two and from auto

34:19 Speaker 2: hyde data. And I wanted to share another table here

34:24 Speaker 2: for the quality assurance of recommendations for tempo total calm

34:28 Speaker 2: ozone slightly different but similar good degree we have quality

34:34 Speaker 2: flag parameters, effective cloud fraction, solar xenath angle viewing zenith

34:39 Speaker 2: angle recommendations. There's a lot more information with contained within

34:45 Speaker 2: the quality flag here compared to two and from Adehyde,

34:48 Speaker 2: including sunglint contamination and so two being present for you

34:54 Speaker 2: can kind of see the different ranges, much higher range

34:57 Speaker 2: with this quality flag or quality use. They recommend less

35:02 Speaker 2: than ten twenty four for the quality flag and for

35:06 Speaker 2: quantitative use for total column ozone. For best highest quality

35:10 Speaker 2: data they simply apply the flag of equal zero and

35:15 Speaker 2: if you want to retain more adequate data, you can

35:19 Speaker 2: also keep those flags equal to one, two and five

35:23 Speaker 2: effective cloud fraction. There's no filter currently being used for

35:27 Speaker 2: qualitative use in Worldview. However, for quantitative use and research

35:32 Speaker 2: they recommend a uneffective cloud fraction of less than point five.

35:37 Speaker 2: And you're seeing similar solar interviewing zenith angle quantitative use

35:41 Speaker 2: restrictions for our total ozone column as shown for.

35:45 Speaker 3: Two and from all hide.

35:48 Speaker 2: And now we're comparing the different quality assurance threshold shown

35:53 Speaker 2: here same day on September twentieth, twenty twenty five, as

35:58 Speaker 2: shown for two, but this is for total calm ozone

36:01 Speaker 2: and applying the qualitative thresholds that were shown in the

36:06 Speaker 2: previous slide and the more quantitative research thresholds, so as

36:11 Speaker 2: we would expect when applying a cloud for action. On

36:15 Speaker 2: the bottom panels, here we're seeing more areas removed due

36:19 Speaker 2: to cloud. Also there are there is some influence from

36:23 Speaker 2: applying that quality flag equal to zero as well. Some

36:27 Speaker 2: of these areas around here were also removed due to

36:29 Speaker 2: that quality flag. So again taking account kind of what

36:34 Speaker 2: happens going from our different qualitative to quantitative thresholding for

36:39 Speaker 2: this product. All right, so now we're zooming out and

36:43 Speaker 2: look at the tempo field reguard for total calumn ozone

36:47 Speaker 2: and the top here we are looking at the quantitative

36:51 Speaker 2: more research thresholds versus the qualitative thresholds that were shown

36:55 Speaker 2: on the previous slide. And as you can see, when

36:58 Speaker 2: we apply that stricter cloud fraction, we remove the clouds

37:03 Speaker 2: from the scene. But also there's some influence from that

37:06 Speaker 2: quality flag equal to zero. You can especially to see

37:10 Speaker 2: that quality flag equal zero and the sunlin contermination removed

37:15 Speaker 2: from the far south extent here. This is sun glint

37:19 Speaker 2: that's removed when we apply that quality flag equal to

37:22 Speaker 2: zero versus retaining that sun glint area and the bottom map,

37:27 Speaker 2: and there's also some influence from applying that quality flag

37:32 Speaker 2: in other areas of the field reguard as well. Okay,

37:37 Speaker 2: we're going to actually show some examples of our level

37:40 Speaker 2: to trace gas products and for the example, we're going

37:46 Speaker 2: to be focused on that version for data time period.

37:50 Speaker 2: This is zoomed into the South US US on October fifteenth,

37:56 Speaker 2: twenty twenty five. To the left here where's showing a

38:00 Speaker 2: beer's true color image and on the right showing the

38:05 Speaker 2: animation of tempo troposphere No. Two erical column densities on

38:10 Speaker 2: this one day from morning to evening across the domain,

38:15 Speaker 2: and what we're seeing here are high n O two

38:20 Speaker 2: BCD columns over Dallas Houston. We're seeing some elevated n

38:26 Speaker 2: O two across the major traffic corridors and over the

38:30 Speaker 2: premium basin as well. So there's a large oil and

38:33 Speaker 2: gas infrastructure and across the premium basin and we're seeing

38:37 Speaker 2: higher some higher n O two columns tied to that

38:40 Speaker 2: premium basin, along with some small hot spots of n

38:44 Speaker 2: O two tied to power plant locations, and also some

38:49 Speaker 2: fires that were occurring across the domain. Across east here especially,

38:54 Speaker 2: we had some fires that were recurring across Louisiana, for example,

38:58 Speaker 2: and some of these two hot spots were tied to

39:00 Speaker 2: those spires. And then this is another animation for total

39:05 Speaker 2: calm from adehyde the same day, October fifteenth, twenty twenty

39:09 Speaker 2: five morning to evening from adehyde information across the South ust.

39:16 Speaker 3: And one of the areas of higher from.

39:19 Speaker 2: Moto hyde concentrations if you look at the animation closely

39:23 Speaker 2: occurs over the Houston area, So that is a typical

39:27 Speaker 2: urban area tied to high from Otahyde concentration. So we

39:31 Speaker 2: see that in our from Autohyde animation here, along with

39:34 Speaker 2: other areas of elevated from ata hyde even over the

39:37 Speaker 2: Permium basin tied to the O two in that region

39:41 Speaker 2: and the oil and gas industry. And this is animation

39:47 Speaker 2: for total calm ozone for the same day. And again

39:53 Speaker 2: it's this total calm, So this takes an account built

39:55 Speaker 2: the proposphere and the stratosphere over you know, ninety percent

40:00 Speaker 2: on average of the ozone is contained within the stratosphere.

40:05 Speaker 2: That being said, we're still seeing some variations in total

40:09 Speaker 2: calum ozone in this animation with some high higher concentrations

40:15 Speaker 2: across the various regions, including in around the Houston area.

40:20 Speaker 2: Of course, now we have our tempo ozone profile data

40:25 Speaker 2: which is shown here. This is the tempo tropospheric ozone

40:30 Speaker 2: column data for that same day, and now we're seeing

40:34 Speaker 2: some stronger, more apparent distinct variations in our tropospheric ozone

40:41 Speaker 2: calum amounts as we retrieve this information only in the troposphere,

40:45 Speaker 2: and I'll highlight again one of the areas of interest

40:48 Speaker 2: over the Houston area as these higher ozone concentrations that

40:54 Speaker 2: occur in the later afternoon area during that animation, we

40:58 Speaker 2: can even get down to the lower troposphere with their

41:02 Speaker 2: tempo ozone profile product. So same day, but looking at

41:07 Speaker 2: the zero two kilometer ozone column information, and this is

41:12 Speaker 2: really neat what we see here. We'll replay the animation

41:16 Speaker 2: again from morning to evening and as you will see

41:20 Speaker 2: in a later afternoon evening hour, we see a definitely

41:25 Speaker 2: distinct high values that are occurring in and around the

41:29 Speaker 2: Houston area, mainly to the southwest where we actually where

41:34 Speaker 2: the ground monitors observed higher ozone columns, higher ozone concentrations

41:40 Speaker 2: on this day as well.

41:41 Speaker 3: So really neat that we're.

41:42 Speaker 2: Picking out some of the elevated lower tropospheric ozone in

41:46 Speaker 2: the area where the ground monitors measured higher ozone concentrations

41:50 Speaker 2: and unhealthy ozone levels.

41:55 Speaker 3: So we're changing day.

41:56 Speaker 2: We're changing days here to September twenty five because we

42:02 Speaker 2: had a large wildfire event occur on this day. So

42:06 Speaker 2: again to the left, I'm sharing the veer's true color

42:10 Speaker 2: image here on the in the Pacific Northwest, and we're

42:14 Speaker 2: seeing these large grayish color intensities here tied to wildfire

42:20 Speaker 2: smoke on this day. And to right, we're looking at

42:24 Speaker 2: the Tempo ultra violet aerosol index and this variable is

42:28 Speaker 2: contained within the tempo futal ozone product. So it's really

42:33 Speaker 2: neat about this as we see these higher UV aerosol

42:37 Speaker 2: index values tied to these wildfire smoke plumes, and that's

42:42 Speaker 2: due to the absorbing the absorbing nature of smoke particles

42:47 Speaker 2: which leads to the high UV aerosol index value. So

42:52 Speaker 2: this parameter can be very useful in terms of monitoring

42:55 Speaker 2: and tracking smoke plumes throughout the day. Okay, So I

43:03 Speaker 2: don't want to share some comparisons between TEMPO and our

43:08 Speaker 2: current low earth observing satellite instruments, the NASA Ozone Monitoring

43:14 Speaker 2: Instrument and TRIPOMI. All these results here are based on

43:18 Speaker 2: our version three Tempo products. So this is the comparison

43:25 Speaker 2: between our low earth orbiting Ozone Monitoring Instrument versus Tempo. Again,

43:33 Speaker 2: OMI is a low worth orbiting so it only provides

43:35 Speaker 2: the midday observations of air pllutants across the globe. And

43:39 Speaker 2: it's spatial resolution it is much coarser at thirteen by

43:43 Speaker 2: twenty four kilometers squared compared to what we have with Tempo.

43:48 Speaker 3: And this is showing just as zoomed.

43:50 Speaker 2: Out you know, over the west region of the US

43:54 Speaker 2: OMI triposophic two versus Tempo, and it is you can

43:59 Speaker 2: see the really incredible information content now we get from

44:03 Speaker 2: Tempo n O two compared to TRIPOM. And if we

44:07 Speaker 2: zoom in look at the zoomed in images over the

44:10 Speaker 2: La basin area versus Tempo versus OMI, we can see

44:16 Speaker 2: that Tempo in O two tropospheric vertical column densities are

44:19 Speaker 2: larger than OMI in urban areas and this shows over

44:24 Speaker 2: the LA area over MEXICALI two. And this is a

44:30 Speaker 2: combination of both the much higher spatial resolution of Tempo

44:34 Speaker 2: that's able to capture these areas are high n O two,

44:38 Speaker 2: but also the higher instrument sensitivity of Tempo as well,

44:41 Speaker 2: So this is really incredible. Now we're seeing with Tempo

44:45 Speaker 2: compared to our previous legacy instruments like OMI. Okay, now

44:50 Speaker 2: we're looking at comparison between Tempo and Tropomy. So Tropomy,

44:57 Speaker 2: like OMI, is a low worth urbaning satellite instrument, was

45:01 Speaker 2: launched in twenty seventeen and provides higher spatial resolution.

45:06 Speaker 3: Information than OMY.

45:07 Speaker 2: So now we're TROPOMI is providing two observations at five

45:12 Speaker 2: point five by three point five kilometers squared. So we

45:15 Speaker 2: can definitely see right away that the comparison between Tropomy

45:19 Speaker 2: and OMI is much closer than OMI and Tempo number one.

45:24 Speaker 2: There are differences for sure tied to the different instrument

45:28 Speaker 2: retrieval techniques, but also the spatial resolution as well. So

45:34 Speaker 2: if we can kind of see that if we zoom

45:36 Speaker 2: in closer to the Tempo versus TRIPOMI comparison across the

45:41 Speaker 2: La Basin area here where we are seeing in general

45:46 Speaker 2: higher two tropospheret N two from Tempo across the La Basin. Again,

45:52 Speaker 2: this likely has to do with both the different instrument

45:55 Speaker 2: retrieval techniques, but also the in general, the higher spatial

45:59 Speaker 2: resolute of tempo is contributing to these different inn O

46:04 Speaker 2: two differences. But it is important to note that we're

46:07 Speaker 2: still capturing similar in O two gradients tempo versus tropomi.

46:12 Speaker 3: There's differences in the values here.

46:16 Speaker 2: Okay, Now we're going to dig into a little more

46:18 Speaker 2: quantitative comparison between tempo and tropo ME two at these

46:22 Speaker 2: selected areas across the field. Regard shown here in the

46:26 Speaker 2: map in the top right you take home points here

46:31 Speaker 2: is that tempo is higher than tropo ME at these

46:34 Speaker 2: locations of interest. In general, tempo tends to be a

46:38 Speaker 2: little bit lower than tripo ME in non attainment areas.

46:42 Speaker 2: Tropomi tends to have a high bias in these low

46:45 Speaker 2: in two conditions. And if we do more of a

46:48 Speaker 2: detailed comparison between tempo, tropospheric O two and tripomi at

46:55 Speaker 2: the Boulder and the LA South Coast locations here, we're

46:59 Speaker 2: seeing that we're getting better correlations between tempo and tripomi

47:03 Speaker 2: across the Boulder area compared to the LA South Coast area.

47:07 Speaker 2: So we grapping correlations of point seventy three in Boulder

47:12 Speaker 2: in correlations of point three six across the LA South coast.

47:15 Speaker 2: So the main take home here is that Tempo and

47:18 Speaker 2: Tropomei they generally are an agreement, but they can also

47:23 Speaker 2: be their comparisons can vary at different locations across the

47:27 Speaker 2: field for guard as well. Okay, so now we're gonna

47:32 Speaker 2: do a little more in detail with the validation of

47:35 Speaker 2: Tempo products. This is again a validation comparison done for

47:40 Speaker 2: our version three Tempo products, and this is a comparison

47:46 Speaker 2: Tempo validation done against our Pandora network. So our Pandora

47:52 Speaker 2: is our main ground based validation network for Tempo and

47:58 Speaker 2: Pandora provides two and from autohyde observations or measurements like Tempo,

48:05 Speaker 2: but from the ground based perspective. And here we're showing

48:08 Speaker 2: the correlations and mean biases between Tempo and two and

48:15 Speaker 2: Pandora and Tempo from Adehyde and Pandora from Audehyde. And

48:20 Speaker 2: the key take home notes are that Tempo N two

48:23 Speaker 2: in general reproduces the spatial variability of Pandora. It correlates

48:28 Speaker 2: well at most of the Pandora sites, and for Tempo

48:33 Speaker 2: from Adehide we have similar outcomes here with barely high

48:38 Speaker 2: correlation at the Pandora site level, there is some reasonable

48:42 Speaker 2: bias with some of the individual sites that are requiring

48:46 Speaker 2: further investigation which we're currently doing. This validation work is

48:50 Speaker 2: ongoing with our version four data as well, so more

48:55 Speaker 2: to come in terms of our Tempo validation versus Pandora

48:58 Speaker 2: for our version four data. In general, our Tempo version

49:01 Speaker 2: three data showed good comparisons between are with our tempo

49:06 Speaker 2: with our Pandora observations of n O two and from Atohyde.

49:11 Speaker 2: And if we go into looking at Tempo versus Pandora,

49:16 Speaker 2: taking account all those sites that we just showed in

49:19 Speaker 2: the previous slide and look at the per month comparisons

49:23 Speaker 2: and variability, this is showing Tempo versus Pandora in O

49:28 Speaker 2: two at the top and from Autohyde at the bottom,

49:32 Speaker 2: And for both of these comparisons and figures, we are

49:36 Speaker 2: showing that Tempo two and from Autohyde captures the seasonal

49:41 Speaker 2: variation as well with pretty good bias as well low bias.

49:45 Speaker 2: So this is showing again that Tempo is able to

49:48 Speaker 2: capture these seasonal variations as shown by Pandora, and we

49:55 Speaker 2: can get into more detail in terms of the variations

49:59 Speaker 2: per day of the day they met before. But looking

50:03 Speaker 2: at tempo versus pandora in O two at the top

50:06 Speaker 2: and from at a hyde at the bottom for different

50:10 Speaker 2: time periods throughout the day for all those sites that

50:13 Speaker 2: we saw. So the key one of the key messages

50:17 Speaker 2: here is that the n O two for tempo during

50:23 Speaker 2: the time periods in the morning in the evening with

50:26 Speaker 2: higher solar as vienap angles, we see a lower correlation,

50:31 Speaker 2: but still good correlation between tempo and pandora.

50:36 Speaker 3: This is expected.

50:37 Speaker 2: We know that our tempo retrievals have lower accuracy during

50:41 Speaker 2: the morning and the evening.

50:43 Speaker 3: It's not as apparent in the framato hyde.

50:46 Speaker 2: When you're looking at the formati hyde comparison between tempo

50:50 Speaker 2: and pandora throughout the day, our correlations are in general

50:54 Speaker 2: lower than n O two, and that's to be expected

50:57 Speaker 2: because the formato hyde retrieval it's more difficult from both

51:02 Speaker 2: the tempo and pandora perspective, so great that we're seeing

51:06 Speaker 2: these overall good correlations in general with from audehyde as well. Okay,

51:13 Speaker 2: now we're going to go into some tempo data visualization

51:16 Speaker 2: and worldview and Christina will take it over from here.

51:22 Speaker 1: Thanks Arin for a great overview of tempo. Now we're

51:25 Speaker 1: going to get into some of the data exploration that

51:28 Speaker 1: you can do with the Tempo that are available on

51:30 Speaker 1: the NASA Worldview. So we are just going to go

51:33 Speaker 1: to NASA Worldview from a normal web browser. You can

51:37 Speaker 1: type it in it's worldview dot earth data dot NASA

51:40 Speaker 1: dot gov. You can see I've typed that and it's

51:44 Speaker 1: already populated. So when you go to Worldview, this is

51:48 Speaker 1: the landing page that you'll see. Worldview has some great

51:53 Speaker 1: stories to peruse the different data sets that they have here.

51:56 Speaker 1: I encourage you to check these out at a later date.

51:58 Speaker 1: Many of them are relevant to air quality topics, but

52:02 Speaker 1: we're not doing that today. So we're just going to

52:03 Speaker 1: x out this welcome page. And I should note you

52:08 Speaker 1: are welcome to follow along with what we're doing here.

52:10 Speaker 1: But this is also a recording, so you can go back.

52:13 Speaker 1: You can follow it through later as well at your

52:16 Speaker 1: own pace, pose it and everything. Some of the homework

52:19 Speaker 1: questions are going to involve similar examples, so that will

52:21 Speaker 1: be another opportunity to go through these tools on your own.

52:26 Speaker 1: So we are going to go to explore in more detail.

52:30 Speaker 1: The wildfires of January twenty twenty five in the LA area.

52:37 Speaker 1: Note to orient you a little bit to worldview. Down

52:40 Speaker 1: here at the bottom left, we can see it has

52:42 Speaker 1: defaulted to the current date twenty twenty six, January twelfth.

52:47 Speaker 1: Since we're going to go into the LA area, I'm

52:50 Speaker 1: just going to use my cursor scroll wheel to scroll

52:53 Speaker 1: into roughly the California area for those who may not

52:57 Speaker 1: be most familiar, We're just going to put on some

52:59 Speaker 1: placeland for orientation. And because we're going to January. Oh,

53:05 Speaker 1: that's the wrong way, because we are going to January

53:08 Speaker 1: twenty twenty five. I'm just going to click this arrow down.

53:10 Speaker 1: So now it's twenty twenty six to twenty twenty five,

53:13 Speaker 1: and we are going to go to January ninth of

53:20 Speaker 1: twenty twenty five, and again for some quick orientation. Here

53:26 Speaker 1: this is the LA area. We can see that the

53:29 Speaker 1: fires were already in process. The images that show up

53:37 Speaker 1: now are true color imagers and they're represented as base layers.

53:42 Speaker 1: So it has defaulted to terror modus and that's what

53:46 Speaker 1: we're seeing here, so it has more of a morning overpass.

53:49 Speaker 1: We can turn off this layer. We can turn on

53:51 Speaker 1: the aqua modis. We can see that it had a

53:53 Speaker 1: little bit of a different view later in the day,

53:56 Speaker 1: or if we want to do a Noah twenty veers,

53:59 Speaker 1: we can see that this was a little bit more

54:02 Speaker 1: in the middle of the of the We can see

54:07 Speaker 1: it was a little bit more in the middle of

54:09 Speaker 1: the satellite swath for this particular time. But what we're

54:14 Speaker 1: interested in today is the tempo data. So we are

54:17 Speaker 1: going to go to this red button down here the

54:20 Speaker 1: plus ad layers. We're going to click on it and

54:24 Speaker 1: it brings up this page which has again a number

54:28 Speaker 1: of topics. Feel free to peruse them at your own leisure.

54:31 Speaker 1: We have some air quality topics, some specific air quality

54:34 Speaker 1: topics like fires, dust, ash, et cetera. But since we

54:38 Speaker 1: are going to tempo, what we're going to go is

54:41 Speaker 1: we're going to go to this top search bar and

54:43 Speaker 1: type in tempo and as we do that, we can

54:48 Speaker 1: see that a number of different products show up. So

54:51 Speaker 1: these are all of the products for Tempo which are

54:54 Speaker 1: currently available in Worldview. So we can click on formaldehyde

54:58 Speaker 1: and it'll give you an idea the temporal coverage, the

55:03 Speaker 1: a little bit of the metadata. We want to explore

55:08 Speaker 1: all of these, so We're just going to click the

55:09 Speaker 1: little checkbox here, and you can see behind this panel

55:14 Speaker 1: that I'm in right now, these data sets are slowly

55:23 Speaker 1: populating underneath here. You can see it's a little bit

55:29 Speaker 1: above this formaldehyde here, So if I we're to turn

55:33 Speaker 1: off the formaldehyde, we can see that the nitrogen dioxide

55:35 Speaker 1: is adding itself to the list of variables the layers

55:41 Speaker 1: that we're going to have available for this particular data.

55:45 Speaker 1: And as you can see, sometimes it can be a

55:47 Speaker 1: little bit tedious to do all of these things to

55:49 Speaker 1: check all of these boxes. So we are just going

55:53 Speaker 1: to go to a window where I've already had all

55:57 Speaker 1: of these loadeds, so we're just gonna pull the cake

56:00 Speaker 1: out of the oven. Already done. One thing again to note,

56:04 Speaker 1: so now we have all of the tempo layers loaded

56:08 Speaker 1: and they are hidden, so we will go through and

56:11 Speaker 1: look at them one at a time. So I've hidden

56:13 Speaker 1: all of those. And another thing to note for when

56:18 Speaker 1: we first opened Worldview, we had year, month, day. You'll remember,

56:22 Speaker 1: you'll notice since we included the tempo layers, we're now

56:26 Speaker 1: having an hour and minute and this is in time

56:30 Speaker 1: of UTC. And there's also this little increment feature here

56:35 Speaker 1: it is defaults to six minutes because the tempo revisit

56:38 Speaker 1: time is hourly. We're just going to click on this.

56:42 Speaker 1: When we click on these arrows, it's going to be

56:44 Speaker 1: an hourly timestep. So first here, what we're going to

56:48 Speaker 1: do is look at the nitrogen dioxide since Aaron showed

56:51 Speaker 1: some great examples there. So we're just going to click

56:54 Speaker 1: on this level three layer. And what are we seeing here?

57:00 Speaker 1: This may be something that will happen as you're exploring worldview.

57:04 Speaker 1: You can see that the values here are fairly uniform.

57:07 Speaker 1: As I'm mousing my cursor over the map, you can

57:10 Speaker 1: see on the left hand panel there's a little triangle

57:14 Speaker 1: in line that's giving you an idea what the numerical

57:17 Speaker 1: values are. They're fairly uniform values. What has happened here

57:20 Speaker 1: and you might encounter this. We have perhaps inadvertently put

57:25 Speaker 1: on the stratosphere level three layer. So this is nitrogen

57:28 Speaker 1: dioxide level three vertical column stratosphere, which is not maybe

57:33 Speaker 1: necessarily what we want to do if we are looking

57:35 Speaker 1: at the air quality issues. So now we are looking

57:40 Speaker 1: at nitrogen dioxide level three vertical column, tropospheric and sub daily.

57:45 Speaker 1: This is a little more what we would expect as

57:48 Speaker 1: we're looking at nitrogen dioxide near surface values, and again

57:53 Speaker 1: we can use the cursor to mouse over the areas

57:57 Speaker 1: of very high concentrations. You can see the values here

58:02 Speaker 1: almost three times ten at the sixteenth molecules per centimeter squared.

58:06 Speaker 1: If we move the cursor out towards the ocean, where

58:10 Speaker 1: it's a little bit more pristine or uncontaminated, we can

58:13 Speaker 1: see that the values get much lower. We can see

58:16 Speaker 1: again some other values from general other sources of pollution

58:23 Speaker 1: in the San Francisco Bay area and the central value

58:26 Speaker 1: of California, so we're seeing different sources in addition to

58:30 Speaker 1: the very strong smoke plume that we're seeing in the

58:33 Speaker 1: LA area. So now we're going to use these arrows

58:36 Speaker 1: that we set up to increment hourly, and as we

58:39 Speaker 1: do that, we're seeing again the bar down at the bottom.

58:44 Speaker 1: You can see it's stepping by an hour at a time,

58:49 Speaker 1: and we can see that the values are also changing.

58:51 Speaker 1: So we're able to see the different time steps that

58:54 Speaker 1: are going on throughout the course of this day. Looking

58:58 Speaker 1: at the scene and looking at the values that we're

59:02 Speaker 1: observing here, what has happened here? We have gone to

59:06 Speaker 1: January tenth, at twelve forty five am UTC. This corresponds

59:14 Speaker 1: to four forty five pm local California time, which is

59:21 Speaker 1: after sunset at this time. So you'll remember tempo is

59:24 Speaker 1: only available during daylight hours, and it's a little bit

59:28 Speaker 1: difficult to see, but if you look on this course,

59:33 Speaker 1: if you look on the time bar here, there's a

59:36 Speaker 1: little bit of a different shading from the nighttime values here,

59:42 Speaker 1: so it's a little bit darker, and then there's a

59:48 Speaker 1: change where this is showing where the values are the

59:52 Speaker 1: data are available over the course of the daytime. So

59:57 Speaker 1: that can give you a sense of where data are

59:59 Speaker 1: available from this instrument for this particular time. So we

1:00:05 Speaker 1: can step backwards in time again. So now we've reset

1:00:14 Speaker 1: to sort of the beginning of the day. Okay, so

1:00:17 Speaker 1: that is nitrogen dioxide level three data. We can also

1:00:22 Speaker 1: look at the two level two data, and again we're

1:00:26 Speaker 1: picking the tropospheric values. As we turn off this layer

1:00:31 Speaker 1: the tropospheric level three data and turn on the level

1:00:34 Speaker 1: two data, you can see this is how that looks.

1:00:41 Speaker 1: As Aaron mentioned, this is the level two data are

1:00:45 Speaker 1: presented in this way. One thing that you can do

1:00:49 Speaker 1: there's a little tool here called view options. If we

1:00:52 Speaker 1: click on these little slider options, we're going to increase

1:00:56 Speaker 1: the granule count here from its default of one to nine,

1:01:00 Speaker 1: because again there's nine granules in the Tempo overpass. And

1:01:07 Speaker 1: so now we have the full US one snapshot of

1:01:15 Speaker 1: Tempo is visible at a time, and we can see

1:01:18 Speaker 1: if we mouse over these, it's giving us a little

1:01:21 Speaker 1: bit more specific timestep for this swath sixteen forty five,

1:01:27 Speaker 1: sixteen thirty eight, sixteen thirty one. So these are in

1:01:30 Speaker 1: approximately six minute intervals, going from starting at the earliest

1:01:39 Speaker 1: time fifteen fifty two as it progresses from east to

1:01:43 Speaker 1: west across the US So we can see that the

1:01:50 Speaker 1: Level two data are visualized a little bit differently. Here,

1:01:56 Speaker 1: again we can step through and see the evolution of

1:02:04 Speaker 1: the scene over the course of the day to give

1:02:10 Speaker 1: a sense of the differences between Level three and Level

1:02:12 Speaker 1: two data. This is just a comparison mode. We're going

1:02:17 Speaker 1: to talk about this more on day two. Here we're

1:02:20 Speaker 1: comparing the level three data on the left and the

1:02:24 Speaker 1: Level two data on the right. And if we drag

1:02:27 Speaker 1: the slider, this is the level two data and the

1:02:31 Speaker 1: Level three data and You'll remember, as Aaron described, the

1:02:36 Speaker 1: native grid of tempo in the level two data are

1:02:39 Speaker 1: a little bit more of an angled grid, whereas the

1:02:42 Speaker 1: level three data are averaged over point zero two degree grids,

1:02:48 Speaker 1: so it's a little bit more regular. This is a

1:02:50 Speaker 1: way to sort of visualize a little bit the differences

1:02:54 Speaker 1: between the two between level two versus level three. So

1:03:00 Speaker 1: that's the No two level two data. We can also

1:03:05 Speaker 1: look at the ozone column values, so we can only

1:03:09 Speaker 1: look at the ozone total column. But let's see what

1:03:12 Speaker 1: we have. This is the ozone level three column values. Again,

1:03:20 Speaker 1: if we mouse into different places of the map, we

1:03:23 Speaker 1: can see that the values on the color bar to

1:03:27 Speaker 1: the left are changing with the different values to giving

1:03:31 Speaker 1: you a sense of the numerical values that we're looking

1:03:35 Speaker 1: at here. The main takeaway here with the ozone total

1:03:40 Speaker 1: column is it's very difficult to see any signal from

1:03:45 Speaker 1: the wildfire event that we know is happening in the

1:03:47 Speaker 1: troposphere in the column values. So that's ozone. We can

1:03:53 Speaker 1: again hide that layer and now let's look at for aldehyde.

1:03:57 Speaker 1: And looking at for aldehyde, it is similar to what

1:04:01 Speaker 1: we saw in the nitrogen dioxide. We're seeing a clear

1:04:07 Speaker 1: plume a very distinct plume associated with the wildfire smoke

1:04:14 Speaker 1: that we know is happening as it goes from beginning

1:04:21 Speaker 1: of the day. We can sort of step through hour

1:04:28 Speaker 1: by hour and see the changes the evolution in the

1:04:33 Speaker 1: scene and the values that we're seeing informaldehyde over the

1:04:37 Speaker 1: course of the day and as it disperses off the coast,

1:04:42 Speaker 1: and again we can mouse over and get a sense

1:04:44 Speaker 1: of quality the quantitative values that are within this plume

1:04:50 Speaker 1: as well. And finally, what we're going to look at,

1:04:57 Speaker 1: We're going to briefly look at the aerosol index Aaron

1:05:00 Speaker 1: mentioned as well. It's retrieved in conjunction with the ozone

1:05:04 Speaker 1: product here and higher aerosol index indicates the presence of

1:05:10 Speaker 1: absorbing aerosols, so such as dark smoke in this scene,

1:05:14 Speaker 1: which we certainly can tell is happening here. So going

1:05:19 Speaker 1: to the beginning of the day, we can see that

1:05:23 Speaker 1: detection of absorbing particles is present in this plume as

1:05:30 Speaker 1: it was observed by Tempo as it evolves over the

1:05:33 Speaker 1: course of the day. So that's the aerosol index as well.

1:05:38 Speaker 1: So one final thing that we can look at for

1:05:40 Speaker 1: this event. As Aaron mentioned, there were special operations special

1:05:46 Speaker 1: every ten minute observations that were happening a few days

1:05:49 Speaker 1: later during this same event. And so here is the

1:05:54 Speaker 1: custom interval selector one thing we're going to do. Because

1:05:57 Speaker 1: they were every ten minutes currently we're set out. We've

1:06:00 Speaker 1: had it set to be every hour. We're going to

1:06:02 Speaker 1: go to Custom and do intervals of every ten minutes,

1:06:06 Speaker 1: and so now our increment is going to be every

1:06:08 Speaker 1: ten minutes. The special operations happened on January sixteenth and seventeenth,

1:06:17 Speaker 1: so we're just going to navigate to the sixteenth, and

1:06:21 Speaker 1: let's start at eight o'clock Zulu, and then we are

1:06:29 Speaker 1: going to look at the too level two tropospheric values here,

1:06:36 Speaker 1: so we'll turn on that layer. And then because we're

1:06:40 Speaker 1: looking at special operations, we are going to go back

1:06:43 Speaker 1: into our tools and reduce the granule counts so we're

1:06:49 Speaker 1: only looking at let's say two at a time. So

1:07:07 Speaker 1: now we're only looking at two granules at a time,

1:07:10 Speaker 1: so we can x this out. And so now we

1:07:15 Speaker 1: are looking at January sixteenth, eight o'clock Zulu, which is

1:07:20 Speaker 1: noon Pacific standard time, and we are seeing there are

1:07:25 Speaker 1: two overpasses here or two swaths here. Nineteen fifty four,

1:07:31 Speaker 1: nineteen fifty nine. If we increment another ten minutes, now

1:07:36 Speaker 1: we're seeing twenty oh four twenty oh nine. As we

1:07:43 Speaker 1: step through the time incrementor of every ten minutes, we

1:07:47 Speaker 1: can see how the scene evolves over the course of

1:07:50 Speaker 1: this day. At this ten minute resolution, where we have

1:07:55 Speaker 1: the high high revisit time, high temple observations during this

1:08:03 Speaker 1: particular hour. So I encourage you to go to the

1:08:06 Speaker 1: Tempo log and see if there were any other special

1:08:10 Speaker 1: operations that may have happened at scenes of interest. So

1:08:18 Speaker 1: that concludes our visualization exercise in Worldview for today. We

1:08:22 Speaker 1: will be back in Worldview on day two to explore

1:08:27 Speaker 1: a couple more case studies that we can observe with

1:08:30 Speaker 1: these data that are available. There are also some other

1:08:36 Speaker 1: tools that you may want to explore for visualization of

1:08:40 Speaker 1: tempo data. The slides are of it will be available

1:08:44 Speaker 1: on the website after the after the training is done.

1:08:51 Speaker 1: So here are some other tools that you may be

1:08:55 Speaker 1: interested in. So this is the conc illusion of our

1:09:01 Speaker 1: part one Tempo Geostationary trace gas training to summarize what

1:09:07 Speaker 1: we went over today. Tempo's hyperspectral capability enables us to

1:09:13 Speaker 1: retrieve criteria trace gas pollutants in the troposphere at unprecedented

1:09:18 Speaker 1: spatial and temporal resolution across Greater North America. It allows

1:09:24 Speaker 1: us to observe small scale emission sources and fine scale

1:09:27 Speaker 1: pollutant gradients that have not been adequately resolved by prior

1:09:32 Speaker 1: satellite missions. The geostationary nature of its observations allow for

1:09:37 Speaker 1: monitoring of rapidly evolving pollutants which support air quality analysis

1:09:41 Speaker 1: and forecasting. The nitrogen dioxide and formaldehyde products show overall

1:09:47 Speaker 1: good agreement to the Pandora Validation Network and tropomeat satellite data.

1:09:52 Speaker 1: The ozone Profile product, which is currently in beta, will

1:09:56 Speaker 1: offer enhanced capabilities to monitor and characterize ozone concentrations within

1:10:01 Speaker 1: the troposphere. And that's a worldview applies tempo quality assurance,

1:10:06 Speaker 1: which is suitable for qualitative visualization. For more quantitative analysis,

1:10:12 Speaker 1: users will need to apply quality assurance variables to both

1:10:16 Speaker 1: level two and Level three products, which remove the tempo

1:10:20 Speaker 1: data with larger errors or uncertainties from that analysis, with

1:10:25 Speaker 1: the caution that using strict quality assurance tends to remove

1:10:28 Speaker 1: large nitrogen dioxide columns from wildfire smoke. To look ahead

1:10:36 Speaker 1: to our Part two, which again will happen in two days.

1:10:40 Speaker 1: In Part two, we will evaluate the tempo trace gas

1:10:42 Speaker 1: products in Worldview to anticipate short term air quality risks

1:10:46 Speaker 1: such as high concentrations of ozone precursors. And we will

1:10:49 Speaker 1: determine the spatial patterns, temporal trends, and likely sources of

1:10:53 Speaker 1: trace gases related to wildfire, smoke and urban air pollution

1:10:56 Speaker 1: events using tempo data in Worldview, and we will look

1:11:00 Speaker 1: at two case studies, one on some fires in the

1:11:04 Speaker 1: Colorado Front Range and then urban air quality conditions and

1:11:10 Speaker 1: pollution sources over East Texas. As a quick reminder, this

1:11:16 Speaker 1: training series will have one homework assignment. You will be

1:11:20 Speaker 1: able to access the homework assignment from the training web

1:11:22 Speaker 1: page starting January twenty second after Part two, and answers

1:11:26 Speaker 1: should be submitted via Google Forms within two weeks so

1:11:28 Speaker 1: by February fifth. Certificates of completion will be issued to

1:11:32 Speaker 1: participants who attend all of the live trainings and submit

1:11:35 Speaker 1: the homework before that deadline. The certificates will be issued

1:11:39 Speaker 1: via email about two months after the training. Here we've

1:11:44 Speaker 1: included the contact information for the speakers today, as well

1:11:48 Speaker 1: as links to the our set website and YouTube channel.

1:11:51 Speaker 1: We encourage you to sign up for our mailing list.

1:11:55 Speaker 1: Here are some additional links and resources. Thank you for

1:11:58 Speaker 1: attending today. I'm going to transition to the Q and

1:12:02 Speaker 1: A portion of the session. Okay, we've been getting some

1:12:17 Speaker 1: Q and A already coming into the WebEx. Anybody who

1:12:22 Speaker 1: has questions there should be a Q and A box

1:12:26 Speaker 1: to the bottom in the bottom right of the WebEx.

1:12:29 Speaker 1: Feel free to put your questions in that chat and

1:12:31 Speaker 1: we will try to get to them. So question one,

1:12:36 Speaker 1: I guess we'll start there. Can you explain why Tempo

1:12:39 Speaker 1: is using a continuous UV visible spectrum instead of discrete

1:12:42 Speaker 1: bands like other satellites, Aaron, do you want to elaborate

1:12:46 Speaker 1: a little bit on that?

1:12:50 Speaker 2: Yeah, hopefully can hear me? Okay, yep, sound great to me,

1:12:53 Speaker 2: all right, great, Yeah, So Tempo is using a spectrometer

1:12:59 Speaker 2: that very fine spectral resolution across that UV visible spectrum.

1:13:04 Speaker 2: So that's very critical in terms of detecting that very

1:13:08 Speaker 2: fine absorption spectrum and the differences in the absorption between

1:13:13 Speaker 2: those critical trace gases that Tempo was designed to provide

1:13:17 Speaker 2: to the community. So that's really the key part, and

1:13:21 Speaker 2: that really goes even further with the ozone concentrations because

1:13:26 Speaker 2: that the uv VIZ spectrometer does also enable improve sensitivity

1:13:31 Speaker 2: to ozone in the troposphere layer so compared to using

1:13:35 Speaker 2: distrect discrete bands like other satellites. So in general and overall,

1:13:40 Speaker 2: the uv VIZ spectrometer provides that higher precision and improves

1:13:44 Speaker 2: sensitivity to trace gases compared to using a satellite with

1:13:48 Speaker 2: distrect with discrete bands for example.

1:13:52 Speaker 3: Hopefully that clears things up.

1:13:55 Speaker 1: Thanks. Yeah, And to remind everyone, this Q and a

1:13:58 Speaker 1: document will be posted on the training web page after

1:14:01 Speaker 1: the fact as well, so you'll be able to go

1:14:04 Speaker 1: back over these. So question two is regarding the special scans.

1:14:09 Speaker 1: So basically I think is what happens when the when

1:14:14 Speaker 1: the special scans are happening. Is there a special instrument

1:14:18 Speaker 1: or or how does how do the special scans affect

1:14:20 Speaker 1: the standard scans that are going on?

1:14:28 Speaker 3: Do you mean to go for that one?

1:14:31 Speaker 1: Sure?

1:14:32 Speaker 2: Okay, yeah, so I think you kind of answer it

1:14:35 Speaker 2: there on the document here. But yeah, so we're always

1:14:39 Speaker 2: there's no special no, no separate special instrument. It's all

1:14:42 Speaker 2: the tempo spectrometer that does the special versus the standard scans, right,

1:14:47 Speaker 2: so we just transition the tempo instrument from performing the

1:14:52 Speaker 2: standard normal hourly daylight operations to a special scan mode

1:14:58 Speaker 2: during a day, right. So you saw that example in

1:15:02 Speaker 2: the training where we did that same kind of special

1:15:06 Speaker 2: scan operation for a four day time period. So you

1:15:10 Speaker 2: know we can do that too, where we set the

1:15:12 Speaker 2: special scan mode to operate for four continuous days and

1:15:16 Speaker 2: then have those ten minute you know, for example, ten

1:15:19 Speaker 2: minute scans over a certain portion of the field of

1:15:23 Speaker 2: our guard, you know, happening for that four day time

1:15:28 Speaker 2: period during the daylight.

1:15:31 Speaker 1: Yeah, And in the slides, I'll also mentioned in the

1:15:34 Speaker 1: slides there's a link to the Tempo Operations LUG which

1:15:37 Speaker 1: can give a record of when special scans were happening.

1:15:41 Speaker 1: And so if if there's data missing, it could in

1:15:45 Speaker 1: a particular time, it could be because there was a

1:15:46 Speaker 1: special scan happening at a different in a different time

1:15:49 Speaker 1: and place, or in a different place during the same time. Rather,

1:15:53 Speaker 1: we're getting a number of questions, I'll just combine them

1:15:56 Speaker 1: all together. There's a number of questions about the stratospheric

1:16:00 Speaker 1: tropospheric division and how that's, how that is, how that

1:16:04 Speaker 1: process is done, and how that distinction is made. Aaron,

1:16:07 Speaker 1: do you want to talk a little bit more about

1:16:09 Speaker 1: that in sort of a general term. I think I'm

1:16:11 Speaker 1: combining a bunch of these questions, but they're all similarly themed.

1:16:15 Speaker 2: Yeah, I did notice there's a handful of questions related

1:16:18 Speaker 2: to that. So what we really do is we actually

1:16:24 Speaker 2: use the total column, right, and we subtract out the

1:16:27 Speaker 2: tropospheric column. So really we're not retrieving information directly on

1:16:32 Speaker 2: the stratospheric portion of that. We have a you know,

1:16:36 Speaker 2: TEMPO provides a total column and information directly we use

1:16:41 Speaker 2: in a retrieval process to get at the tropospheric colum

1:16:46 Speaker 2: amount and that and after that we're able to at

1:16:49 Speaker 2: least get an idea on the stratospheric portion. That's how

1:16:53 Speaker 2: it separate out in the Tempo data prodcasts well, for

1:16:57 Speaker 2: n O two, for example, we have you can get

1:16:59 Speaker 2: information on the total column, the triposphere column, and the

1:17:03 Speaker 2: stratospheric colum portion within the Tempo data product for two.

1:17:10 Speaker 2: For formalde high, there's a question later on regarding formaldehyde,

1:17:14 Speaker 2: I believe, and for formalde high, the reason why we

1:17:17 Speaker 2: don't do the stratosphere troposphere separation is that the formatohyde

1:17:24 Speaker 2: concentrations in the stratosphere are very low, so there's not

1:17:29 Speaker 2: really a need to do that stratosphere triposphere separation for

1:17:34 Speaker 2: the formaldehyde product for example. And I think there's a

1:17:38 Speaker 2: question also relayed to ozone and the separation of that.

1:17:42 Speaker 3: Yeah, so the ozone is similar to No.

1:17:45 Speaker 2: Two in regards that we can get at information or

1:17:48 Speaker 2: we do use a separation technique right to get at

1:17:52 Speaker 2: the stratospheric and the tropospheric concentrations for ozone, which are

1:17:57 Speaker 2: within the data product itself. Ozone is special in regards

1:18:02 Speaker 2: to we're able to retrieve additional information in the lower

1:18:06 Speaker 2: triposphere layer, so we actually have a you know, you

1:18:10 Speaker 2: can get at the tropospheric or the lower tripospheric column

1:18:13 Speaker 2: amounts for or ozone within the ozone profile product. I

1:18:20 Speaker 2: think that might clear up those three or four questions

1:18:24 Speaker 2: related to troposphere for stretosphere I'm misstanding.

1:18:29 Speaker 1: Let me know, Thanks, and yeah, feel free if anybody

1:18:33 Speaker 1: has any follow up questions, you can again put them

1:18:35 Speaker 1: in that Q and A box in the WebEx. Let's see,

1:18:40 Speaker 1: so we covered those, there's question number five which is

1:18:45 Speaker 1: just about the orbital period of the satellite. Again, this

1:18:49 Speaker 1: isn't we have that in the chat this is a

1:18:53 Speaker 1: geostationary satellite, so it is constantly the orbital period is

1:18:57 Speaker 1: one day. It's constantly staring at north of And for

1:19:02 Speaker 1: more detail on that, we have some fundamentals of remote

1:19:05 Speaker 1: sensing self paced training. If you would like to get

1:19:09 Speaker 1: some more detail about the different warbal types I see numbers.

1:19:18 Speaker 1: Question six is about the lower signal to noise ratio

1:19:22 Speaker 1: with formaldehyde. Aaron, do you want to do that one

1:19:25 Speaker 1: now or should we leave it for it?

1:19:29 Speaker 2: Kind of skip over that one and let me let's

1:19:31 Speaker 2: go back to that one. I can go to the

1:19:34 Speaker 2: following question.

1:19:36 Speaker 1: Let's see how about question seven? There there was some

1:19:41 Speaker 1: striping striping feature on the trope Sphirit profile that you showed.

1:19:48 Speaker 1: Do you want to talk a little bit about what

1:19:49 Speaker 1: causes that?

1:19:51 Speaker 2: Yeah, So the ozone profile, I wouldn't make it clear

1:19:55 Speaker 2: that we are still in the validation phase of that product,

1:19:59 Speaker 2: like a like you mentioned in the training, it is

1:20:01 Speaker 2: a beta product right now. So we're undergoing pretty intense

1:20:06 Speaker 2: validation on that ozone profile product currently with the hope

1:20:11 Speaker 2: that we can have a provisional ozone profile by the

1:20:15 Speaker 2: end of this year. So, but you know, during our

1:20:18 Speaker 2: early validation results and one of our main mechanisms for

1:20:23 Speaker 2: validating our tempo ozone profile product is the whull net stations,

1:20:30 Speaker 2: which I'm not sure people know about tollnet, but it's

1:20:32 Speaker 2: a ground based validation network for our tempo ozone profile.

1:20:37 Speaker 2: Our early validation results have shown that this horizontal striping

1:20:41 Speaker 2: is kind of similar to what we see with Tropomi

1:20:44 Speaker 2: and only data, for example, in terms of insufficient calibration

1:20:49 Speaker 2: across a tempo wavelength spectrum, especially those wavelengths that are

1:20:54 Speaker 2: used for retrieving of ozone information. So it's kind of

1:20:58 Speaker 2: so much to like the row not now homies row

1:21:01 Speaker 2: phenomenomalies that you see in the trop Amy data that

1:21:05 Speaker 2: are also tied to calibration.

1:21:08 Speaker 3: Calidate calibration problems.

1:21:12 Speaker 2: And again we hope to have much more information regarding

1:21:15 Speaker 2: this and also additional improvements in the product once we

1:21:19 Speaker 2: have that. Once we have the provisional product available later

1:21:24 Speaker 2: on in twenty twenty.

1:21:25 Speaker 1: Six, sounds good. I think we're all looking forward to

1:21:30 Speaker 1: further validation and further improving of this product. But it's

1:21:33 Speaker 1: a really cool product. So question eight, how you can

1:21:40 Speaker 1: sort of reconstruct information about atmosphere processes or pollutant emissions

1:21:47 Speaker 1: during periods where the satellite is not making observations. The

1:21:51 Speaker 1: written response I think covers this. I don't know, Aaron,

1:21:54 Speaker 1: if you have anything to chime in, but essentially this

1:21:56 Speaker 1: is something that I don't think Temple can't do it

1:21:59 Speaker 1: by it else. This is a case where you would

1:22:02 Speaker 1: bring in some other perhaps modeled data or other observations

1:22:06 Speaker 1: to be able to get the fuller picture there. And

1:22:08 Speaker 1: there's links to some other trainings which may be relevant

1:22:11 Speaker 1: to people interested in pursuing that question. I don't know

1:22:14 Speaker 1: if you have anything else to add, Aaron.

1:22:18 Speaker 2: Nothing from my end, I think that response right there

1:22:22 Speaker 2: tells it all.

1:22:23 Speaker 1: Yeah. So question nine, I think you touched on this

1:22:28 Speaker 1: already about how the data calibrated with ground based observation systems.

1:22:34 Speaker 1: You mentioned Pandora, and you mentioned tole nets, and again

1:22:37 Speaker 1: there was a whole training specifically to the ground networks

1:22:41 Speaker 1: that NASA uses to be able to get at air

1:22:46 Speaker 1: quality and climate applications. Let's see. So question ten, I

1:22:56 Speaker 1: think we already covered that. That's the trope tropospheric stratospheric division.

1:23:08 Speaker 1: The question eleven's a little bit different with converting the

1:23:14 Speaker 1: vertical column values into something that is like what is

1:23:19 Speaker 1: measured at the surface and PPB or PPM, and I

1:23:23 Speaker 1: think this again goes to, yeah, there's other information needs

1:23:27 Speaker 1: to be incorporated, something like models or something like ground

1:23:31 Speaker 1: based sensors. I don't know, Eron, if you have anything

1:23:34 Speaker 1: else to add to that.

1:23:37 Speaker 2: I think that is a great response right there. It

1:23:39 Speaker 2: just takes further work and further research to really get

1:23:42 Speaker 2: at that PPM or PPB value when you're looking at

1:23:46 Speaker 2: tempo data.

1:23:48 Speaker 1: Yeah, that's always the million dollar question is converting from

1:23:51 Speaker 1: the column values to something at the surface level. So

1:23:55 Speaker 1: question twelve, I don't know Eron if you want to

1:23:58 Speaker 1: elaborate a little bit more on this, but I think

1:24:00 Speaker 1: this is essentially talking about how can researchers best deal

1:24:05 Speaker 1: with places where there are a lot of clouds going on?

1:24:10 Speaker 1: And we talked about it with Quality Assurance. Temple can't

1:24:13 Speaker 1: see through clouds, so you need to sort of filter

1:24:16 Speaker 1: with that. Any anything else you want to add with

1:24:20 Speaker 1: how how to deal with areas which are very cloudy.

1:24:24 Speaker 2: Yeah, it's gonna be a problem, right, I mean, when

1:24:26 Speaker 2: you're looking at cloud covered areas that are you know,

1:24:32 Speaker 2: cloud fractions with the cloud cover is encompassing too much

1:24:35 Speaker 2: of the tempo footprint right where the tempo is trying

1:24:37 Speaker 2: to perform that retrieval, it's going to be very difficult

1:24:41 Speaker 2: to provide a high quality retrieval. So depending on what

1:24:46 Speaker 2: is meant by cloud noise, which I guess is just

1:24:50 Speaker 2: a tempo footprint that has a lot of cloud within

1:24:52 Speaker 2: it to really get you really can't get high quality

1:24:57 Speaker 2: information on from tempo in those areas where the cloud

1:25:01 Speaker 2: for action is high in the footprint. So, like we

1:25:05 Speaker 2: mentioned in the quality assurance discussion, even when doing a

1:25:10 Speaker 2: qualitative look at tempo data, we recommend using filtering out

1:25:15 Speaker 2: any tempo footprint or tempo data that's tied or associated

1:25:20 Speaker 2: with the cloud for action of a grade and fifty percent.

1:25:22 Speaker 2: So that's probably at the level where this cloud always

1:25:26 Speaker 2: become a problem.

1:25:30 Speaker 1: Yeah, one of the challenges in doing remote sensing observations

1:25:36 Speaker 1: is those darn clouds sometimes get in our way. So

1:25:40 Speaker 1: question thirteen is asking about whether the AOD product is

1:25:44 Speaker 1: going to be in worldview. It's not currently. We are

1:25:49 Speaker 1: planning to have a tempo aerosol product training later this

1:25:54 Speaker 1: year when those data are available, or more broadly, I

1:26:03 Speaker 1: don't know if we have any better timeline on that

1:26:05 Speaker 1: but I think I think later this year is what

1:26:09 Speaker 1: I think we can confidently say hopefully that the AOD

1:26:13 Speaker 1: product will be available.

1:26:17 Speaker 2: Yeah, okay, hopefully may ish, potentially after, but yeah, hopefully

1:26:25 Speaker 2: may ish. Yes, the question I should clarify here the

1:26:30 Speaker 2: level four AOD product, it's really a level four PM

1:26:33 Speaker 2: two point five product.

1:26:35 Speaker 3: The level the AOD is actually a level two product.

1:26:41 Speaker 1: Is there a level two and a level three AOD

1:26:43 Speaker 1: that will be No.

1:26:45 Speaker 2: So that is one difference between what we're generating for

1:26:48 Speaker 2: our aerosol products versus our versus our trace gas products.

1:26:53 Speaker 3: So for our as you saw.

1:26:55 Speaker 2: On the table earlier and some of the presentation material,

1:27:00 Speaker 2: we provide the trace gas products on both the standard

1:27:05 Speaker 2: tempo footprint but also on a standard grid level three

1:27:10 Speaker 2: standard regular grid, which is referred to our level three product.

1:27:15 Speaker 2: They are we are not doing that for the aerosol.

1:27:17 Speaker 2: So we'll have the aerosols as a level or AOD

1:27:21 Speaker 2: as a level two product, and then additional retrieval processes

1:27:27 Speaker 2: are used to drive that level four PM two point

1:27:31 Speaker 2: five product.

1:27:35 Speaker 1: Sounds good. Thanks for the clarification there. Question fourteen, Yeah,

1:27:41 Speaker 1: I think that's an interesting question, but a little bit

1:27:44 Speaker 1: beyond the scope of what we are doing right here

1:27:48 Speaker 1: in terms of how solar radiation modification will affect these products.

1:27:53 Speaker 1: That's I don't know that we have anything else to

1:27:59 Speaker 1: add beyond what's written on the screen right there, but

1:28:03 Speaker 1: it could be interesting. Question fifteen about the validation of

1:28:10 Speaker 1: tropospheric ozone products. I think we covered this in detail

1:28:13 Speaker 1: and in some of the in the presentation some of

1:28:16 Speaker 1: the other Q and A. I don't know if we

1:28:18 Speaker 1: have anything else specific to add there, Aaron, But that's

1:28:23 Speaker 1: done with toll net and done with Pandora. Yeah.

1:28:28 Speaker 2: Again, like our our Pandora network was really more for

1:28:33 Speaker 2: validing our total column of zone product.

1:28:38 Speaker 3: The toll net, the toll.

1:28:40 Speaker 2: Net program with the ozone profile retrievals from tollnet are

1:28:44 Speaker 2: really the main, the main validation mechanism for our tempo

1:28:50 Speaker 2: ozone profile. So yeah, we're more work to be done

1:28:53 Speaker 2: there and we'll have much more results and progress by

1:28:56 Speaker 2: late later on in twenty twenty six.

1:29:02 Speaker 1: Sounds good. Question sixteen talks just about the difference between

1:29:06 Speaker 1: geostationary orbit and polar orbiting. I think that was covered

1:29:10 Speaker 1: in the slides as well, and the fundamentals of remote sensing.

1:29:14 Speaker 1: I know we're at our ending time. But let's can

1:29:20 Speaker 1: we just perhaps Aaron, if you want to end on

1:29:23 Speaker 1: question seventeen, which is the when is the reprocessing for

1:29:27 Speaker 1: the version four going to be available? Do we have

1:29:31 Speaker 1: a timeline on that?

1:29:35 Speaker 3: Man, I would love to provide a timeline.

1:29:38 Speaker 2: Question with the Tempo team back and forth on this,

1:29:42 Speaker 2: and we don't really have a good timeline for when

1:29:46 Speaker 2: all that reprocessing will be done. But I will provide

1:29:51 Speaker 2: an email, and we will provide an email to all

1:29:54 Speaker 2: the Tempo folks and on our various email chains and

1:30:00 Speaker 2: email communication chains when that reprocessing is complete. But I

1:30:05 Speaker 2: hope I'm gonna throws a ballpark. We hope by the

1:30:09 Speaker 2: middle of us a year, it'll be all done and reprocessed.

1:30:12 Speaker 2: But I don't have a good timeline. Unfortunately, there's a

1:30:17 Speaker 2: lot of data to.

1:30:18 Speaker 1: Go back over and there's uh yeah, it's I want

1:30:22 Speaker 1: to be sure we're doing it right, and it can

1:30:24 Speaker 1: take some time to go back over for years.

1:30:28 Speaker 2: The key is will provide an email notification when that's done,

1:30:32 Speaker 2: so everybody will know if you're on our emails.

1:30:36 Speaker 1: So sign sign up for the emails exactly when that

1:30:42 Speaker 1: will happened, so I know we're at time. I think

1:30:50 Speaker 1: a couple.

1:30:52 Speaker 3: We can try to maybe address question six. That was

1:30:55 Speaker 3: a bit.

1:30:57 Speaker 2: I guess the main thing with from out of hide

1:31:00 Speaker 2: it's just a challenging trace gas. Do you know when

1:31:04 Speaker 2: you're trying to use a satellite to measure from aldehyde,

1:31:07 Speaker 2: it has a lower atmospheric composition compared to n O two.

1:31:13 Speaker 2: We have much weaker not much, but a weaker spectral

1:31:16 Speaker 2: signature compared to n O two. For example, we don't

1:31:20 Speaker 2: have that same sensitivity that we do with from Autohyde

1:31:24 Speaker 2: as n O two. But that being said, we can

1:31:27 Speaker 2: still we have shown that our from the hide to

1:31:30 Speaker 2: retrieval is good from tempo, just not as good as

1:31:33 Speaker 2: our n O two due to these challenging you know,

1:31:36 Speaker 2: these other challenging issues tied to from Autohyde in the

1:31:39 Speaker 2: in the proposphere.

1:31:47 Speaker 3: Hopefully that clears that up.

1:31:58 Speaker 1: Right, you know, there were a couple more questions that

1:32:03 Speaker 1: came in. I don't know if we want to stay

1:32:06 Speaker 1: on or regardless, these will all be answered in this

1:32:11 Speaker 1: document and posted on the website. So question eighteen is

1:32:16 Speaker 1: referring to question thirteen. Yeah, so PM two point five,

1:32:31 Speaker 1: I think the link I think we're asking for the Okay,

1:32:39 Speaker 1: I think the link to where the aerosol tempo training

1:32:44 Speaker 1: will happen. I think if you sign up for the

1:32:49 Speaker 1: our set mailing list, the announcement will go out when

1:32:52 Speaker 1: we have dates on that. We're we don't have set

1:32:55 Speaker 1: dates for when the Tempo Aerosol training will be happening

1:33:00 Speaker 1: at the moment because we want to make sure the

1:33:02 Speaker 1: products are available at that point. But I think sign

1:33:07 Speaker 1: up for the mailing list and you'll get you'll be

1:33:10 Speaker 1: the first to know when the registration opens for that

1:33:14 Speaker 1: that training.

1:33:18 Speaker 3: Do you see that question nineteen that came in there?

1:33:20 Speaker 3: If we want to address that or not?

1:33:23 Speaker 1: That one's all you do? You want to have anything

1:33:26 Speaker 1: to add to that?

1:33:27 Speaker 3: Well, people need to do that, No they don't.

1:33:30 Speaker 2: I mean, right, like we mentioned in the training, we

1:33:34 Speaker 2: use the GEOCF model for this CTM and conversion of

1:33:40 Speaker 2: in the in the generation of a MSS two for

1:33:43 Speaker 2: our products. So and the GEOCF is at a quarter

1:33:47 Speaker 2: degree a resolution or grid spacing.

1:33:51 Speaker 3: So I'll just leave it at that.

1:33:52 Speaker 2: I mean there are things that can be done with

1:33:55 Speaker 2: other models to bring in that may be higher resolution, right,

1:33:59 Speaker 2: but that another.

1:34:02 Speaker 3: More intense training.

1:34:04 Speaker 1: And I think for for purposes of this training, probably

1:34:08 Speaker 1: you all on the back end, on the on the

1:34:10 Speaker 1: tempo side of things have already incorporated that. So I

1:34:14 Speaker 1: think what I would think is people don't need to

1:34:18 Speaker 1: worry about that, and you can use the products as

1:34:23 Speaker 1: they are and don't need to worry about that. Yeah, okay,

1:34:32 Speaker 1: I think we might be at the end then. Thanks

1:34:34 Speaker 1: to everyone who attended, and hope to see everyone two

1:34:40 Speaker 1: days from now on on Thursday, same time or a

1:34:45 Speaker 1: few hours later. You can find the link on the website.

1:34:47 Speaker 1: There's there's two sessions each day. Thanks everyone for joining

1:34:52 Speaker 1: and I'll see you next time.

1:34:56 Speaker 2: Thank you as