NASA ARSET NASA ATL24 Bathymetric Data for Coastal and Near-Shore Applications

0:00 Speaker 1: Thank you for joining today's r SET training NASA satellite

0:03 Speaker 1: laser altimetry for coastal and nearshore bathymetry. I'm Sean McCartney

0:08 Speaker 1: from NASA Goddard Spaceflight Center in Maryland, serving as your

0:11 Speaker 1: RSET trainer. I'm excited to begin this first part of

0:15 Speaker 1: our comprehensive two part satellite bathymetry training with all of you.

0:21 Speaker 1: For newcomers to our SET trainings, the following slides present

0:24 Speaker 1: a short program introduction. The Applied Remote Sensing Training Program

0:29 Speaker 1: or r SET, part of NASA's Earth Action Capacity Building program,

0:34 Speaker 1: offers free, accessible training on remote sensing technologies, methods, and applications.

0:41 Speaker 1: Arset's mission is to increase Earth science, remote sensing and

0:45 Speaker 1: model data use in decision making by providing training for

0:48 Speaker 1: professionals in the public and private sectors, environmental managers, and policymakers.

0:55 Speaker 1: On the right of the slide you can see all

0:57 Speaker 1: the different thematic areas in which we conduct trainings. Our

1:01 Speaker 1: SET offers both online and in person training sessions designed

1:05 Speaker 1: for beginners through advanced users. All trainings are provided at

1:09 Speaker 1: no cost through live instruction, instructor led sessions, or self

1:14 Speaker 1: paced formats like the fundamentals of remote sensing course on

1:17 Speaker 1: arset's Learning management system. Materials are available in multiple languages

1:23 Speaker 1: and can be freely used and adapted for educational purposes.

1:27 Speaker 1: Users of our set, methods and data should acknowledge the

1:30 Speaker 1: NASA applatter Remote Sensing training program. Additional information is available

1:35 Speaker 1: on our website. The following slides provide an overview of

1:40 Speaker 1: the two part webinar series NASA satellite laser altimetry for

1:44 Speaker 1: coastal and nearshore bathymetry. So why would somebody want to

1:48 Speaker 1: take this training? Well, nearly eighty percent of the world's

1:52 Speaker 1: oceans are unexplored and unmapped. Successful coastal management depends on monitoring,

1:59 Speaker 1: and coastal bathymetry is essential for navigational hazards, for vessel operations,

2:06 Speaker 1: tidal modeling and prediction, tsunami risk assessment and forecasting, underwater

2:12 Speaker 1: cultural heritage preservation, and environmental change monitoring. Satellite derived bathymetry

2:18 Speaker 1: enhances capacity for collecting high resolution, accurate depth data supporting

2:24 Speaker 1: varied applications. By the end of this two part training,

2:28 Speaker 1: participants will be able to identify NASA bathymetry data used

2:32 Speaker 1: for global, coastal and nearshore bathymetry mapping for risk reduction

2:36 Speaker 1: relating to shipping and navigation. Identify the applications and limitations

2:41 Speaker 1: of ice AT two bathymetry data ATL twenty four for

2:45 Speaker 1: coastal and nearshore bathymetry mapping, Plot and download i AT

2:49 Speaker 1: two bathymetry data using slide roll web client, and analyze

2:54 Speaker 1: i AT two bathymetry data using slide roll Python climate

2:58 Speaker 1: Cli client. The prerequisites for the training are two are

3:03 Speaker 1: set courses Fundamentals of Remote Sensing course and mapping and

3:07 Speaker 1: monitoring lakes and reservoirs with satellite observations. From December second

3:13 Speaker 1: to December fourth, there will be two one and a

3:16 Speaker 1: half hour sessions which will include presentations, demonstrations, and question

3:21 Speaker 1: and answer sessions. All materials and recordings for each session

3:25 Speaker 1: will be available from the training web page. If you

3:28 Speaker 1: are not able to attend one part, a recording will

3:31 Speaker 1: be made available within forty eight hours on the ARSET

3:34 Speaker 1: training web page. Homework opens on December fourth and will

3:39 Speaker 1: be due on December thirty first. You will be able

3:43 Speaker 1: to access the homework from the training page on Thursday,

3:46 Speaker 1: December fourth. A certificate of completion will be awarded to

3:51 Speaker 1: those who attend all live sessions and complete the homework

3:54 Speaker 1: assignment by the given due date. Part one of the

3:58 Speaker 1: training is focused on NASA ei AT twenty four bathymetric

4:02 Speaker 1: data for coastal and nearshore applications. The objectives of Part

4:07 Speaker 1: one of the training are as follows. By the end

4:10 Speaker 1: of today, participants will be able to identify nasabathymetry data

4:15 Speaker 1: used for global coastal and nearshore bethymetry mapping for risk

4:18 Speaker 1: reduction relating to shipping and navigation. Identify the applications and

4:23 Speaker 1: limitations of i AT two bethymetry data ATL twenty four

4:27 Speaker 1: for coastal and nearshore bethymetry mapping and plot, and download

4:31 Speaker 1: IAT two bethymetry data using slider roll web client. Please

4:38 Speaker 1: put your questions in the questions box and we will

4:40 Speaker 1: address them at the end of the webinar. Feel free

4:43 Speaker 1: to answer your questions as we go. We will try

4:46 Speaker 1: to get to all the questions during the Q and

4:47 Speaker 1: A session. After the webinar, the remainder of the questions

4:50 Speaker 1: will be answered in the Q and A doc which

4:53 Speaker 1: will be posted to the training website about one week

4:55 Speaker 1: after today's training. It is not my pleasure to introduce

5:00 Speaker 1: the guest trainers for Today's webinar Doctor Lauria Magruder, doctor

5:04 Speaker 1: Dian Fritz, and Joseph Paul Swinsky. Laura Magruder serves as

5:09 Speaker 1: Associate professor in Aerospace Engineering at the University of Texas

5:12 Speaker 1: at Austin, with previous roles at Jet Propulsion Laboratory Ann JOHNS.

5:17 Speaker 1: Hopkins Applied Physics Laboratory. Following nine years of science team

5:21 Speaker 1: leadership on NASA's i SAT two mission, encompassing mission development, implementation, guidance,

5:27 Speaker 1: and operational instrument pointing strategies, Magruder has transitioned to leadership

5:32 Speaker 1: positions on additional remote sensing science and instrument teams. Dian

5:37 Speaker 1: Fritz holds a position of Associate Scientist at the National

5:40 Speaker 1: Snow and Ice Data Center NSIDC, located in Boulder, Colorado,

5:46 Speaker 1: and serves as a Remote sensing educator at the University

5:49 Speaker 1: of Colorado. Her academic credentials include a doctorate in geological Sciences. J. P.

5:56 Speaker 1: Swinskey is a twenty four year NASA Goddard software developer

6:00 Speaker 1: who spent most of his career creating flight software for

6:02 Speaker 1: spacecraft and science instruments. He led flight software development for

6:06 Speaker 1: the Atlas instrument from twenty ten to twenty sixteen. In

6:10 Speaker 1: twenty eighteen he transitioned to cloud based data processing solutions,

6:14 Speaker 1: and in twenty twenty co developed slide rule with the

6:17 Speaker 1: ice AT two Project Scientist and University of Washington, which

6:20 Speaker 1: is a public web service providing fast cloud access to

6:24 Speaker 1: ice AT two data. Laurie over to you.

6:31 Speaker 2: Thanks Sean. I'm so happy to be here and talk

6:34 Speaker 2: about IAT to the new bathymetry product that has just

6:39 Speaker 2: come out earlier this year. I can honestly say that

6:43 Speaker 2: I SAT too is my favorite satellite, and I worked

6:46 Speaker 2: on it for many, many years, and so this is

6:48 Speaker 2: just a really exciting time for me to be here

6:51 Speaker 2: and provide an overview of the mission and introduction to

6:55 Speaker 2: the ATL twenty four Coastal and nearshore bathymmetry product. So

7:01 Speaker 2: let's start with the mission overview. The Ice Cloud and

7:06 Speaker 2: Land Elevation Satellite two is an Earth observing laser altimetry

7:10 Speaker 2: mission that launched in twenty eighteen, so we've been on

7:14 Speaker 2: orbit for a little more than seven years. It completed

7:17 Speaker 2: its prime mission and successfully satisfied all its science requirements

7:22 Speaker 2: in twenty twenty one, but all operations are still nominal

7:25 Speaker 2: and it's continuing to give us amazing data. I SAT

7:30 Speaker 2: two carries a single instrument called ATLAS, the Advanced Topographic

7:35 Speaker 2: Laser Altimeter System. ATLAS is a photon counting laser altimeter,

7:39 Speaker 2: meaning it's sensitive to single photon reflections from the Earth's surface.

7:43 Speaker 2: This state of the art technology to date has produced

7:46 Speaker 2: over two trillion laser pulses. Turns out, that's a whole

7:50 Speaker 2: lot more than the two billion from the original IAT

7:53 Speaker 2: mission in two thousand and three to two thousand and nine.

7:57 Speaker 2: Right now, the prediction is that IAT two has all

8:00 Speaker 2: consumables on board, meaning laser energy and fuel to continue

8:05 Speaker 2: operating until twenty thirty three. So ATLAS uses a neodymium

8:14 Speaker 2: doped yttrium orthovanidate laser to produce six individual beams that's

8:21 Speaker 2: configured into three beam pairs that you can see in

8:24 Speaker 2: the schematic here on the slide. Each pair contains a

8:28 Speaker 2: weak and a strong beam with respect to laser energy

8:32 Speaker 2: at a one to four energy ratio. The pair components

8:36 Speaker 2: are separated by ninety meters in the cross track direction,

8:40 Speaker 2: and the pair to pair separation is three point three kilometers.

8:46 Speaker 2: Each of the six beams provides an elevation profile of

8:50 Speaker 2: the illuminated surface with around a twelve meter diameter footprint,

8:55 Speaker 2: so these are just creating pencil lines of elevation across

8:58 Speaker 2: the Earth's surface. Atlas output wavelength is a five to

9:03 Speaker 2: thirty two nanometer which is in the visible green part

9:07 Speaker 2: of the spectrum, and a pulse width of one nanoseconds

9:10 Speaker 2: with one nanosecond which speaks to the high vertical resolution

9:15 Speaker 2: of the system. Photon counting altimetry allows some unique advantages

9:20 Speaker 2: in terms of data collection. So since the receiver is

9:23 Speaker 2: sensitive to single photons, it allows the system to operate

9:27 Speaker 2: at a lower energy level compared to previous or other

9:30 Speaker 2: types of light our technology. In turn, that means that

9:34 Speaker 2: the system can pulse at a faster repetition rate, and

9:38 Speaker 2: in this case, Atlas has a ten killer repetition rate

9:41 Speaker 2: that at the altitude, which is about a five hundred

9:45 Speaker 2: kilometer average altitude, it can produce a laser footprint or

9:50 Speaker 2: puts down a laser footprint every seventy centimeters in the

9:53 Speaker 2: long track direction. The final thing I'll mention here in

9:57 Speaker 2: terms of the collection strategy is the fact that the

10:00 Speaker 2: orbital plane is inclined at ninety two degrees from the

10:03 Speaker 2: equatorial plane, which means that the coverage for I set

10:07 Speaker 2: two goes from eighty eight degrees south latitude to eighty

10:11 Speaker 2: eight degrees north latitude. So I think it's important to

10:16 Speaker 2: get oriented with respect to what the IST two data

10:19 Speaker 2: looks like. So I'm showing you an example elevation profile

10:23 Speaker 2: over West Greenland west Central Greenland. The horizontal axis is

10:27 Speaker 2: latitude and the vertical axis is the elevation. So each

10:31 Speaker 2: of these black dots indicates the Atlas detection a detected photon.

10:37 Speaker 2: By Atlas, every photon has its own latitude, longitude, and elevation,

10:42 Speaker 2: So the challenge with photon counting is the need to

10:45 Speaker 2: separate the signal photons from the noise photons. So although

10:48 Speaker 2: in this case it's pretty easy for us to look

10:50 Speaker 2: at the profile and identify the surface which is this

10:53 Speaker 2: dense black line, it's a lot more challenging to automate

10:57 Speaker 2: the process of signal finding. Signal finding also differs greatly

11:02 Speaker 2: with surface reflectants, surface type, and surface feature length scales.

11:07 Speaker 2: So I set two provides data products for specific surface

11:11 Speaker 2: types that have been produced with unique algorithms optimized for

11:14 Speaker 2: the surface characteristics. So when we start to explore these

11:19 Speaker 2: data over the last seven years on orbit. You can

11:22 Speaker 2: see through aggregation all of the impact that I set

11:26 Speaker 2: to has had with respect to the specific data products

11:29 Speaker 2: that provides. We see sea ice measurements both in the

11:33 Speaker 2: Southern Ocean and the Arctic. We see the land ice

11:36 Speaker 2: and mass change for Greenland and Antarctica. And next comes

11:42 Speaker 2: the brown or the terrain, the land surface elevations, and

11:46 Speaker 2: then the vegetation canopy heights, which will be in green

11:50 Speaker 2: here in just a minute. Additionally, we have data products

11:55 Speaker 2: over the ocean and inland water surfaces. However, what was

12:02 Speaker 2: somewhat unexpected was the ability of YET two to measure

12:06 Speaker 2: beneath the water surface or measure underwater topography, which is

12:11 Speaker 2: also called bathymetry. It's common knowledge that the green laser

12:16 Speaker 2: wavelength can penetrate water, but it was not fully anticipated

12:19 Speaker 2: that a photon counting system from three hundred miles in

12:22 Speaker 2: space would perform as well as it does. You can

12:27 Speaker 2: see here an example of the bathymetry that we've been

12:30 Speaker 2: seeing for quite some time. It's a sea surface which

12:34 Speaker 2: is this dark black line horizontal on the elevation profile,

12:39 Speaker 2: and then underneath is the underwater topography or bothymmetry. In

12:43 Speaker 2: this case we're at the Great Barrier Reef. Studies of

12:47 Speaker 2: this performance or the capability of bothymmetry has happened in

12:52 Speaker 2: various locations, and ultimately we know that ATLAS can measure

12:56 Speaker 2: bathymetry up to forty meters in depth for clear water conditions. Certainly,

13:02 Speaker 2: it varies depending on where you are geographically. However, until

13:07 Speaker 2: this year, we've not had a dedicated withymmetry data product

13:10 Speaker 2: from the mission. I thought it'd be important to just

13:13 Speaker 2: show you a snapshot of the data products for iat IO.

13:18 Speaker 2: The level zero and Level one products are telemetry and

13:22 Speaker 2: converted telemetry products. The level two products are ATLO three,

13:27 Speaker 2: which is the geolocated point cloud, and then ATILO four,

13:31 Speaker 2: which are normalized backscatter profiles. ATLO three is the input

13:38 Speaker 2: to all of the Level three A products, and those

13:41 Speaker 2: are those blue boxes horizontal or in the middle of

13:45 Speaker 2: the slide. These are a long track products produced from

13:50 Speaker 2: these optimized algorithms that I mentioned previously. These Level three

13:54 Speaker 2: A blue boxes include those data we saw from Land

13:57 Speaker 2: ice Heights. C ice cis free board land and vegetation inland,

14:03 Speaker 2: water and ocean. But then you can see this this

14:08 Speaker 2: newly added box here ATIL twenty four, which is the

14:11 Speaker 2: along track near shore coastal bathymetry product which I'm happy

14:16 Speaker 2: to introduce to you today as an introduction. ATL twenty

14:22 Speaker 2: four was developed by the collaborative team. My team here

14:27 Speaker 2: at University of Texas at Austin Center for Space Research

14:31 Speaker 2: Jeff Perry and Matthew Holwell, and our teammates at Oregon

14:36 Speaker 2: State University Chris Parrish, Keana Keef, JP Swinsky at slide

14:41 Speaker 2: ROLL and all of the mission or Project Science Office

14:44 Speaker 2: partners Jeff Lee, Tom Newman, Dennis Felixon. We're all working

14:50 Speaker 2: together and released a TAIL twenty four on April first,

14:54 Speaker 2: earlier this year. Ultimately, ATAIL twenty four is designed to

14:59 Speaker 2: provide the community signal finding and photon classification for sea

15:03 Speaker 2: surface and seafloor along global coastlines. We had the goal

15:08 Speaker 2: of creating a robust algorithm that could be successful at

15:12 Speaker 2: the global scale, provide appropriate corrections to the data and

15:16 Speaker 2: confidence values of the signal classifications, and then also a

15:20 Speaker 2: per point uncertainty. So these were the goals of ATAIL

15:24 Speaker 2: twenty four when we first initiated development, which was a

15:29 Speaker 2: couple of years ago. What I'd like to emphasize here

15:33 Speaker 2: is the significant challenge with developing a single algorithmic approach

15:38 Speaker 2: that could satisfy all of those product goals. However, to

15:42 Speaker 2: meet this challenge, we use a novel machine learning approach

15:46 Speaker 2: called an ensemble. The ensemble creates a robust capability in

15:51 Speaker 2: the environment that is highly variable when you consider that

15:56 Speaker 2: everywhere there's a different bottom type, a different surface reflection,

16:00 Speaker 2: changing water turbidity levels, and sea surface conditions. So AHL

16:05 Speaker 2: twenty four also provides a confidence value for each photon

16:09 Speaker 2: that indicates an estimate of the classification fidelity, and then

16:13 Speaker 2: additionally we correct this with an index of refraction correction.

16:19 Speaker 2: The inputs of the ensemble are classification predictions from several

16:23 Speaker 2: base algorithms, and when I say classification predictions, it is

16:27 Speaker 2: that photon a sea surface, is it the theymetry or

16:31 Speaker 2: is it unknown right now? The ensemble gathers input from

16:35 Speaker 2: six distinct methods, and since the errors of each classification

16:40 Speaker 2: based algorithm are uncorrelated, the ensemble is able to predict

16:44 Speaker 2: the symmetry from the combination of predictions from the base

16:48 Speaker 2: algorithm input. So this creates a scenario of the whole

16:53 Speaker 2: is greater than the sum of the parts, meaning that

16:56 Speaker 2: the ensemble performs at least as well as the best

16:59 Speaker 2: performing algorithm for a given location. Ultimately, we're taking advantage

17:04 Speaker 2: of the fact that the single that a single signal

17:06 Speaker 2: finding algorithm cannot perform equally everywhere. So what I want

17:11 Speaker 2: to point out here in the upper right is just

17:14 Speaker 2: a granular a transect that we pulled out, and it

17:17 Speaker 2: shows the different colors just below the area where the

17:20 Speaker 2: turquoise photons are those classified as sea surface and the

17:24 Speaker 2: brighter blue photons are those that the ensemble predicts as

17:27 Speaker 2: the symmetry. And then I'll point out here this plot

17:32 Speaker 2: on the bottom right is each column is the F

17:38 Speaker 2: one score of the base algorithms and the ensemble for

17:42 Speaker 2: predicting sea surface and seafloor. So the ensemble is the

17:45 Speaker 2: column in the middle, and then you have three base

17:48 Speaker 2: algorithms on each side to the left and to the right.

17:52 Speaker 2: F one score is a metric that we often use

17:55 Speaker 2: in combination with a lot of other metrics to understand

17:58 Speaker 2: how good we're doing in terms of accurately classifying the

18:02 Speaker 2: photons as sea surface or c floor. So the green

18:05 Speaker 2: labels are each algorithm of how they predict C surface,

18:10 Speaker 2: and you can see that each of those values is

18:13 Speaker 2: almost they're pretty equal and almost to one, which would

18:17 Speaker 2: be a perfect classification score F one score. But the

18:22 Speaker 2: brown values are the the F one scores for classifying

18:26 Speaker 2: C floor. And you can see here that there's variable

18:29 Speaker 2: in this location. Each algorithm performs differently based on those

18:34 Speaker 2: variations of values. But the ensemble the column in the

18:37 Speaker 2: middle shows the highest F one score for classifying sea surface.

18:42 Speaker 2: And like I said that the ensemble is created to

18:46 Speaker 2: perform at least as well as the highest performing base algorithm.

18:52 Speaker 2: Another key benefit to this ensemble, which is an XG boost,

18:57 Speaker 2: is that it can look at its decision trees and

19:00 Speaker 2: create a probability of classification correctness if you will. And

19:07 Speaker 2: that's what we've turned into a confidence score. This is

19:12 Speaker 2: a really good visualization, and then I'll get to the

19:14 Speaker 2: confidence score of the comparisons of classifications of the base algorithms.

19:20 Speaker 2: So you can see here the top is manually labeled

19:24 Speaker 2: reference data. So this is what we say is the

19:26 Speaker 2: quote ground truth of sea surface and sea floor for

19:30 Speaker 2: this particular profile. Then each profile moving down from the

19:34 Speaker 2: manually labeled reference data is the output. So the top

19:38 Speaker 2: one is the ensemble, and you can see as you

19:41 Speaker 2: go through we have the six input I think we

19:45 Speaker 2: just have five because I couldn't sit in on the slide,

19:47 Speaker 2: but you can see the variation, Like if you look

19:50 Speaker 2: at the bathy pathfinder algorithm that over here towards the right,

19:54 Speaker 2: it gets a little confused and grabs some of the

19:56 Speaker 2: water column and labels it as the symmetry, whereas maybe

20:00 Speaker 2: if we skip down to the media and filter, which

20:02 Speaker 2: is second from the bottom profile, you can see that

20:05 Speaker 2: for whatever reason, that algorithm misses some of those bathymetry

20:09 Speaker 2: points on the right. So each of these algorithms has

20:12 Speaker 2: a different type of error associated with it or capability

20:17 Speaker 2: in correctly signal finding, but we can use all of

20:21 Speaker 2: them in the ensemble to create the best guess of classification.

20:27 Speaker 2: So the confidence as I mentioned, is an output of

20:30 Speaker 2: the XG boost ensemble, and it's kind of internally generated,

20:36 Speaker 2: but it has a lot of variables, and it's also

20:39 Speaker 2: based on this current composition of base algorithms. So what

20:44 Speaker 2: I want to emphasize here is that right now for

20:47 Speaker 2: version one release, when we feel that the confidence level

20:50 Speaker 2: of zero point six or higher is a good way

20:54 Speaker 2: to users can filter the data for high confidence classifications. However,

21:00 Speaker 2: that's based on the input algorithms, as I said, and

21:04 Speaker 2: so next versions might have a different threshold recommended of confidence,

21:10 Speaker 2: so be mindful of that as you look to versions

21:14 Speaker 2: two release excuse me, version one, Release two, and version

21:19 Speaker 2: two that comes out probably next year, that those confidence

21:21 Speaker 2: thresholds are filtering might change. This shows you those C

21:27 Speaker 2: floor classification photon classifications from about point six too point one,

21:33 Speaker 2: and you can see it is just by visually inspecting

21:37 Speaker 2: that it seems pretty accurate and there's very few outliers

21:40 Speaker 2: in terms of what the ensemble is predicting. So in

21:46 Speaker 2: terms of the full A tail twenty four workflow, I

21:50 Speaker 2: have a version here. So let's start in the upper

21:56 Speaker 2: left box that shows the intersection of the ATL of

22:00 Speaker 2: three granules and then our ATL twenty four search mask.

22:05 Speaker 2: The search mask is a global mask that uses a

22:09 Speaker 2: derived retrievability score for probable I SAT to bothymmetry detection,

22:14 Speaker 2: and this is based on water turbidity and maximum depth

22:17 Speaker 2: of a given region. And then we take that mask

22:21 Speaker 2: and we widen vertically and horizontally to ensure that we

22:25 Speaker 2: don't miss anything. And the mask guides the processing of

22:29 Speaker 2: ATL twenty four to just our coastline near shore environments,

22:34 Speaker 2: so moving down along the less left side, we process

22:38 Speaker 2: the data and use a function we created called costnet

22:41 Speaker 2: to make a first pass at identifying water surface photons.

22:45 Speaker 2: This is because we have these six independent algorithms for

22:49 Speaker 2: classifying bathymmetry photons, but some of them need a starting

22:52 Speaker 2: point of where the water surface is, and that's why

22:55 Speaker 2: we do a first approach to water surface classification. Each

23:00 Speaker 2: of these these algorithms use a different approach to signal finding.

23:04 Speaker 2: Some of them are histogramming approach or density evaluations, some

23:08 Speaker 2: of them are machine learning approaches, but again all of

23:11 Speaker 2: these in all of these predictions from these base algorithms

23:14 Speaker 2: are fed into the ensemble. Once the ensemble provides the

23:20 Speaker 2: classification predictions again which is supposed to be the greater

23:25 Speaker 2: than the some of the parts, we perform some blunder

23:27 Speaker 2: detection and outlier removal, which are if we go down

23:31 Speaker 2: the gray boxes in the middle of the flow chart,

23:34 Speaker 2: and then we do a final sea surface finding and

23:37 Speaker 2: provide a correction for the index of refraction of water

23:42 Speaker 2: for any of the photons that occur beneath the sea surface.

23:47 Speaker 2: The uncertainty calculation is calculated from a subaqueous uncertain D

23:54 Speaker 2: lookup table, which is the parallelogram at the bottom, and

23:59 Speaker 2: then it all so uses the coefficient of diffusion diffusion

24:04 Speaker 2: the diffuse coefficient excuse me from veers the KD four

24:08 Speaker 2: ninety value. So then that's our uncertainty value on per

24:12 Speaker 2: point level. And then our data product is generated with

24:16 Speaker 2: a variety of parameters. Some of them are listed here

24:19 Speaker 2: on the far right C floor, ellipsoid height, C floor

24:23 Speaker 2: orthometric height, and uncertainty estimates along with the confidence value.

24:30 Speaker 2: And I should also mention for the refraction correction. As

24:34 Speaker 2: you can see up in this top parallelogram, we do

24:38 Speaker 2: have our own refractive index layer that's created from global

24:44 Speaker 2: level temperature and salinity values to create the most accurate

24:49 Speaker 2: correction in time and space. So we kicked off the

24:55 Speaker 2: full production of a TAIL twenty four back in February.

24:59 Speaker 2: It took about thirteen days to produce twenty seven point

25:03 Speaker 2: six terabytes of data. We processed this in AWS.

25:08 Speaker 3: I thought these.

25:09 Speaker 2: Metrics were pretty interesting to say that a total linear

25:12 Speaker 2: coverage of bathymmetry was thirteen point seven million kilometers and

25:17 Speaker 2: then you can see here the number of our symmetry points,

25:21 Speaker 2: which we do follow the ASPRS convention of classification, so

25:25 Speaker 2: ourthymmetry is class forty and our c surface is class

25:28 Speaker 2: forty one. We collected a little over seven billion bathymetry points,

25:34 Speaker 2: and if we just filtered those by our confidence value

25:37 Speaker 2: of point six, then we had more than a half

25:41 Speaker 2: identified as high confidence pathymmetry. And you can see here

25:44 Speaker 2: in the figure the just the coverage in yellow of

25:48 Speaker 2: where we have but symmetry within the product. We chose

25:54 Speaker 2: eight sites that had that we had pretty consistent reference

25:57 Speaker 2: data to compare to to understand what how accurate not

26:02 Speaker 2: the classification confidences, but how accurate the actual bathymmetry elevations were.

26:09 Speaker 2: And you can see here this is on the left column,

26:13 Speaker 2: are the accuracy test results for all of the symmetry

26:17 Speaker 2: points that those eight sites. And you can say we

26:19 Speaker 2: have an RMS of point six eight meters and then

26:25 Speaker 2: if we threshold those values for high confidence on the

26:29 Speaker 2: right plots showing a reduction in RMS of down to

26:34 Speaker 2: forty three centimeters, and so it does improve the overall

26:38 Speaker 2: quality of the symmetry elevations by just using high confidence,

26:43 Speaker 2: but you also have to weigh the impact of removing

26:49 Speaker 2: some of those symmetry points that could very well be bothymmetry,

26:54 Speaker 2: but not all of the algorithm's predictions supported that. One

27:00 Speaker 2: of the things that is I'm sure would be interesting.

27:03 Speaker 2: Are some strengths and limitations to the current version of

27:07 Speaker 2: a TAIL twenty four, And I just want to emphasize

27:10 Speaker 2: that the ATAIL twenty four is a global coverage of

27:14 Speaker 2: the Earth's coastlines, and we have yet to have a

27:18 Speaker 2: space based laser bathymetry system that provides the wealth of

27:24 Speaker 2: data that I set to is providing. I think our

27:28 Speaker 2: approach for signal finding and classification is pretty robust because

27:32 Speaker 2: it can leverage the advantage of each individual algorithm or

27:36 Speaker 2: base algorithm where it works well and we don't lose

27:40 Speaker 2: any quality of the output data and signal finding. It's

27:46 Speaker 2: accessible in a variety of ways. You can download the

27:49 Speaker 2: data from several methods, which sure is part of this training.

27:55 Speaker 2: The architecture of the algorithm is also configured for continual improvement,

28:00 Speaker 2: meaning that we can right now we have six base algorithms,

28:03 Speaker 2: but we could up that to twenty four base algorithms,

28:06 Speaker 2: and it's all set up the structure is set up

28:08 Speaker 2: to continue to add algorithms as methods are identified and implemented.

28:16 Speaker 2: And then also the accuracy of this product is pretty great,

28:20 Speaker 2: and it has been proven suitable for calibration and validation

28:23 Speaker 2: of spectral satellite derived but symmetry in most cases. Some

28:27 Speaker 2: of the limitations, though you can mention here, are that

28:31 Speaker 2: some of the conditions do have results in variable data quality,

28:36 Speaker 2: as you can imagine, and then the current uncertainty estimates

28:39 Speaker 2: are overly optimistic, but we're still working on refining that

28:43 Speaker 2: and making them a little bit more realistic. The data

28:47 Speaker 2: exists only on the coastlines and coincident with the current

28:50 Speaker 2: search masks, so the ATIL twenty four algorithm is not

28:53 Speaker 2: run anywhere except the intersection of the ATL three grain

28:57 Speaker 2: rules and the search mask. And then the photon classification.

29:02 Speaker 2: Conflicts between a TAIL twenty four sea surface and AT

29:06 Speaker 2: eight terrain surface do exist. So if you are interested

29:09 Speaker 2: in seamless you know topobathy profiles along our coastlines and

29:14 Speaker 2: you pull in at eight terrain heights, there's some confliction

29:18 Speaker 2: in the in the photon classification, so it's not there's

29:23 Speaker 2: not that continuity yet in those two along track products,

29:30 Speaker 2: and I wanted to point out a few resources that

29:32 Speaker 2: might be helpful in using a TIL twenty four. Understanding

29:36 Speaker 2: more about a TEL twenty four. The first two bullets

29:40 Speaker 2: are papers that were recently published this year in Earth

29:44 Speaker 2: and Space Science. The first one perish at All. It

29:47 Speaker 2: provides more detail about the accuracy analysis that those eight

29:52 Speaker 2: sites that I was showing earlier. And then the second one,

29:56 Speaker 2: my Magruder at All, is in depth description the base

30:00 Speaker 2: algorithms and the ensemble and talks a little bit more

30:03 Speaker 2: about the metrics associated with understanding the performance of the classification.

30:09 Speaker 2: And then the algorithm. Theoretical basis document is the last bullet,

30:14 Speaker 2: and that gives you even more information about the algorithm.

30:17 Speaker 2: It includes the refraction correction approach. It also includes uncertainty

30:24 Speaker 2: evaluation and all of the other details about the product.

30:30 Speaker 2: And I think that's all I have. I'm so happy

30:33 Speaker 2: to be here today with you, and I will throw

30:35 Speaker 2: it back over to Sean.

30:42 Speaker 1: Thank you Laurie for the excellent overview of I two

30:44 Speaker 1: and the ATL twenty four bathymetry product. We will now

30:48 Speaker 1: transition to Diane Fritz from the National Snow and Ice

30:51 Speaker 1: Data Center to present on ATL twenty four data resources

30:54 Speaker 1: and access. Diane over to you.

31:00 Speaker 4: Hi.

31:01 Speaker 5: I'm Diane and I work at the National Snow and

31:04 Speaker 5: Ice Data Center NSIDC, and I am going to do

31:10 Speaker 5: a quick segment to show you ATL twenty four data,

31:14 Speaker 5: that bathymetry product and the resources and access that we

31:19 Speaker 5: have for you at the DECK or Distributive Active Archives Center.

31:24 Speaker 5: So what is the role of NSIDC. We preserve many,

31:30 Speaker 5: many data sets, including all of the products from i

31:34 Speaker 5: SAT too. Although we are the National Snow and Ice

31:38 Speaker 5: Data Center focused on the criosphere. You'll notice from Professor

31:43 Speaker 5: mcgrider's previous video that she showed I SAT two is

31:49 Speaker 5: a global data set, all of the ones that get

31:53 Speaker 5: derived from that mission, and so we have all of

31:56 Speaker 5: those products at NSIDC. We have a lot of products

32:00 Speaker 5: from other satellite missions, airborne surveys, field observations, all sorts

32:06 Speaker 5: of different things related to the Christ fear and we

32:09 Speaker 5: are one of eleven DAX and you can see on

32:12 Speaker 5: the map here where some of those are located. We

32:16 Speaker 5: are in Boulder, Colorado, but we are accessible all over

32:21 Speaker 5: the globe by emailing us at NSIDC. At NSIDC dot org.

32:29 Speaker 5: Data is a free for anyone. You just need an

32:32 Speaker 5: Earth Data log in to be able to access it.

32:35 Speaker 5: But it's free to create that as well, and it's

32:40 Speaker 5: helpful to be able to work with geotips, net CDF,

32:45 Speaker 5: HDF files, thing other kinds of science files like that.

32:49 Speaker 5: But we have user guides also to help you, and

32:53 Speaker 5: my job and the User Services office at NSIDC is

32:57 Speaker 5: to help anybody use the data that we host at

33:01 Speaker 5: the deck. So this is an example of one of

33:07 Speaker 5: our data set landing pages. In particular, this is the

33:10 Speaker 5: ATL twenty four Bathymetry Product landing page, and I just

33:14 Speaker 5: want to show you some useful things that if you

33:18 Speaker 5: go here, you'll be able to find a user guide

33:22 Speaker 5: that talks a little bit more about what the data is.

33:27 Speaker 5: We have a citation so you can easily give credit

33:31 Speaker 5: to the people that produce this data, but also know

33:34 Speaker 5: who they are and be able to look up the

33:36 Speaker 5: papers and other research that they've done. And we also

33:41 Speaker 5: have a subscribe possibility here, so if you want to

33:46 Speaker 5: know more about a particular data set, you can sign

33:50 Speaker 5: up for email updates of that data. So if there

33:54 Speaker 5: is an update, something that is a new parameter or

33:59 Speaker 5: something like that gets added to a new version, you

34:01 Speaker 5: can get a notification about that. But on our data

34:07 Speaker 5: set landing pages. In addition to these icons at the top,

34:11 Speaker 5: we also have a menu for being able to find

34:15 Speaker 5: things on the side that right menu that you see there,

34:19 Speaker 5: So this gets you access tools, documentation, and help articles.

34:27 Speaker 5: What I have on this screen right here is an

34:30 Speaker 5: example of two different access tools that we have for

34:34 Speaker 5: ATL twenty four. One is the NASA Earth Data Search,

34:38 Speaker 5: which I'm going to do a quick little demo of

34:41 Speaker 5: how you can use that to find ATL twenty four

34:45 Speaker 5: data and actually subset it. There's also slide rule Earth

34:50 Speaker 5: as one of the tools, and many more, and you'll

34:53 Speaker 5: get to hear more about Earth later. We also have documentation,

34:59 Speaker 5: these tech documents if you want to dive further into

35:02 Speaker 5: the algorithms that we're created to get the bathymetry product out,

35:09 Speaker 5: data dictionaries and user guides, and then we also have

35:13 Speaker 5: help articles that are going to help you maybe just

35:18 Speaker 5: understand how to work with a certain kind of platform

35:22 Speaker 5: with HDF files for example, or know a little bit

35:25 Speaker 5: more about the data. So all of these resources are

35:29 Speaker 5: available from that data Set landing page. Now, if I

35:34 Speaker 5: go to the tool NASA Earth data search. I will

35:39 Speaker 5: see some kind of visualization like this, and I'll demonstrate this.

35:45 Speaker 5: You can search for data by mission, so I can

35:51 Speaker 5: put in IAT too and see what's available. I can

35:53 Speaker 5: look for keywords. If I know the data set short

35:57 Speaker 5: name ATL twenty four, I can put that in and

36:01 Speaker 5: look for it. So this is a way to access

36:06 Speaker 5: data from all of the NASA docks, not just NSIDC.

36:12 Speaker 5: There are customization services for select data sets in here

36:16 Speaker 5: as well, and so I'm going to show you what

36:20 Speaker 5: it's like to just pull as small spatial chunk of

36:23 Speaker 5: an ATL twenty four file and visualize that. So this

36:32 Speaker 5: is a picture of panoply. It is a nice kind

36:38 Speaker 5: of visual browse platform and it does some more things

36:42 Speaker 5: as well. But NASA developed this and it's a really

36:45 Speaker 5: nice way to quickly look at a ATL twenty four

36:50 Speaker 5: granule and see what is available in it. As Professor

36:55 Speaker 5: mcgrader mentioned, there are classes of photon classifications in this

37:00 Speaker 5: data set, and that variable is class underscore pH for photon,

37:06 Speaker 5: and forty represents a bathymetry photon and forty one represents

37:12 Speaker 5: those sea surface photons. So on the left here and

37:16 Speaker 5: the small side, I've got a full granule download, and

37:21 Speaker 5: I've zoomed in and changed the scale so that forty

37:26 Speaker 5: the bathymetry photons show up in yellow. And you can

37:30 Speaker 5: see that in here that we have some pretty good

37:34 Speaker 5: returns in the area of the Bahamas for bathymetry. So

37:40 Speaker 5: I'm going to show a little demo using NASA Earth

37:44 Speaker 5: Data Search to grab one of these ATIL twenty four

37:49 Speaker 5: granules near the Bahamas and show you what it's like

37:53 Speaker 5: when we open it in paniply. So I'm going to

37:59 Speaker 5: dive over to my browser now, and this is the

38:05 Speaker 5: dataset landing page for ATL twenty four. The URL is

38:10 Speaker 5: an SIDC dot org slash data slash ATL twenty four.

38:14 Speaker 5: You don't have to type in the rest, the version

38:17 Speaker 5: will self populate. But if I scroll down, you'll see

38:23 Speaker 5: a map that shows that the ATL twenty four coverage

38:26 Speaker 5: is global, and you'll see these different data access tools

38:31 Speaker 5: that you can get slide rules down there. We're going

38:35 Speaker 5: to go to NASA Earth Data Search and click on

38:39 Speaker 5: this and it's going to take us to a page

38:45 Speaker 5: already with ATL twenty four loaded into this. Now I

38:50 Speaker 5: can hover over these different granules. I'm going to click

38:54 Speaker 5: on this one and it'll take me to a place

38:58 Speaker 5: where this one exists. So this is that one that

39:03 Speaker 5: I showed on the slide deck. I can do spatial

39:08 Speaker 5: search here. I can also do temporal search with these controls.

39:14 Speaker 5: I'm going to just draw a little box and grab

39:19 Speaker 5: a little bit of this granule in this upper part.

39:23 Speaker 5: But in doing that spatial search constraint, I still have

39:28 Speaker 5: all of these other granules here as well, so you

39:30 Speaker 5: can see that I've I still have two hundred and

39:33 Speaker 5: twelve that go through this over the course of the mission.

39:37 Speaker 5: I'm for the purpose of this demo, I am just

39:40 Speaker 5: going to copy this granual name and search by it

39:47 Speaker 5: as well and reduce it down to one. But you

39:49 Speaker 5: can use wild cards in here, or you can do

39:52 Speaker 5: a temporal search to narrow it down a little bit more.

39:57 Speaker 5: So with this, I'm going to say I want to

40:02 Speaker 5: download all which is down to my one granule at

40:05 Speaker 5: this point, and it's going to take me to this

40:10 Speaker 5: platform where it's going to give me an option to

40:15 Speaker 5: say do you want to just download this data straight

40:18 Speaker 5: or do you want to customize it? And Harmony is

40:22 Speaker 5: the tool that we have for doing spatial and temporal

40:27 Speaker 5: subsetting on some IAT to data sets. So I'm going

40:32 Speaker 5: to choose a service. The only option here for this

40:36 Speaker 5: data set is the Harmony trajectory subsetter, and once I've

40:42 Speaker 5: selected that, I can scroll down and see that had

40:46 Speaker 5: I had more data sets or a temporal constraint, I

40:52 Speaker 5: could use that, but I still have just that one granule.

40:55 Speaker 5: So I have my spatial subsetting here that looks good

40:59 Speaker 5: to me. I'm going to clear done, go to click

41:03 Speaker 5: download data, and it's going to start putting in that

41:09 Speaker 5: order for me. With that, taking that bounding box and

41:12 Speaker 5: actually trimming my granule just to that area. I can

41:18 Speaker 5: go to my order status and see how Harmony is doing,

41:24 Speaker 5: and it looks like that job it's going to look

41:28 Speaker 5: at that job code it was already successful. This takes

41:32 Speaker 5: a little bit of a lag time to show that,

41:35 Speaker 5: but that's where you can go and get that link directly.

41:40 Speaker 5: It can also notify you when everything is finished processing

41:47 Speaker 5: and you can pull data from here. So I'm going

41:50 Speaker 5: to hop over to panoply. I've already downloaded a subsetted

41:57 Speaker 5: file before this, and this is what it looks like

42:00 Speaker 5: when you open that and you can see the three

42:03 Speaker 5: beams here and when I expand one of those from

42:07 Speaker 5: the This is the ATL twenty four product, and we

42:11 Speaker 5: have our class of photons. We have heights of different

42:17 Speaker 5: kinds and flags and different kinds of information, and then

42:22 Speaker 5: I can plot these and if I'm looking at my

42:26 Speaker 5: class of photons, and again I've changed my scale here

42:31 Speaker 5: in panoply from zero to forty one to thirty nine

42:35 Speaker 5: to forty one, just so I can see that bathymetry

42:39 Speaker 5: photons in the side, so I can see that I've

42:42 Speaker 5: got some good response there. I can also bring up

42:48 Speaker 5: the height data as well. And so if you recalled

42:56 Speaker 5: the way that I grabbed that data, there was a

42:58 Speaker 5: little bit of an island going across here, and you

43:01 Speaker 5: can see some of that height shown in this area.

43:09 Speaker 5: So that's how panoply can help you inspect these HDF

43:14 Speaker 5: files pretty quickly. I'm going back to our data order

43:19 Speaker 5: and see that it's here. I can actually click on

43:23 Speaker 5: this link and get this subseted file directly downloaded from

43:29 Speaker 5: clicking here. But if I have a lot more granules

43:32 Speaker 5: in here, I can use this Earth Data download and

43:37 Speaker 5: it can pull everything in one download for me. This

43:41 Speaker 5: is a different software you need to put on your

43:43 Speaker 5: computer to have that work, but it's really quite simple

43:47 Speaker 5: and we are happy to help out with that. So

43:52 Speaker 5: I have one more slide to show, and that's just

43:58 Speaker 5: some resources that I've provided. This is a link to

44:02 Speaker 5: the ATL twenty four data set landing page, a way

44:06 Speaker 5: to get the Earth Data log in for NASA data

44:09 Speaker 5: in case you don't have that already. Here's a direct

44:13 Speaker 5: link to NASA's Earth Data search for all the decks

44:17 Speaker 5: the Panopolice software from NASA, and we have some help

44:22 Speaker 5: articles the Earth Data Search Guide. This will tell you

44:24 Speaker 5: how to use Harmony subsetting a little bit, but we

44:27 Speaker 5: also have Harmony in as an API as well, and

44:33 Speaker 5: there is a Python library called Harmony Pi that you

44:37 Speaker 5: can use as well if you're doing a programmatic subsetting

44:40 Speaker 5: in a Python notebook for example. And then here's also

44:44 Speaker 5: a link to slide rule Earth. But other people will

44:48 Speaker 5: be talking about that. So I hope this helped you

44:53 Speaker 5: get a good orientation to resources that we have for

44:56 Speaker 5: supporting ATL twenty four and in the User Services Office.

45:01 Speaker 5: We are here to help you, so feel free to

45:04 Speaker 5: email us at NSIDC at NSIDC dot org.

45:13 Speaker 1: Thank you, Diane. Next we'll hear from JP Swinsky, who

45:16 Speaker 1: will present on the slide rule web service for plotting

45:20 Speaker 1: and downloading ATL twenty four data. JP over to you.

45:26 Speaker 3: Okay, thanks for that introduction.

45:29 Speaker 6: I'm going to be talking about slide rule Earth, which

45:32 Speaker 6: is a collaboration between NASA, Goddard and University of Washington.

45:38 Speaker 3: So what is slide rule.

45:40 Speaker 6: Slide rule is a public web service that has rest

45:43 Speaker 6: like APIs for processing science data. The goal is to

45:48 Speaker 6: provide researchers with low latency access to one demand data products,

45:53 Speaker 6: and we provide those data products using parameters supplied at

45:57 Speaker 6: the time of the request, so it's like a real

45:59 Speaker 6: time data access system. The entire system runs in Amazon

46:04 Speaker 6: Web Services US West two and we have access very

46:09 Speaker 6: efficient access to all of NASA's Earth data products stored

46:13 Speaker 6: in US West two like ISA two, Jedi, Lansat, and

46:18 Speaker 6: a growing list of others. As a team, we support

46:22 Speaker 6: multiple clients for accessing slide Rule. The two main clients

46:27 Speaker 6: are a Python web client and our Python client and

46:32 Speaker 6: our web client which I'll be talking about today, but

46:35 Speaker 6: we also support no JS and for those power users,

46:40 Speaker 6: you can access our system with CURL. Here's on the

46:44 Speaker 6: screen you see it just a cartoon of what the

46:46 Speaker 6: system looks like at a very high level, the users

46:51 Speaker 6: typically interact with slide rule using either a Jupiter notebook

46:55 Speaker 6: or their web browser.

46:56 Speaker 3: If they're using the web client.

46:58 Speaker 6: We run a bunch of under the domain slide earth

47:01 Speaker 6: dot io, and then behind those servers we have access

47:05 Speaker 6: to things like NASA's Common Metadata Repository, which is their

47:09 Speaker 6: index for the NASA's Earth data products, as well as

47:13 Speaker 6: the Earth Data Cloud that I mentioned before. You can

47:18 Speaker 6: find us more information about us on our website, which

47:22 Speaker 6: is slide wile worth do I Owe. You can point

47:23 Speaker 6: your browser to that. You're always free to contact me.

47:26 Speaker 6: I have my contact information there and for developers, please

47:31 Speaker 6: look us up on GitHub. We're completely open source under

47:34 Speaker 6: a BSD three clause license.

47:38 Speaker 3: So why did we create slide role.

47:42 Speaker 7: So the.

47:44 Speaker 6: Idea is that we're hopefully improving the way data is

47:47 Speaker 6: accessed in the cloud. So here's just a picture of

47:51 Speaker 6: what it's like to access data in the cloud without

47:55 Speaker 6: a service. So here you have different s three buckets

47:59 Speaker 6: holding different data products like the ATL three data from

48:03 Speaker 6: I Set two, the Jedi data, the Blue Topo data,

48:07 Speaker 6: the Harmonized lands At Sentinel data, and it immediately becomes

48:13 Speaker 6: evident that if you're working from a single Jupiter Hub

48:16 Speaker 6: instance or your laptop, it is very challenging to work

48:20 Speaker 6: with these data sets at a global scale. There's just

48:22 Speaker 6: too much data to work with. You also need to

48:27 Speaker 6: know the format of all these different data sets. For instance,

48:30 Speaker 6: if you're trying to correlate ATLO six ice surfaces to

48:37 Speaker 6: the Arctic dem you need to be able to figure

48:40 Speaker 6: out how to get the acquisition date from the Arctic

48:42 Speaker 6: dem rasters, and that's going to be different than the

48:45 Speaker 6: way you're going to get an acquisition date from HLS,

48:47 Speaker 6: and different then it's going to be from Blue Topo

48:49 Speaker 6: and so forth. And then lastly, there's going to be

48:54 Speaker 6: common operations that you're going to want to do in

48:58 Speaker 6: the data that are going to be very compute intensive.

49:01 Speaker 6: For instance, if you're working with a photon cloud like

49:04 Speaker 6: we are with ATL twenty four, it is very compute

49:07 Speaker 6: intensive to take that photon cloud and resolve it to

49:11 Speaker 6: a single elevation over a certain aggregated area.

49:17 Speaker 3: So that is why.

49:18 Speaker 6: We came up with slide rule as a service for

49:21 Speaker 6: accessing this data. We provide optimized direct access to the

49:26 Speaker 6: raw data. So if you want to just read Atail

49:29 Speaker 6: twenty four or blue topo. You can use our services

49:33 Speaker 6: and we will very efficiently in a parallel way, read

49:37 Speaker 6: the data as fast as possible and give it to

49:40 Speaker 6: you directly. But we also supply efficient subsetting services. So

49:45 Speaker 6: if you want to take the Atail twenty four data

49:49 Speaker 6: and look at it just in an area of interest,

49:55 Speaker 6: you can provide that polygon to slide Rule and it

49:57 Speaker 6: will go out and it'll pull all the different grands

50:00 Speaker 6: and in a parallel way provide them to you in

50:02 Speaker 6: one single response.

50:05 Speaker 3: And then we do on demand processing right next to

50:07 Speaker 3: the data.

50:08 Speaker 6: So we are running in US West two, right next

50:11 Speaker 6: to the buckets that are stored in that same data center,

50:14 Speaker 6: and we can produce customized data products on demand using

50:18 Speaker 6: parameters supplied by your request. So that's in a nutshell

50:24 Speaker 6: what slide rule is. And now I'm going to transition

50:28 Speaker 6: to talk about to give a demonstration of our web client.

50:32 Speaker 6: So this is pretty new. It's it just came out

50:34 Speaker 6: this year, and this is one of the ways, maybe

50:37 Speaker 6: the easiest way to accessing slide Rule, which is doing

50:41 Speaker 6: so from your web browser. So this is the this

50:52 Speaker 6: is the web client. You can get to it going

50:55 Speaker 6: to client, do slide rule or do io, and you

50:59 Speaker 6: can also find links to it. You can also find

51:03 Speaker 6: links to it from our main web page.

51:08 Speaker 3: So when you.

51:08 Speaker 6: Navigate to the web plan, this is the screen that

51:11 Speaker 6: you'll see. And the first thing you're going to want

51:13 Speaker 6: to do for accessing the ATL twenty four Coastal Pathymetry

51:18 Speaker 6: data is you're going to want to go over here

51:20 Speaker 6: and select the ice AD to Coastal Bathymetry selection box.

51:25 Speaker 6: What that does is that puts a search mask up

51:29 Speaker 6: onto overlays it onto the map, showing you.

51:33 Speaker 3: Where there's data available.

51:34 Speaker 6: So everything that you want to do should be within

51:37 Speaker 6: that search mask so that you're sure to get a

51:42 Speaker 6: response back from our service. So I went and did

51:47 Speaker 6: a little poking around and I found a spot that

51:51 Speaker 6: has some good bathymetry off the coast of Cuba. So

51:58 Speaker 6: I just searched for it and zooms me in there,

52:01 Speaker 6: and I know that there's some good pathymmetry in this

52:03 Speaker 6: area here. So what I'm going to do with the

52:05 Speaker 6: I SID two Coastal Pathymetry box selected is I'm going

52:08 Speaker 6: to take the box tool and I'm going to draw

52:12 Speaker 6: just a bounding box over this area and you can

52:16 Speaker 6: see that it automatically shows you the size of the

52:19 Speaker 6: area for requests using the our public service. For a

52:24 Speaker 6: tail twenty four, we ask that it be underneath a

52:29 Speaker 6: less than one thousand square kilometers, so five hundred and

52:32 Speaker 6: fifty some square kilometers is just fine, and you can.

52:35 Speaker 3: See what it is if you hover over it.

52:37 Speaker 6: Once you have the bounding box, you can just click

52:39 Speaker 6: run side rule and it will generate that request for you.

52:44 Speaker 6: Send that request to the servers running in US West two.

52:49 Speaker 6: Those servers will subset all of the data that needs

52:53 Speaker 6: to be subseted, collect it all in a single response,

52:55 Speaker 6: and send it back to the browser that you're running.

53:03 Speaker 6: Once the data is all loaded, it'll pop up this

53:06 Speaker 6: analysis screen on your browser. The first couple times through

53:11 Speaker 6: you're gonna get a lot of helpful advice. I think

53:15 Speaker 6: it's after five requests, we figure that you've read all

53:19 Speaker 6: the helpful advice and you'll no longer see those. If

53:21 Speaker 6: you ever want to reset or get it back, you

53:27 Speaker 6: can always reset the system here and you'll go back

53:30 Speaker 6: to being like a new user. So now we have

53:35 Speaker 6: a display of all the different tracks. You can hover

53:37 Speaker 6: over it to see the information of the different photons returned.

53:43 Speaker 6: These are all the bathymetry photons in that bounding box

53:47 Speaker 6: in ATL twenty four. So you can hover over. You

53:52 Speaker 6: can see the confidence level, the rithometric height, the reference groundtrack,

53:56 Speaker 6: things like that. You can select the different tracks, and

54:01 Speaker 6: on the right you'll see a plot of the along

54:06 Speaker 6: track photon cloud, just the bothymetry photons. So let me

54:12 Speaker 6: just I'll just show a couple of different ones and

54:14 Speaker 6: see what you can see.

54:15 Speaker 3: What they look like.

54:18 Speaker 6: That's a neat one. Okay, well, let's let's choose maybe

54:27 Speaker 6: choose this one here. Let's choose this one here. It

54:29 Speaker 6: goes down to five five meter depth. So now this

54:34 Speaker 6: shows me as zoomed in view of the pathymmetry. But

54:38 Speaker 6: I want to get a better context. But I want

54:39 Speaker 6: to look at the whole photon cloud that it comes from.

54:42 Speaker 6: So if I click this show photon cloud, it actually

54:45 Speaker 6: goes out and makes another request to the server to

54:49 Speaker 6: grab the ATL O three raw photon cloud and combine

54:55 Speaker 6: it with the ATL twenty four photon classifications, and then

54:59 Speaker 6: it gives now this full view of the data. So

55:03 Speaker 6: what we're looking at now is the input data, if

55:07 Speaker 6: you will, to the ATL twenty four algorithms that produce

55:10 Speaker 6: the classifications.

55:11 Speaker 3: That what we saw on the previous plot.

55:13 Speaker 6: And so if we zoom, let's zoom in here. So

55:19 Speaker 6: I just zoomed into a section. You can see the

55:22 Speaker 6: white here is our sea surface. That's the sea surface

55:26 Speaker 6: classified photons. The purple or the unclassified photons. The ATAIL

55:32 Speaker 6: twenty four algorithm is just saying it's neither sea surface.

55:34 Speaker 3: Nor with imagery.

55:36 Speaker 6: And then the green is the photons that the algorithm

55:40 Speaker 6: identified as a bathometric return. Okay, I think one other

55:49 Speaker 6: thing I'll show here is let's go back to this

55:54 Speaker 6: is sometimes it's useful to just get a quick view

55:58 Speaker 6: of what the whole data set looks like, and so

56:00 Speaker 6: we do for that, we do provide a three D

56:04 Speaker 6: view of it.

56:06 Speaker 3: And so.

56:08 Speaker 6: This here is showing what the what the whole data

56:12 Speaker 6: set looks like. And you can see there's some sea

56:15 Speaker 6: floor here at the bottom there.

56:22 Speaker 3: I guess also.

56:26 Speaker 6: We can you have a lot of things at your

56:29 Speaker 6: disposal as far as controlling what the what the plot

56:33 Speaker 6: looks like. So here we are coloring the points based

56:37 Speaker 6: on the arithometric height. If we wanted to color them

56:41 Speaker 6: based on their confidence level. This is now the same plot,

56:46 Speaker 6: but it's showing which photons have a very high confidence

56:50 Speaker 6: in the algorithm of being a mathemetric return versus ones

56:53 Speaker 6: that have a low confidence. So you can see up

56:56 Speaker 6: here we have some really high confident ones, and then

57:00 Speaker 6: right here appropriately so it's a little bit less confident. Lastly,

57:10 Speaker 6: I will just say there is for those that are

57:13 Speaker 6: power users. You can see what the actual request to

57:17 Speaker 6: the server looked like here. Okay, so this flow has

57:23 Speaker 6: so far been just a very simple, straightforward flow of

57:28 Speaker 6: getting access to the ATAIL twenty four data. We do

57:32 Speaker 6: have in the web client some options for some more

57:35 Speaker 6: advanced access, So now go over here, go back to

57:39 Speaker 6: the request and go up here to this selection button

57:44 Speaker 6: and choose advanced. So now we see we're selecting the

57:49 Speaker 6: atail twenty four x API, and this is going to

57:52 Speaker 6: give us access to more features that the server exposes. Specifically,

57:59 Speaker 6: I'm going to go now to this ATL twenty four

58:02 Speaker 6: X fields. Before I do that, I'll draw a bounding box,

58:08 Speaker 6: and I'm gonna make this a much smaller bounding box

58:10 Speaker 6: because we're gonna get some more data. You'll see why. Okay,

58:16 Speaker 6: So I'm gonna go here and now in my request

58:19 Speaker 6: of atail twenty four instead of saying, you know, I

58:23 Speaker 6: don't just want the bathymetry, I want all of it,

58:27 Speaker 6: the unclassified and the sea service, similar to what we

58:29 Speaker 6: saw when we did the ATLO three requests and used

58:31 Speaker 6: the ATL twenty four as a classification, and then.

58:34 Speaker 3: Also there, I don't want just the main fields.

58:39 Speaker 6: I want all of the fields that are in the

58:42 Speaker 6: data product.

58:44 Speaker 3: So here I'm just gonna select.

58:47 Speaker 6: This whether or not this compact selector, and I'm gonna

58:50 Speaker 6: leave it off now because I don't want the compact version,

58:55 Speaker 6: which happens to be the default version when we do

58:57 Speaker 6: just a general request. If you wanted to, you could

59:01 Speaker 6: also do some filters here, like you could filter on

59:03 Speaker 6: the confidence threshold. You could filter on the turbidity in

59:07 Speaker 6: the in whether or not there was valid turbidity or

59:10 Speaker 6: wind speed, and whether or not it was at night

59:14 Speaker 6: and the sensor death was exceeded. You could also even

59:16 Speaker 6: pull in some ancillary fields from the data product.

59:22 Speaker 3: But we'll leave it just like this, and.

59:27 Speaker 6: We'll show the request parameters and you're going to see,

59:29 Speaker 6: now we've got some additional parameters.

59:31 Speaker 3: In our request. We'll make this request.

59:38 Speaker 6: So this is again a smaller area because My guess

59:41 Speaker 6: is it's going to pull in a lot more data

59:43 Speaker 6: because we're asking for all of the photons and not

59:45 Speaker 6: just some of the photons, not just the pothymmetry photons.

59:48 Speaker 3: Yeah, look at that.

59:50 Speaker 6: So you can see this just in this really small area.

59:54 Speaker 6: And I would we would just caution if you want

59:58 Speaker 6: all of the photons beyond just the pathymetry photons, you

1:00:03 Speaker 6: do want a small area not just for the load

1:00:05 Speaker 6: on our surface, but it can very easily overwhelm a

1:00:09 Speaker 6: typical client laptop running this in your browser as well.

1:00:12 Speaker 6: So this very small request had seven hundred and nine

1:00:15 Speaker 6: thousand photons returned.

1:00:17 Speaker 3: But you can see.

1:00:20 Speaker 1: This is.

1:00:22 Speaker 6: A plot here, and then we can do things like now,

1:00:26 Speaker 6: instead of doing that with the metric, right, let's do

1:00:29 Speaker 6: the photon classification. And you can see that over here

1:00:33 Speaker 6: we have these higher classifications, which is sea surface and pathymetry.

1:00:38 Speaker 6: So right away I can look at this and I

1:00:40 Speaker 6: can say, over here's land and over here is your

1:00:44 Speaker 6: sea surface and pathymetry. And if that wasn't super clear

1:00:48 Speaker 6: just by looking at it here, I'm gonna keep my

1:00:51 Speaker 6: pointer here and I want you to look over at

1:00:55 Speaker 6: the left side of the screen where you see the

1:00:59 Speaker 6: dear graphic map and you should see a moving target

1:01:03 Speaker 6: showing which photon you're highlighting on the map, corresponding to

1:01:08 Speaker 6: this elevation plot on the right, And that just confirms that, yeah,

1:01:13 Speaker 6: this is what you're what you're looking at here is

1:01:15 Speaker 6: land and what you're looking at here is the water.

1:01:20 Speaker 6: Now what you can do when you get all of

1:01:23 Speaker 6: the data is you can export it if you want.

1:01:26 Speaker 6: So we can go to our table and this table tab,

1:01:30 Speaker 6: and now we can run an SQL query against the data.

1:01:34 Speaker 6: We can get a full Now all of the raw

1:01:36 Speaker 6: data is being shown, well actually just the first thousands,

1:01:39 Speaker 6: so you can if this is more for power users,

1:01:42 Speaker 6: but it gives you full access to all of the

1:01:45 Speaker 6: data that you're looking at. I should note right now

1:01:49 Speaker 6: that this entire application, everything you're seeing here is running

1:01:52 Speaker 6: in the client. So this is a browser side, client

1:01:54 Speaker 6: side application, single page application. All of the querying, all

1:01:59 Speaker 6: of it is done on your computer, so you own

1:02:02 Speaker 6: the data. This is not being stored on the server anywhere,

1:02:06 Speaker 6: and so we're just giving you tools to be able

1:02:08 Speaker 6: to interact with the data data that is being returned

1:02:11 Speaker 6: by the server. So you make the request of the

1:02:13 Speaker 6: server using this client the server returns the data, and

1:02:17 Speaker 6: then you own that data and can interact with it

1:02:20 Speaker 6: how you please. And if you build a query that

1:02:24 Speaker 6: you really like, you can export the CBS to your computer.

1:02:31 Speaker 6: Or if you want to just export the raw data

1:02:34 Speaker 6: that came back from the server, you can go over

1:02:36 Speaker 6: here and choose either.

1:02:39 Speaker 3: Geopark or CSV as the format. The export.

1:02:43 Speaker 6: Coming soon is LAS that should be out probably mid December.

1:02:49 Speaker 6: Now we support it on the server side, and very

1:02:54 Speaker 6: soon we're going to tie it in on the client

1:02:56 Speaker 6: side as well. Okay, I think with that, I'm done.

1:03:10 Speaker 1: What a great demo on the use of slade roll

1:03:12 Speaker 1: web service for accessing and plotting a TAIL twenty four data.

1:03:15 Speaker 1: Thank you JP. The following summarized the concepts covered in

1:03:21 Speaker 1: Part one of the webinar series. NASA's ice AT two

1:03:25 Speaker 1: carries a photon counting laser altimeter, which is named Atlas.

1:03:29 Speaker 1: Atlas provides near contiguous a long track sampling using six

1:03:33 Speaker 1: individual beams of green light at five hundred and thirty

1:03:36 Speaker 1: two nanimeter wave length, providing high vertical resolution. Every photon

1:03:41 Speaker 1: detected by ATLAS has a latitude, longitude, and elevation associated

1:03:46 Speaker 1: with it. Atlas provides surface specific data products developed by

1:03:50 Speaker 1: experts and include products for land ice, sea ice, the atmosphere,

1:03:55 Speaker 1: vegetation and land oceans and inland water. ATL twenty four

1:04:01 Speaker 1: product provides ATLAS derived coastal bathymetry via ensemble machine learning models,

1:04:06 Speaker 1: where each photon classified as bathymetry will have a confidence

1:04:09 Speaker 1: value assigned. The National Snow and Ice Data Center or

1:04:14 Speaker 1: NSIDC provides access to ATL twenty four data metadata and tools.

1:04:20 Speaker 1: NASA's Earth Data Search is used to find and spatially

1:04:24 Speaker 1: subset ATL twenty four bathymetry data. Slide roll is a

1:04:29 Speaker 1: public web service with low latency access to on demand

1:04:32 Speaker 1: data products stored in S three and slide roll provides

1:04:36 Speaker 1: on demand processing next to the data for generating customized

1:04:40 Speaker 1: data products using parameter supplied in the user's request. Looking

1:04:47 Speaker 1: ahead to Part two of the webinar series, we will

1:04:49 Speaker 1: be covering the following topics. How to generate satellite derived

1:04:53 Speaker 1: bathymetry using methodologies combining ATL twenty four bathymetry with optical

1:04:58 Speaker 1: satellite data from Sentinel TIS, two Python workflows for filtering, visualizing,

1:05:03 Speaker 1: and analyzing bathymetric point clouds, Integration of ATL twenty four

1:05:08 Speaker 1: into existing Noah workflows for nautical charting and coastal zone management.

1:05:14 Speaker 1: An overview of Noah's satbathy tool and its capabilities for

1:05:18 Speaker 1: automated satellite drive bethymmetry generation using ATL twenty four. Before

1:05:25 Speaker 1: we transition to the question and answer session, I want

1:05:28 Speaker 1: to remind you there will be one homework assignment which

1:05:30 Speaker 1: you will be able to access from the training page

1:05:32 Speaker 1: on December fourth. Answers must be submitted by Google Form

1:05:36 Speaker 1: with the due date of December thirty first. To receive

1:05:40 Speaker 1: a certificate of completion, you must attend all three live

1:05:43 Speaker 1: webinars and complete the homework assignment by the deadline. You

1:05:47 Speaker 1: receive a certificate via email approximately two months after completion

1:05:51 Speaker 1: of the course. We want to thank once more doctor

1:05:56 Speaker 1: Lauri Magruder from the University of Texas at Austin, Doctor

1:05:59 Speaker 1: Diane Fritz from the National Snow and Ice Data Center,

1:06:02 Speaker 1: and JP Swinsky from NASA Goddard Space Flight Center for

1:06:06 Speaker 1: their presentations and instructive demos. Below is the contact information

1:06:11 Speaker 1: for Lori, Diane and Jp, along with links to the

1:06:14 Speaker 1: ourset website and social media. If you enjoyed today's webinar,

1:06:18 Speaker 1: we hope you will sign up on the arset list,

1:06:20 Speaker 1: serve to receive notifications of future trainings, and follow us

1:06:24 Speaker 1: on social media for future relevant announcements pertaining to NASA's

1:06:27 Speaker 1: Earth Sciences. Below our list of resources relevant to the

1:06:33 Speaker 1: material covered in part one of the training. We will

1:06:39 Speaker 1: now transition to the question and answer portion of today's training.

1:06:48 Speaker 8: Great, and thank you for everybody it's been submitting your questions.

1:06:51 Speaker 8: We've got a lot of good ones so far, so

1:06:54 Speaker 8: jumping right into it. Question number one, can we develop

1:06:57 Speaker 8: watershed models using the data from the I st to

1:07:00 Speaker 8: imagery what would be the minimum resolution?

1:07:07 Speaker 5: I'll go ahead and take that. I'm not one hundred

1:07:11 Speaker 5: percent sure about the kind of model that is wanting

1:07:16 Speaker 5: to be developed, but I think hopefully the presentation was

1:07:20 Speaker 5: good at describing the data format. So we have a

1:07:25 Speaker 5: long track photon returns, and those tracks, we have six

1:07:31 Speaker 5: beams that are coming three pairs, and the pairs are

1:07:37 Speaker 5: ninety meters apart from each other, with three kilometers separating

1:07:41 Speaker 5: each beam pair. But the along track resolution is quite

1:07:44 Speaker 5: good at around seventy centimeters. And so it's really the

1:07:50 Speaker 5: elevation data that you get from IAT too, regardless of

1:07:54 Speaker 5: which product that you're looking at is really good validation

1:07:58 Speaker 5: data at a long that track, So I hope that

1:08:04 Speaker 5: helps answer that question.

1:08:10 Speaker 8: Hey, Diane, thank you. Question number two, can we derive

1:08:14 Speaker 8: data for different time periods to do comparisons? It would

1:08:18 Speaker 8: be useful for cebid mobility assessments.

1:08:23 Speaker 5: This was me as well, So a TAIL twenty four

1:08:27 Speaker 5: as many of the other data products are derived from

1:08:31 Speaker 5: the main ATLO three photons and those have been gathered

1:08:37 Speaker 5: since twenty eighteen, and you can separate out those files

1:08:42 Speaker 5: with temporal queries and pick what you're looking at at

1:08:46 Speaker 5: different time periods. Of course, the return time of the

1:08:52 Speaker 5: I SAT to satellite is ninety one days, so between

1:08:58 Speaker 5: that and getting good data, your particular location may limit

1:09:03 Speaker 5: some of what you're trying to do with those comparisons,

1:09:06 Speaker 5: but those are the parameters for it.

1:09:09 Speaker 8: Awesome, Thank you, Diane. Question three is the ATL twenty

1:09:12 Speaker 8: four product already corrected for tides or is this the

1:09:15 Speaker 8: post processing step we have to consider, especially for locations

1:09:19 Speaker 8: where tidal range is considerably wide.

1:09:22 Speaker 4: This is Chris, I'm as speaker on the Thursday session,

1:09:26 Speaker 4: but I can answer this if nobody else wants to

1:09:29 Speaker 4: jump in. We actually don't need an explicit Todd correction

1:09:34 Speaker 4: step in the ATAIL twenty four workflow because we use

1:09:37 Speaker 4: a process similar to what's called allipsoidlarly reference serveang and

1:09:40 Speaker 4: hydrosurveying world. Basically, it means that our data, our heights are,

1:09:45 Speaker 4: our depths are not on the instantaneous water surface the

1:09:49 Speaker 4: relative to a vertical data at every step in the process. See,

1:09:52 Speaker 4: they're an orthometric data more an ollipsoidal datum. And for

1:09:55 Speaker 4: people that want to transform to a tidal EDWM like

1:09:59 Speaker 4: mean low Er La water in the US, at least

1:10:01 Speaker 4: we can use a tool called Noah's v d ADAM

1:10:04 Speaker 4: to be able to do vertical datum transformations. So basically

1:10:07 Speaker 4: we don't we don't actually need an explicit TIPES step,

1:10:11 Speaker 4: but we leave it to the user to be able

1:10:12 Speaker 4: to transform to different vertical datams, including title datams, if

1:10:16 Speaker 4: they want to do.

1:10:17 Speaker 8: So. Great, Chris, thank you so much. Question number four,

1:10:24 Speaker 8: that's a very clear scatter plot. Could you tell me

1:10:27 Speaker 8: about the steps taken to check for overfitting?

1:10:32 Speaker 7: I can answer that if no one else is. This

1:10:36 Speaker 7: is Jeff Perry. I'm on, UT's a TIL twenty fourteen,

1:10:40 Speaker 7: and I kind of wrote this assuming I understood the

1:10:44 Speaker 7: question correctly. I don't actually have the plot in front

1:10:48 Speaker 7: of me, but the individual algorithm should have wrote this first.

1:10:52 Speaker 7: The individual algorithms may or may not overfit depending on

1:10:55 Speaker 7: their native They're all different, right, they all are very

1:10:58 Speaker 7: different algorithms. The final model, the ensemble is we use

1:11:04 Speaker 7: xty boost just because it's fast, for no other reason.

1:11:09 Speaker 7: It is prone to overfitting. We can we can fix

1:11:12 Speaker 7: that by adjusting the hyper parameters of the model. But

1:11:16 Speaker 7: after all is said and done, a lot of the

1:11:19 Speaker 7: overfitting we correct by applying this blunder detection. That's, like,

1:11:23 Speaker 7: you know, really obvious errors that some of the models

1:11:27 Speaker 7: can make, like they can some of them can put

1:11:29 Speaker 7: bithymmetry on top of sea surface and obviously you can't

1:11:32 Speaker 7: have that. So that fixes a lot of those overfitting errors.

1:11:38 Speaker 7: I hope I answered that sufficiently.

1:11:43 Speaker 8: Jeff, thank you so much. I really appreciate that. Question

1:11:46 Speaker 8: Number five, you're welcome. How much work will it take

1:11:49 Speaker 8: to extend this approach to inland waters? Is it possible?

1:11:53 Speaker 8: And is any team working on such a product? What

1:11:56 Speaker 8: about in northern latitude lakes and ponds?

1:12:00 Speaker 4: This is Chris again. I can answer this one if

1:12:03 Speaker 4: no one else wants to we in developing ATL twenty

1:12:06 Speaker 4: four were very purposefully did not consider inland waters because

1:12:09 Speaker 4: there's already a dedicated data product for inland waters, which

1:12:12 Speaker 4: is ANTIL thirteen, and ATL thirteen does also include agathymmetry attribute.

1:12:22 Speaker 8: Terrifect. Thank you so much, Chris, and hopefully whoever asked

1:12:25 Speaker 8: that question can go and search for that ATL thirteen product.

1:12:29 Speaker 8: Question six, what applicability does the STATEA have for bethymetry

1:12:33 Speaker 8: mapping of inland and near coastal lakes?

1:12:39 Speaker 4: Is that the sorry, that one similar similar to the

1:12:43 Speaker 4: it's similar to the previous question right inland waters again,

1:12:46 Speaker 4: that's at ATL thirteen.

1:12:49 Speaker 8: Is some inland waters? Okay, great, thank you Chris. Question

1:12:56 Speaker 8: number seven, our positive depth biases ATL twenty four common

1:13:01 Speaker 8: and what can be the cause?

1:13:05 Speaker 4: I can maybe take that one too, unless Jeff for

1:13:07 Speaker 4: somebody else wants to go ahead. I think that's still

1:13:10 Speaker 4: to be determined. I think that's a great case where

1:13:14 Speaker 4: we'd like to hear from users of the data as

1:13:16 Speaker 4: they do comparisons against independent reference data if they're seeing

1:13:21 Speaker 4: positive depth biases and that accuracy test that was in

1:13:25 Speaker 4: Laurie mcgruder's presentation, we actually didn't see a bias being

1:13:33 Speaker 4: consistently in one direction or another. I think we saw

1:13:35 Speaker 4: biases of around ten centimeters that were that could either

1:13:38 Speaker 4: be positive or negative. There are some reasons why we

1:13:43 Speaker 4: might expect a bias. For example, one of the things

1:13:47 Speaker 4: that are ATL twenty fourteen has started talking about recently

1:13:51 Speaker 4: is what's known as forward scattering bias and mathematric ladder.

1:13:56 Speaker 4: But I think we just haven't. I mean, part of

1:13:59 Speaker 4: the importance of ATL twenty four and I I sat

1:14:02 Speaker 4: it to pythymmetry in general, is that there's just so

1:14:04 Speaker 4: many areas in the world where we don't have really

1:14:07 Speaker 4: good existing pathymmetry, and so that's kind of a double

1:14:12 Speaker 4: edged sword. It makes the data products quite valuable, but

1:14:15 Speaker 4: it also means that there's in many places in the

1:14:18 Speaker 4: world not a lot of existing data to compare to.

1:14:23 Speaker 8: Great Chris, Thank you so much. Question number eight. Does

1:14:27 Speaker 8: the slide Real web client apply any de noising or

1:14:30 Speaker 8: classification algorithms before presenting ATL twenty four bythymmetry points or

1:14:35 Speaker 8: are we viewing the raw ATL twenty four photon returns.

1:14:39 Speaker 3: This is JP.

1:14:42 Speaker 6: What was shown when we show it through the accessing

1:14:45 Speaker 6: it through the web client it's it's just the raw pathymmetry.

1:14:50 Speaker 3: In the general tab.

1:14:52 Speaker 6: We do by default filter any low confidence pthymmetry just

1:14:58 Speaker 6: to keep requests as fast as possible, but then allow

1:15:02 Speaker 6: the user to drill down and raise the threshold if

1:15:07 Speaker 6: they want now in a couple of days. Well, there's

1:15:13 Speaker 6: other APIs where you can apply different processing steps to

1:15:18 Speaker 6: the data being returned. Those steps aren't available in the

1:15:23 Speaker 6: web client, but are available through the Python client. Also,

1:15:28 Speaker 6: I guess, just going back a couple questions for those

1:15:31 Speaker 6: interested in ATL thirteen the inland bodies of water, there

1:15:35 Speaker 6: is also Cydebrugle also supports accessing that data from the

1:15:39 Speaker 6: web browser as well. Instead of the Coastal bothymetry tab,

1:15:44 Speaker 6: you just select the Inland bodies of Water tab and

1:15:47 Speaker 6: you'll get the ATL thirteen data product.

1:15:51 Speaker 8: Great JP, thanks so much. Question number nine, how does

1:15:54 Speaker 8: ice have to you distinguish between photon returns from the

1:15:57 Speaker 8: sea surface and photon returns from the seafloor in shallow

1:16:01 Speaker 8: coastal waters, especially considering issues like photon noise and surface

1:16:05 Speaker 8: wave variability.

1:16:11 Speaker 7: This is Jeff. Again, Matt had answered this, but he's

1:16:15 Speaker 7: having technical difficulties. Otherwise he would probably be speaking up.

1:16:23 Speaker 7: But as you can see in his answer there, Sorry

1:16:26 Speaker 7: now I'm just kind of reading it. But the question is,

1:16:31 Speaker 7: how does it distinguish between photon returns and shallow coastal waters.

1:16:38 Speaker 7: I mean, like he says, we train on a lot

1:16:41 Speaker 7: and a lot of data, and sometimes in really you know,

1:16:44 Speaker 7: high waves, we do make mistakes in the trough of

1:16:48 Speaker 7: a wave can be classified as seafloor. In release we've

1:16:55 Speaker 7: had so what we've been calling version one of the algorithms.

1:16:59 Speaker 7: We've had two releases of this, and the second release

1:17:03 Speaker 7: fixes a lot of those. It goes back and looks,

1:17:08 Speaker 7: especially for those places where this trough of the wave

1:17:12 Speaker 7: is misclassified as bethymetry and fixes them. So that's one

1:17:16 Speaker 7: way we deal with that.

1:17:22 Speaker 8: Okay, Jeff, thank you so much for chiming in, and

1:17:25 Speaker 8: we apologize that Matt is not able to access or

1:17:28 Speaker 8: unmute himself. But moving on question number ten, how does

1:17:32 Speaker 8: the ATAIL twenty four handle the challenge of separating true

1:17:36 Speaker 8: seafloor returns from water column backscatter in shallow coastal zones

1:17:40 Speaker 8: where turbidity is high.

1:17:44 Speaker 7: Hia, it's Jeff again, So I mean, obviously, if there's

1:17:48 Speaker 7: too much turbidity and the signal disappears. There's nothing you

1:17:53 Speaker 7: can do. You know, at some point that's going to happen, right,

1:17:56 Speaker 7: And if the noise is too high, well there's you know,

1:17:59 Speaker 7: if you lose the signal, you lose a signal. But uh,

1:18:03 Speaker 7: you know, again, we have an ensemble here, multiple algorithms,

1:18:07 Speaker 7: and so for example, the algorithm costnet. It operates on

1:18:12 Speaker 7: high uh you know, imagery with high vertical resolution. You

1:18:16 Speaker 7: can very clearly separate the two signals if there's not

1:18:19 Speaker 7: too much sturbidity, and so it separates those out. And again,

1:18:24 Speaker 7: as we keep saying, the ensemble has an opportunity if

1:18:28 Speaker 7: some algorithm misses it or confuses the two signals, or

1:18:31 Speaker 7: makes assumptions about the signal. Like I mentioned in the beginning,

1:18:35 Speaker 7: some algorithms don't don't won't even predict bathymetry if the

1:18:39 Speaker 7: bathymetry signal is stronger than the sea surface. But the

1:18:44 Speaker 7: ensemble and the blunder detection has a chance to go

1:18:47 Speaker 7: back and fix all that. It doesn't always fix it,

1:18:50 Speaker 7: but but it has that opportunity.

1:18:55 Speaker 8: Great, Jeff, thank you so much. Question number eleven is

1:18:58 Speaker 8: the symmetry available for large room systems like the Mississippi

1:19:01 Speaker 8: River or are there too many issues with turbidity or

1:19:04 Speaker 8: orbittle path shoreline geometry.

1:19:13 Speaker 4: I think that's again similar to the previous questions on

1:19:16 Speaker 4: inland waters, where that would be the A Tail thirteen

1:19:19 Speaker 4: product and not ATL twenty four, which is coastal focused.

1:19:25 Speaker 8: All right, Chris, thank you. Question twelve in slide role,

1:19:30 Speaker 8: is the data derived per line only so the spatial

1:19:32 Speaker 8: variability is poor?

1:19:36 Speaker 6: Yeah, So I think what you're asking is all of

1:19:40 Speaker 6: the data products right now that we generate or provide

1:19:43 Speaker 6: services for a long track, and they are. We would

1:19:48 Speaker 6: love to have produce image data raster data sets. We

1:19:54 Speaker 6: hope to in the future, but right now we're funded

1:20:00 Speaker 6: only to support the along track products.

1:20:06 Speaker 8: All right, thanks GP. Question thirteen does the Python library

1:20:10 Speaker 8: allow for larger bounding boxes?

1:20:17 Speaker 5: So this this is just limited by the retrieval of

1:20:23 Speaker 5: the granules from the cloud hosted data. So you can

1:20:27 Speaker 5: make larger bounding boxes, but your your code may stall

1:20:32 Speaker 5: a little bit. But I put in a reference here

1:20:36 Speaker 5: for being able to look up Harmony and look into

1:20:40 Speaker 5: the Harmony pie library and see. So I to be honest,

1:20:45 Speaker 5: I haven't done a very large grab myself, so I

1:20:49 Speaker 5: don't feel one hundred percent confidence about this answer.

1:20:56 Speaker 6: And for side rule, it's really easy, and I'm guessing

1:21:04 Speaker 6: the harmony it would have the same fundamental problem. It's

1:21:10 Speaker 6: it's really easy when you're looking at a globe to

1:21:13 Speaker 6: not realize how.

1:21:14 Speaker 3: Big of a request you're making.

1:21:18 Speaker 6: I don't know if you remember when I was drawn

1:21:21 Speaker 6: that really small box in the demo, and that had

1:21:24 Speaker 6: seven hundred thousand points in it. You can imagine a

1:21:28 Speaker 6: factor of one hundred, factor of a thousand on that easily.

1:21:33 Speaker 6: You know, you're easily talking a billion rows and and

1:21:37 Speaker 6: so with these photon photon cloud data sets, it's very

1:21:42 Speaker 6: easy to get huge, huge returns. So all that being said,

1:21:48 Speaker 6: on side rule, we let we let you do it,

1:21:52 Speaker 6: and and we'll do the best we can. But it's

1:21:58 Speaker 6: typically the client crashes before the servers do. That's just

1:22:01 Speaker 6: what we notice.

1:22:04 Speaker 5: Yeah, So, as JP is saying, these data sets are massive,

1:22:10 Speaker 5: so really kind of doing some pre look of you know,

1:22:14 Speaker 5: what is your temporal subset that you might want or

1:22:16 Speaker 5: your spatial subset and really restricting it to what you

1:22:20 Speaker 5: truly need is the best way.

1:22:23 Speaker 8: Yeah, thank you, Diane and GP. Question fourteen is turbidity

1:22:31 Speaker 8: the main thing affecting the confidence level? Does that come

1:22:33 Speaker 8: from veers?

1:22:42 Speaker 7: I'm not sure who who put this answer in? So

1:22:45 Speaker 7: I can answer if no one else's the I mean yes,

1:22:50 Speaker 7: the answer the first question is yes, turbidity and background

1:22:53 Speaker 7: noise is what's affecting the confidence level. But it's implicit.

1:22:57 Speaker 7: It's not we don't measure the turbidity and then measure

1:23:02 Speaker 7: the background noise and then make assumptions, you know, make

1:23:05 Speaker 7: predictions based upon that. It's just implicit in most of

1:23:09 Speaker 7: the algorithms. When there's more turbidity and more background noise,

1:23:13 Speaker 7: it's just going to have a harder time, uh, deciding

1:23:16 Speaker 7: what you know, what the label, what, what classification of

1:23:20 Speaker 7: photon is and and so uh. And then down in

1:23:24 Speaker 7: the answer the X you boost model. It's a tree

1:23:27 Speaker 7: based model, and so it uses that tree structure to

1:23:31 Speaker 7: determine what the what the competence level is. It's like

1:23:36 Speaker 7: a probability. It's a soft max if you know what

1:23:38 Speaker 7: that is. But uh, it's it's not strictly speaking, it's

1:23:42 Speaker 7: not a probability, but it looks like a probability and

1:23:44 Speaker 7: acts like a probably. So so that's where that comes from.

1:23:49 Speaker 7: Is it's it's part of the UH decision tree model.

1:23:54 Speaker 7: UH will will for every photon, assign probabilities or SoftMAC

1:24:00 Speaker 7: probabilities to each label and then choose the most probable one.

1:24:07 Speaker 8: Thank you, Jeff. Question fifteen, can we use it for

1:24:10 Speaker 8: detecting plate movement of tidal areas to get information about

1:24:14 Speaker 8: upcoming earthquakes.

1:24:26 Speaker 5: I'm just going to unmute and say that this is

1:24:29 Speaker 5: a really interesting application idea. Chris, if you please comment,

1:24:35 Speaker 5: But I think that this is beyond the production of

1:24:39 Speaker 5: ATL twenty four data. But it's great to have people

1:24:44 Speaker 5: thinking this way and thinking about applications for this data.

1:24:49 Speaker 4: I was going to say the exact same thing. It's

1:24:51 Speaker 4: really interesting, but beyond the scope of what we've looked at.

1:24:53 Speaker 4: I don't know if the special resolution and accuracy support that. Yeah,

1:24:58 Speaker 4: if anybody looks into that, i'd I'm very curious to

1:25:03 Speaker 4: hear what they find.

1:25:06 Speaker 8: Diane and Chris, thank you. Question sixteen I think has

1:25:10 Speaker 8: been addressed in previous questions. So for those that are

1:25:13 Speaker 8: interested in inland athymmetry, definitely check out the ATL thirteen product.

1:25:18 Speaker 8: Question seventeen. In satellite arrived bathymmetry workflows that use ATL

1:25:23 Speaker 8: twenty four as reference data for training models on sentinel

1:25:26 Speaker 8: to imagery. Are tubidity and bottom type limitations generally encountered

1:25:30 Speaker 8: first in the multi spectral reflectance rather than in the

1:25:34 Speaker 8: ATL twenty four lighter signal.

1:25:40 Speaker 4: I was just reading.

1:25:43 Speaker 8: This question.

1:25:44 Speaker 4: Oh, somebody scrolled and I'm not sure which one was

1:25:47 Speaker 4: this again seventeen in the Thursday sessions this week, and

1:25:54 Speaker 4: the presentations that both I give and gretch Nami from Noah,

1:25:57 Speaker 4: We're gonna We're gonna talk about using ATL twenty four

1:26:02 Speaker 4: in combination with satellite drive with imetry from multi spectral imagery,

1:26:09 Speaker 4: So hopefully those presentations will help answer this. Let me

1:26:13 Speaker 4: read this question again. Though our turbidity and bottom type

1:26:16 Speaker 4: limitations generally encountered first in the multi spectral reflectance. They're

1:26:24 Speaker 4: sort of independent workflows, but some of the constraints are

1:26:31 Speaker 4: the same. So obviously, in any method of satellite drive

1:26:34 Speaker 4: with imagery relying on spectral data and relying on having

1:26:39 Speaker 4: returns from the sea floor, turbidity is one of the

1:26:43 Speaker 4: main limiting factors, and that's the case also with with

1:26:47 Speaker 4: IAT with imetry. Generally the ranges in which they both

1:26:52 Speaker 4: work are are pretty similar, which makes ATL twenty four

1:26:57 Speaker 4: a pretty good choice for calibration and validation or just

1:27:01 Speaker 4: integration with special drift immetry. I hope that answers the question,

1:27:07 Speaker 4: and I hope that the talks that Gretchen and I

1:27:09 Speaker 4: give on Thursday will also help answer that.

1:27:14 Speaker 8: Made Chris thank you so much. Question eighteen. Are their

1:27:18 Speaker 8: tools available for accessing the ATL thirteen products such as

1:27:21 Speaker 8: slide rule and Diane thank you so much for adding

1:27:23 Speaker 8: this link so that they can access that data set.

1:27:25 Speaker 8: Question nineteen. Is there a gridded product available or planned

1:27:29 Speaker 8: for the future for the US East West? It's covered

1:27:34 Speaker 8: by coasts? Thank you.

1:27:45 Speaker 4: I don't think our team has explicitly talked about doing

1:27:48 Speaker 4: a gridded product. From the beginning, the intent was for

1:27:53 Speaker 4: ATIL twenty four to be in a long track product.

1:27:57 Speaker 4: I think that certainly they're their tools and options that

1:28:02 Speaker 4: will help end users create credited products in some areas.

1:28:08 Speaker 4: I think we're just starting to see enough coverage from

1:28:11 Speaker 4: all of the different i ST two overpasses that we

1:28:13 Speaker 4: maybe have enough data in places to do like a

1:28:16 Speaker 4: gridded bathmetric DM solely from i ST to data, But

1:28:22 Speaker 4: in other areas where people want a gridded product kind

1:28:26 Speaker 4: of like wall to wall coverage. Those are cases where

1:28:29 Speaker 4: the integration with specially drived withthymetry from multi spectral imagery

1:28:34 Speaker 4: can be very valuable.

1:28:38 Speaker 8: You thank you, Chris. Question twenty what changes are expected

1:28:42 Speaker 8: at the confidence levels next year? It looks like Matt

1:28:58 Speaker 8: might have answered this one, and unfortunately that does not

1:29:00 Speaker 8: have access to this WebEx but he did provide an

1:29:05 Speaker 8: answer which I will read. But I apologize because I'm

1:29:08 Speaker 8: not the one that contributed this answer. But we are

1:29:11 Speaker 8: actively improving the constituent models and investigating new models to

1:29:14 Speaker 8: add into the mix. The confidence levels are entirely dependent

1:29:17 Speaker 8: on capabilities of the constituent models. After building the ensemble

1:29:21 Speaker 8: and prior to publishing the next version of ATL twenty four,

1:29:24 Speaker 8: we will conduct a study of confidence values relative to

1:29:27 Speaker 8: reference with symmetry data sets to determine the updated best

1:29:30 Speaker 8: practice value to use. Thank you Matt for adding that

1:29:35 Speaker 8: and for answering that question. Question number twenty one. How

1:29:38 Speaker 8: critical is viers KT four ninety to the uncertainty assessment?

1:29:43 Speaker 8: I would think clouds and land adjacency would greatly reduce

1:29:46 Speaker 8: the quantity of KD four ninety retrievals and quality of

1:29:49 Speaker 8: these retrievables would be reduced in optically optically shallow waters

1:29:53 Speaker 8: unless alterations to the standard KD four ninety algorithm are

1:29:57 Speaker 8: used to minimize benthic contamination.

1:30:05 Speaker 4: I'm not sure to type that answer, but I agree

1:30:07 Speaker 4: with what's given. There's the answer. The Deer's Katie four

1:30:11 Speaker 4: ninety is not directly used in the UH the deeporate retrieval.

1:30:18 Speaker 4: It is the only place where that's currently used is

1:30:20 Speaker 4: in the uncertainty calculation. And as it says an answer

1:30:24 Speaker 4: there that that's used in data that's that goes into

1:30:27 Speaker 4: the uncertainty look up tables. Yeah, hopefully that that answers it.

1:30:36 Speaker 8: And Chris, I don't know if you can answer this

1:30:37 Speaker 8: next one as well, but why not use Katie four

1:30:39 Speaker 8: ninety in a green band example? Given the viewers five

1:30:43 Speaker 8: hundred fifty five.

1:30:46 Speaker 4: The the global Katie diffuse attenuation coefficient of downwelling radiance.

1:30:54 Speaker 4: That's what Katie represents. The global data products that we're

1:30:57 Speaker 4: aware of and work with are generally provided as Katie

1:31:01 Speaker 4: four ninety, meaning diffuse attenuation coefficient at four hundred and

1:31:06 Speaker 4: ninety nanimeters. It would be better to use one closer

1:31:10 Speaker 4: to five hundred and thirty two nanometers, which is the

1:31:12 Speaker 4: wave length at which I SAT two's ladder system operates.

1:31:19 Speaker 4: What we do is we actually provide a conversion from

1:31:24 Speaker 4: KT four ninety to KTIE five thirty two. And that's

1:31:29 Speaker 4: that's based on a drived empirical linear relationship between KTI

1:31:34 Speaker 4: four ninety and KT five thirty two. That said, and

1:31:39 Speaker 4: I'm not aware of it, but if there is a

1:31:40 Speaker 4: global readily available vers KDE or Vere's KD five point

1:31:45 Speaker 4: fifty five, the question as a good one that probably

1:31:49 Speaker 4: would be a better product to use.

1:31:54 Speaker 8: Thank you, Chris. Question twenty three, can I get access

1:31:57 Speaker 8: to the data for Nigeria?

1:32:05 Speaker 5: It is a global data set, so along the coast there,

1:32:10 Speaker 5: I don't know what the mask actually looks like in

1:32:14 Speaker 5: that space. So Chris, maybe at the top of your

1:32:16 Speaker 5: mind you might have a better sense. But yes, what

1:32:20 Speaker 5: is available from finding mathymetry from ATLO three to ATL

1:32:25 Speaker 5: twenty four should be in that area.

1:32:31 Speaker 8: All right, allan thank you. Question twenty four could this

1:32:34 Speaker 8: be used to map objects in your coasts like vessels, booies, jetties,

1:32:37 Speaker 8: et cetera.

1:32:43 Speaker 5: I wrote, what is there? But Chris, you will definitely

1:32:48 Speaker 5: have a better sense of this because depending on the

1:32:52 Speaker 5: size of the sunken vessel things like that, and the

1:32:54 Speaker 5: clarity of the water being able to see that in

1:32:58 Speaker 5: the seafloor. Not sure if if anybody has come across

1:33:02 Speaker 5: that in the tests, but possible if if the there

1:33:06 Speaker 5: isn't any turbidity in the water. I would think, Chris,

1:33:15 Speaker 5: I think you're still muted.

1:33:19 Speaker 4: I was just saying, I agree with Diana. It's it's

1:33:21 Speaker 4: really just a function of the of the resolution, and

1:33:23 Speaker 4: I think that one was addressed already. But basically, in

1:33:26 Speaker 4: the along track, along any individual beam track, the resolution

1:33:31 Speaker 4: is approximately seventy centimeters, although it also depends on the

1:33:35 Speaker 4: posymetry cass, it depends on water water clarity, and then

1:33:39 Speaker 4: it's coarser cross tracks passing. So just you know, is

1:33:43 Speaker 4: the size of an object is it? Is it detectable

1:33:46 Speaker 4: at that special resolution at least for things like jetties

1:33:50 Speaker 4: or you know, large you know, coastal engineering structures. Uh,

1:33:55 Speaker 4: if if you happen to get a track line over it,

1:33:58 Speaker 4: you will, you'll you'll most definitely say it. And I've

1:34:00 Speaker 4: seen cases where you can get with immetry right up

1:34:03 Speaker 4: to say a jetty and then and then you'll see

1:34:06 Speaker 4: the see the surface elevations going over the top of

1:34:12 Speaker 4: the jetty.

1:34:15 Speaker 8: Hey Diana, Chris, thank you. A last question twenty five.

1:34:19 Speaker 8: What happens in the presence of algae blooms or sea grass.

1:34:27 Speaker 4: That's a great question, and I think the answer is

1:34:30 Speaker 4: it just it just depends. You know, when you've got

1:34:32 Speaker 4: algae bloms, that's obviously impacting the serbidity, impacting the ability

1:34:37 Speaker 4: of that laser light to penetrate through the water column

1:34:39 Speaker 4: and to get a reflect a detectible reflection off the

1:34:42 Speaker 4: sea floor that makes it back up to the to

1:34:44 Speaker 4: the sensor. And then it's kind of a similar thing

1:34:48 Speaker 4: with with UH with sea grass, UH that can do

1:34:53 Speaker 4: a couple of different things that it can obscure the

1:34:55 Speaker 4: sea floor. It also seagrass tends to have relatively low

1:35:00 Speaker 4: reflectance at five hundred and thirty ten animeters, so it

1:35:03 Speaker 4: can make it difficult to get returns. But more commonly,

1:35:07 Speaker 4: what I've seen in areas for example, you know, we've

1:35:11 Speaker 4: looked at some data around Saint Croix where there might

1:35:17 Speaker 4: be a particular track that goes over a seagrass bed,

1:35:20 Speaker 4: and you know, sometimes we have seen what looks like

1:35:24 Speaker 4: good withthymmetry in those areas. I don't think we've done

1:35:28 Speaker 4: an accuracy assessment specifically over sea grass, so you know,

1:35:31 Speaker 4: how much does that impact the accuracy. I would say

1:35:34 Speaker 4: kind of unknown at this point and probably depends a

1:35:38 Speaker 4: lot on the density of the sea grass, depth and

1:35:44 Speaker 4: other factors.

1:35:47 Speaker 8: Great, Chris, thank you so much. As we wrap up,

1:35:50 Speaker 8: I just want to thank one of all the participants

1:35:52 Speaker 8: that joined wherever you joined from, Thank you so much.

1:35:54 Speaker 8: We do look forward to seeing you in two days

1:35:56 Speaker 8: on Thursday for the second and final part of this

1:35:59 Speaker 8: training series. We also want to thank once again Laurie McGruder,

1:36:03 Speaker 8: Diane Fritz, and JP Swinsky, and also for those that

1:36:06 Speaker 8: were able to join to help answer the questions, Jeff

1:36:09 Speaker 8: Perry and Chris Parrish. Thank you both. And we also

1:36:12 Speaker 8: want to acknowledge that Amy Neely who's also helping out

1:36:15 Speaker 8: and supporting this training as well. So thank you to

1:36:17 Speaker 8: all the people that presented and also helped to answer

1:36:19 Speaker 8: the questions, and we look forward to seeing you all

1:36:21 Speaker 8: in two days for the second part of the webinar series.

1:36:24 Speaker 8: Thank you