NASA ARSET_ Overview of GEOGLOWS for Monitoring and Predicting Flood Risk

0:00 Speaker 1: Hello everyone, Welcome back to this final session of the

0:04 Speaker 1: training on Monitoring and Predicting Floods using Earth Observations for

0:08 Speaker 1: Planning and Preparedness. Part three today is going to focus

0:12 Speaker 1: on overview of Group of Earth Observations Global Water Sustainability

0:17 Speaker 1: tool that is GEOGLOS for monitoring and predicting flood risk.

0:22 Speaker 1: Our guest speakers today are Rachel Megoffin from Aquaveo LLC

0:27 Speaker 1: and Riley Hales from Brigham Young University. We'll start with

0:32 Speaker 1: a brief review of what we saw in Part two.

0:35 Speaker 1: We had doctor Renito Franso. We provided an overview of

0:39 Speaker 1: Opera Dynamic Surface Extent or DSWX products which are derived

0:45 Speaker 1: from harmonized Lancet and Sentinel two optical imagery and Sentinel

0:51 Speaker 1: one synthetic aperture radar imagery. These products are DSWX HLS

0:56 Speaker 1: and DSWXS one. The products are near global and are

1:01 Speaker 1: available at thirty meter resolution. The temporal coverage or that

1:06 Speaker 1: TSWXHLS is from April twenty twenty three to present with

1:11 Speaker 1: approximately three day resolution and DSWXS one started in December

1:17 Speaker 1: twenty twenty three extend to present with six to resolution.

1:24 Speaker 1: We also saw how to select access and composite DSWXHLS

1:30 Speaker 1: and DSWXS one products. DSWX data products can be visualized

1:36 Speaker 1: using NASA Worldview and data can be accessed using nasal Data.

1:40 Speaker 1: We saw that multiday composite of these two products are

1:45 Speaker 1: recommended when there are persistent clouds. Flood detection more accurate

1:51 Speaker 1: with DSWXHLS for airy region, while DSWXS one for inundated

1:58 Speaker 1: vegetation and multiple DSWX data layers merged using QGIS can

2:05 Speaker 1: improve flood detection capability. Doctor Fransov demonstrated that so today

2:13 Speaker 1: in the final part will have overview of geoglows for

2:16 Speaker 1: Monitoring and Predicting flood risk. There will be one homework

2:21 Speaker 1: posted on twenty fifth June, that is today, and it

2:24 Speaker 1: will be due on ninth of July. The homework will

2:27 Speaker 1: be posted on the training webpage. A certificate of completion

2:31 Speaker 1: will be awarded to those who attend all live sessions

2:34 Speaker 1: and complete the homework assignment before the two date. So

2:41 Speaker 1: we'll start with today's session overview of geoglows for Monitoring

2:45 Speaker 1: and Predicting flood risk. Overall objectives for Part three are

2:50 Speaker 1: that by the end of this part participants will be

2:53 Speaker 1: able to identify the capabilities of Geogloss River Forecast System

2:58 Speaker 1: or RFS for global stream flow prediction and use Geoglose

3:02 Speaker 1: Hydroviewer to access globally available retrospective and predicted stream flow

3:08 Speaker 1: for selected reverse Our outline today is that Rachel and

3:14 Speaker 1: Riley they will be talking about Geoglows. They will provide

3:17 Speaker 1: an overview of Geoglows River Forecast System which is based

3:21 Speaker 1: on meteorology model, hydrology model, stream flow routing model and

3:26 Speaker 1: stream analysis for bias estimation and reduction. Then we'll talk

3:31 Speaker 1: about Geoglows RFS data access and there will be a

3:34 Speaker 1: demonstration of how to use Geoglose Hydroviewer. Once again, Please

3:42 Speaker 1: put your questions in the question box and we will

3:44 Speaker 1: address them at the end of the webinar. Feel free

3:48 Speaker 1: to enter your questions as we go and we will

3:50 Speaker 1: try to get all of the questions during the question

3:54 Speaker 1: and answer session. After the webinar, the remainder of the

3:57 Speaker 1: questions will be answered in the Q and a document

4:00 Speaker 1: which will be posted on the training website about a

4:03 Speaker 1: week after the training. With that, I want to introduce

4:08 Speaker 1: our speakers for today. Rachel Hooper macaffin is a hydrologic

4:13 Speaker 1: data scientist at Aquaveayo with a master's degree from Brigham University.

4:20 Speaker 1: She specializes in GIS and hydrologic modeling. For the past

4:25 Speaker 1: five years, she has contributed to the Geogloss River Forecasting System,

4:30 Speaker 1: starting as a student at Brigham Young University and continuing

4:34 Speaker 1: in her role at Aquaveyo. Through projects with Surveer and

4:42 Speaker 1: the World Meteorological Organization, she has helped lead workshops designed

4:47 Speaker 1: to build capacity for data management and flood preparedness across

4:51 Speaker 1: Central America, South America, and Africa. Doctor Riley Hills is

4:57 Speaker 1: a hydrology and river hydrolic researcher at Brigham Young University.

5:02 Speaker 1: He is the technical director of the Geogloss program, where

5:06 Speaker 1: he builds and operates a global coverage hydrological model called

5:10 Speaker 1: the River Forecast System. He has provided hydrologic capacity building

5:16 Speaker 1: and training in more than a dozen countries in Central

5:20 Speaker 1: and South America, Africa, the Middle East, and South Asia.

5:25 Speaker 1: RFS data is used by water and disaster management agencies

5:29 Speaker 1: in these areas for flood preparedness, early warning systems, agricultural planning,

5:35 Speaker 1: water quality monitoring, and other applications. So with that, we

5:39 Speaker 1: invite Rachel and Riley to talk about geoglows.

5:46 Speaker 2: Okay, thank you, Amita for that wonderful introduction. As she said,

5:50 Speaker 2: my name is Rachel Hubert macgoffin. I work for AQUAVEO

5:54 Speaker 2: and I've been working with the Geoglos Initiative for a

5:56 Speaker 2: couple of years now, and I'm really excited to share

5:59 Speaker 2: a little bit about our river forecast system, our hydrologic model.

6:04 Speaker 2: And so as I begin, you may be familiar with GEO,

6:08 Speaker 2: but if you're not, that's a group on Earth observations

6:10 Speaker 2: and it is a partnership of more than one hundred

6:13 Speaker 2: national governments in participating organizations that envisions a future in

6:18 Speaker 2: which decisions and actions for humanities benefits are based on coordinated,

6:22 Speaker 2: comprehensive and sustained Earth observations. And so this is a

6:26 Speaker 2: large partnership in geoglows kind falls under the umbrella of GEOO,

6:32 Speaker 2: and geoglows is focused on global water sustainability, and so

6:37 Speaker 2: GEOGLOWS leverages partnerships, data and resources to deliver accurate, open

6:42 Speaker 2: and accessible hydrological predictions on a global scale, playing a

6:47 Speaker 2: vital role in addressing the complex challenges of water resource management.

6:52 Speaker 2: And so I wanted to begin with this introduction, so

6:54 Speaker 2: you kind of understand that geoglows itself is a larger organization. Today,

6:59 Speaker 2: we're going to focus just on our hydrological model, and

7:02 Speaker 2: we're going to focus specifically on streamflow and so on

7:05 Speaker 2: the hydrologic global scale. We're going to focus on kind

7:08 Speaker 2: of what Geoglow's main flagship thing is is our river

7:12 Speaker 2: forecast system, our hydrologic model that produces streamflow. But if

7:17 Speaker 2: you want to learn more about geogos in general, you

7:19 Speaker 2: can visit our website www dot geoglows dot org. And

7:24 Speaker 2: so also, as I'm getting started, this hydrologic model, the

7:27 Speaker 2: River Forecast System RFS, as you will hear me refer

7:31 Speaker 2: to it, is slightly a paradigm shift from what you

7:34 Speaker 2: may normally see in hydrologic models. We often see these

7:40 Speaker 2: global entities providing dem data, land use data, meteorology data,

7:45 Speaker 2: and we're used to seeing those on global scales, and

7:49 Speaker 2: often what happens is those meteorology data sets are taken

7:55 Speaker 2: and then a lot of funding, resources and trainings are

7:59 Speaker 2: given to local organizations and then it becomes their responsibility

8:03 Speaker 2: to do the hydrologic modeling. And so each place may

8:07 Speaker 2: build their own hydrologic model based on these global data

8:10 Speaker 2: sets and then from there they take their hydrologic model

8:14 Speaker 2: and use it to do things like agriculture planning, water

8:17 Speaker 2: supply management, disaster preparedness. And so they're both doing the

8:22 Speaker 2: modeling and doing the local decision making. And this makes sense.

8:27 Speaker 2: They're getting to make their own local model. Until when

8:31 Speaker 2: some of that funding, some of those resources, some of

8:34 Speaker 2: those trainings go away, maybe the person who was trained

8:37 Speaker 2: to leave the organization, and suddenly we find that those

8:40 Speaker 2: hydrologic models may not be up kept, they may not

8:43 Speaker 2: be sustainable, and this approach may end up breaking. And

8:46 Speaker 2: so Geoglows seeks to solve this problem by shifting it

8:50 Speaker 2: and instead we take in our River Forecast system, we

8:54 Speaker 2: take this global dem land data, meteorology data, and we

9:00 Speaker 2: provide our own global hydrologic model. And so this is

9:05 Speaker 2: done on a global scale, and then we focus the

9:08 Speaker 2: trainings and the resources on specifically taking the hydrology data

9:13 Speaker 2: sets and applying them to local applications of things like

9:18 Speaker 2: agriculture planning, water supply management, disaster preparedness. And so as

9:23 Speaker 2: I talk about the River Forecast system, it's going to

9:25 Speaker 2: be a little different because instead of giving you links

9:28 Speaker 2: to GitHub or to code where you're going to go

9:32 Speaker 2: run the model yourself. We are going to focus on

9:34 Speaker 2: giving you the results from the model, giving you the

9:37 Speaker 2: time series, giving you the products, giving you the visualizations.

9:41 Speaker 2: So then hopefully if you are in a situation where

9:43 Speaker 2: you work for a national hydrologic service or for a

9:45 Speaker 2: local government, you can just focus on doing what you're

9:48 Speaker 2: mandated to do, where you take that data and apply

9:51 Speaker 2: it to the situations that you need. And so with

9:56 Speaker 2: that in mind, our global hydrologic model, as I've said,

10:00 Speaker 2: is the River Forecast System, and so this contains streamflow

10:05 Speaker 2: data for rivers globally. It has both a forecast and

10:09 Speaker 2: a retrospective component to it, and we have used this

10:12 Speaker 2: model to address hydrologic needs in countries around the world

10:16 Speaker 2: by providing global data and allowing for local bias correction.

10:20 Speaker 2: And so we have worked in Central and South America,

10:23 Speaker 2: We've done projects in Africa and the Hinalayan region, and

10:26 Speaker 2: so we've really focused on this idea of global modeling

10:29 Speaker 2: but then taking it to specific locations and having local applications.

10:34 Speaker 2: And I just wanted to mention if you have worked

10:36 Speaker 2: with Geoglows RFS before it has been upgraded to a

10:41 Speaker 2: new version in the last couple of years, and so

10:44 Speaker 2: it has new file formats, more streams, and a longer

10:46 Speaker 2: retrospective simulation. So even if you've worked with this before,

10:51 Speaker 2: you should listen because you may still learn something new.

10:54 Speaker 2: And again, all the information on this and everything I'm

10:57 Speaker 2: going to go over in this presentation are on our

10:59 Speaker 2: training site training dot geoglows dot org, and I'm going

11:02 Speaker 2: to put a plug in for that over and over

11:04 Speaker 2: again because I know it's really hard to learn all

11:07 Speaker 2: this information from one presentation, and so that's a really

11:09 Speaker 2: good resource to go to get more information. So before

11:14 Speaker 2: we get into how to access this data, I want

11:16 Speaker 2: to go a little into the nuts and bolts of

11:18 Speaker 2: our model. So this is going to be a bit

11:19 Speaker 2: of a tangent, and we will then go into how

11:23 Speaker 2: to actually access the data. But I kind of wanted

11:25 Speaker 2: to tie into the remote sensing theme and kind of

11:28 Speaker 2: explain the origins of how the river forecast system operates.

11:32 Speaker 2: And so we know that good meteorology modeling or good

11:37 Speaker 2: hydrologic modeling is based on good meteorology, and so without

11:43 Speaker 2: the meteorology, we're not going to be able to figure

11:45 Speaker 2: out the hydrology.

11:46 Speaker 3: And so a.

11:48 Speaker 2: Meteorology model is basically an energy balanced calculations driven mainly

11:53 Speaker 2: by solar radiation, and so we see three D grid

11:56 Speaker 2: of varying grids cell heights with many layers. These grids

11:59 Speaker 2: are a coarse resolution, so think like a quarter degree

12:04 Speaker 2: twenty five kilometers. So these are a bigger resolution. We're

12:07 Speaker 2: going to use smaller resolution on our streams when we

12:10 Speaker 2: get to our stream catchments, and I'll go over that.

12:12 Speaker 2: But the meteorology model has a courser resolution. And then

12:16 Speaker 2: there's energy transfers happening horizontally in columns and also vertically

12:20 Speaker 2: in layers, and so these type of models are going

12:23 Speaker 2: to be calibrated for atmospheric variables, usually not precipitation or

12:27 Speaker 2: water variables, and so you may be familiar with some

12:32 Speaker 2: of these models. We have the Global Forecast System GFS,

12:36 Speaker 2: which is the American model, and then we have ACMWF

12:41 Speaker 2: their Integrated Forecast System IFS is the European model. This

12:45 Speaker 2: is what we use for the River Forecast System. RFS

12:47 Speaker 2: is based on IFS the ECMWS data, and these models

12:52 Speaker 2: when they produce forecasts, they don't just produce one value

12:56 Speaker 2: to make a guess into the future. Rather, they do

12:59 Speaker 2: ensemble prediction, where they predict a range of values of

13:02 Speaker 2: different different possibilities of things that could happen, and they'll

13:05 Speaker 2: often produce different a short range forecast, a medium range forecast,

13:09 Speaker 2: the long range forecast, and so generally the best models

13:13 Speaker 2: are going to be medium range, and they're going to

13:15 Speaker 2: be good for twelve ish days in advance, depending on

13:18 Speaker 2: your location and a lot of different factors. And so

13:21 Speaker 2: Geoglows RFS uses the ECMWF one, We use the medium range,

13:25 Speaker 2: the fifteen day forecast, which is why our streamflow is

13:27 Speaker 2: also fifteen days. And we use the ERA five data

13:31 Speaker 2: for our retrospective data and the IFS data for our

13:34 Speaker 2: forecast data. And I just wanted to mention that while

13:38 Speaker 2: the data we take in is runoff data from EERA five,

13:42 Speaker 2: this does not come from a satellite, but EA five

13:46 Speaker 2: is a global atmospheric or analysis from ECMWF, and so

13:50 Speaker 2: satellites provide many of the real world observations blended into

13:54 Speaker 2: EERA five, including temperature, humidity, precipitation, soil moisture, and so

13:58 Speaker 2: even if the runoff isn't measured by satellite, there are

14:02 Speaker 2: satellite values and observations that are going in to improve

14:07 Speaker 2: the model as a whole, which should then hopefully improve

14:09 Speaker 2: the runoff values. And that's what we take in for RFS,

14:13 Speaker 2: and so some satellites including Noah fifteen through nineteen, Agua Aqua,

14:19 Speaker 2: and so if you want to learn more about it,

14:21 Speaker 2: go read the EI five observations documentations. But that's kind

14:25 Speaker 2: of how we tie into remote sensing.

14:29 Speaker 3: And so now that we've.

14:30 Speaker 2: Talked about meteorology data, I'm going to move to talk

14:33 Speaker 2: about hydrology models. And so we know that in hydrology

14:39 Speaker 2: we're going to treat everything as if it's happening in

14:41 Speaker 2: a bucket. So precipitation is coming down, evaporation is going up,

14:46 Speaker 2: we have groundwater discharge, we have infiltration, we have snow melt.

14:51 Speaker 2: It's all happening, but it's all happening in this bucket.

14:53 Speaker 2: And this is happening as calculations. So we get you know,

14:56 Speaker 2: precipitation plus soil moisture, but then we take away the

15:00 Speaker 2: of appo transpiration and whatever we have left is what

15:03 Speaker 2: our runoff is. And so water is happening only in

15:07 Speaker 2: a bucket traveling vertically, but then the water is going

15:11 Speaker 2: to leave the bucket as runoff, and we're going to

15:15 Speaker 2: need to route it to put it into channels and

15:17 Speaker 2: move it along the channels into our rivers. And so

15:23 Speaker 2: we're going to start and we're going to take those

15:25 Speaker 2: those weather forecast grids and then in a bucket is

15:29 Speaker 2: happening all our hydrology and we're left with runoff and

15:33 Speaker 2: we can convert that into runoff depths, so we end

15:35 Speaker 2: up with a grid of run runoff depths. And then

15:39 Speaker 2: we're going to take streams and catchments because we know

15:41 Speaker 2: that we're modeling streamflow. We want to get the amount

15:43 Speaker 2: of water in the stream and we can figure out.

15:46 Speaker 3: Where the streams are using.

15:47 Speaker 2: Elevation because we know that water is going to travel downhill,

15:51 Speaker 2: and we can use these runoff grits in these streams

15:53 Speaker 2: and catchments to create a basin runoff volume these tables

15:58 Speaker 2: with kind series, and from there we're going to be

16:01 Speaker 2: able to route the water through our streams to create

16:03 Speaker 2: an overall hydrograph which will represent our time series of

16:07 Speaker 2: our data in our stream. And so again, this runoff

16:12 Speaker 2: depth is happening on grids, but we have no flow

16:15 Speaker 2: between the grids, and instead we have these streams, and

16:18 Speaker 2: so our little stream catchments don't follow the grids, so

16:22 Speaker 2: we need to convert our runoff depth into the basin

16:25 Speaker 2: runoff vout volumes. So we take the grid lines and

16:30 Speaker 2: we intersect the basins and we can get the percentage

16:33 Speaker 2: of the basin that's in each of the different grids,

16:36 Speaker 2: and we can sum them all together, and so we

16:38 Speaker 2: end up with the depth over that polygon. Then we're

16:42 Speaker 2: able to take that and repeat it for every basin,

16:47 Speaker 2: for every timestep. So now we have the volume of

16:49 Speaker 2: water in every basin at every time step, and we're

16:53 Speaker 2: going to be able to use that in our routing process.

16:57 Speaker 2: And so next comes the computations where we have to

17:01 Speaker 2: figure out, based on this timestep, based on these volumes,

17:05 Speaker 2: how is how's my water actually moving through my river?

17:09 Speaker 2: And so we know runoff is going to follow a

17:11 Speaker 2: terrain slope until it enters a river channel, and then

17:16 Speaker 2: river channels concentrate the runoff into an open channel flow

17:20 Speaker 2: and we assume no losses once in this channel flow,

17:24 Speaker 2: and then we can put the water and calculate how

17:26 Speaker 2: fast it moves.

17:27 Speaker 3: And now there's lots.

17:29 Speaker 2: Of different hydrologic routing methods. The one we use in

17:32 Speaker 2: our model is most similar to Muskigum routing and so

17:36 Speaker 2: kind of to go over our model, we start with

17:40 Speaker 2: our satellite data and that gets meteorology and land surface

17:44 Speaker 2: model data, and then we have our digital elevation model

17:48 Speaker 2: which turns into streams and catchments. And now using these

17:51 Speaker 2: two things, we can get our runoff per basin. Those

17:53 Speaker 2: are our volumes on our little basins. Then we need

17:57 Speaker 2: to route our water through our stream. We actually use

18:00 Speaker 2: a Python package. It was actually written by Riley Hales

18:04 Speaker 2: who's going to be showing the High Reviewer later, and

18:07 Speaker 2: we use that to calculate how the water is moving

18:10 Speaker 2: through the street the streams and add in the different

18:15 Speaker 2: volumes as they enter into the stream beds. But then

18:18 Speaker 2: we don't stop there, so then we get river discharged.

18:21 Speaker 2: But we also as part of our process, make data

18:24 Speaker 2: distribution a high priority and so we take our outputs

18:28 Speaker 2: and we put it on web services, We make bulk

18:30 Speaker 2: downloads available, We put it in long term archives on AWS,

18:35 Speaker 2: so all this data that we have collected through the

18:37 Speaker 2: sprouting process is now available for distribution. We also create visualizations.

18:42 Speaker 2: In partnership with EZRI, we have a living outless layer.

18:46 Speaker 2: We also produce plots and maps, reports, all of these

18:50 Speaker 2: and then hopefully people can take all these things we've

18:52 Speaker 2: provided to do local applications. And so the goal is

18:56 Speaker 2: that while it is a global model, that by providing

18:59 Speaker 2: the data to distribution tools and the visualizations, that this

19:02 Speaker 2: can then be taken and applied locally as people need it.

19:08 Speaker 2: So there's three main products I'm going to talk about

19:11 Speaker 2: from RFS today. There's the retrospective stream flow data, which

19:15 Speaker 2: goes back to nineteen forty. There's the fifteen day stream

19:19 Speaker 2: flow forecast data, which is an ensemble of data available

19:23 Speaker 2: daily for fifteen days in the forecast. And then we

19:26 Speaker 2: also have our hydrography data all available for download as well,

19:29 Speaker 2: including our streamlines and catchments and some other products I.

19:32 Speaker 3: Will talk about as well.

19:35 Speaker 2: So to start with, the retrospective data is based on

19:40 Speaker 2: the E five data set. I kind of mentioned that earlier.

19:43 Speaker 2: We use the runoff value as the value that we

19:45 Speaker 2: wrote in and this is a deterministic result and so

19:49 Speaker 2: there's just one value, it's not a range of value.

19:53 Speaker 2: The retrospect is deterministic and we have about six point

19:56 Speaker 2: eight million stream segments and so this is available on

19:59 Speaker 2: an hourly timestep eighty five years of data on six

20:02 Speaker 2: point eight million streams globally, and it updates with a

20:06 Speaker 2: five day lag from present, so it's pretty up to

20:09 Speaker 2: date in terms of how recently it's available and also

20:15 Speaker 2: from the retrospective data. We use all these tools to

20:18 Speaker 2: create derivative products, so that way, if someone doesn't want

20:22 Speaker 2: to download the whole time series and do the computations themselves,

20:25 Speaker 2: we do have derivative products already available for download. And

20:28 Speaker 2: so this includes return periods like the two year, five year,

20:32 Speaker 2: ten year, twenty five year, and fifty twenty five year,

20:35 Speaker 2: fifty and one hundred year return periods calculated using the

20:38 Speaker 2: Gumbele distribution and the log Pierson Type three distribution, and

20:41 Speaker 2: so this is like your one hundred what the value

20:43 Speaker 2: of your one hundred year flood is, your fifty year flood,

20:46 Speaker 2: your ten year flood. We also pre compute all the

20:50 Speaker 2: monthly averages, the yearly averages, the daily averages, which this

20:53 Speaker 2: is really nice if you're doing projects where you need

20:56 Speaker 2: to access all this information. It can save on a

20:58 Speaker 2: lot of streams. It can save you time for me

20:59 Speaker 2: to download all the data and compute those yourself. We

21:03 Speaker 2: also have flowdoration curves both available, like the overall flow

21:06 Speaker 2: duration curve for the river, but we also have the

21:09 Speaker 2: monthly floatation curve available for each month for each river.

21:13 Speaker 2: And so these are some examples. Riley's going to show

21:16 Speaker 2: the hydro viewer after and he will show that there's

21:18 Speaker 2: also a lot of other plots in charge that can

21:21 Speaker 2: be made using the retrospective data, and you're able to

21:24 Speaker 2: compute those and download those. But this is just some

21:27 Speaker 2: examples of some of the derivative products that we have

21:29 Speaker 2: available for download. We also have our forecast data and

21:34 Speaker 2: so this is a fifteen day ensemble forecast. So I

21:38 Speaker 2: mentioned earlier that the IFS data is produced in an ensemble.

21:42 Speaker 2: They produce several different possibilities, and we take each of

21:46 Speaker 2: their ensemble values for their runoff and we route each

21:48 Speaker 2: one of them. So therefore we end up having ensemble

21:54 Speaker 2: of forecast members as well. And so this is a

21:59 Speaker 2: fifteen day forecast. It's available every day, there's a new forecast.

22:03 Speaker 2: It's at three hour time steps, and there's one control

22:06 Speaker 2: forecast and fifty ensemble members, so you will see fifty

22:10 Speaker 2: one different possibilities. And we also have forecast statistics available

22:16 Speaker 2: and that will help provide the uncertainty bounds. You can

22:18 Speaker 2: kind of see your range of possible answers you could

22:20 Speaker 2: get as well as what is considered to be.

22:23 Speaker 3: The most likely.

22:25 Speaker 2: I also want to mention that all our forecast data

22:28 Speaker 2: we save, and so we have about two years at

22:32 Speaker 2: this point worth of old retrospective data old forecast data available,

22:38 Speaker 2: and so if you want to go look at your

22:40 Speaker 2: forecast data from about a year ago, you could go

22:43 Speaker 2: look at that. And so for some reason you need

22:45 Speaker 2: old forecast data that is available as well. And then

22:50 Speaker 2: I also mentioned that we have our hydrography data available.

22:54 Speaker 2: So as a background, the RFS hydrography data is based

22:57 Speaker 2: on TEDx hydro which is from the Tandem satellite and

23:00 Speaker 2: it's delineated using taodem, and so we took the TEDx

23:05 Speaker 2: hydro stream delineations, but we did make some improvements to

23:10 Speaker 2: it to help our model run smoother and to fix

23:12 Speaker 2: some of the errors that were present. And so some

23:14 Speaker 2: of these will be applicable to everyone, and some of

23:17 Speaker 2: the improvements we made any will be applicable to anyone

23:20 Speaker 2: who wants to use the hydrography data. But also some

23:23 Speaker 2: of it was specifically for our model and what we

23:25 Speaker 2: needed to happen. So we combined some of the headwaters,

23:29 Speaker 2: like to order one streams we might combine with their

23:32 Speaker 2: order to stream to be one single basin. We combined

23:35 Speaker 2: some of the really tiny short streams as well as

23:37 Speaker 2: we're moving some water sheds and flat deserts where we're

23:40 Speaker 2: not seeing any flow ever in them. And so we

23:43 Speaker 2: did these things to help the model run smoother and

23:45 Speaker 2: to help make our numbers more accurate in the model

23:48 Speaker 2: overall better. And so the version that we ended up

23:51 Speaker 2: with of our hydrography is what we have available now

23:53 Speaker 2: for other people to download. And so this data, the

23:57 Speaker 2: hydrography data, is all the into VPUs, and a VPU

24:02 Speaker 2: is a vector processing unit which is basically just a

24:05 Speaker 2: collection of watersheds that we bundled together to make more

24:08 Speaker 2: efficient processing and data management. And so if you know

24:12 Speaker 2: your VPU, you can download just that VPU instead of

24:15 Speaker 2: needing to download globally all the streams. Because with six

24:17 Speaker 2: point eight million streams, that sends up to.

24:19 Speaker 3: Be a lot of data.

24:20 Speaker 2: So being able to download just your VPU of data

24:23 Speaker 2: is a way to make it a lot more efficient,

24:25 Speaker 2: and so ours is one hundred and twenty five globally,

24:29 Speaker 2: and you'll see that in the a US buckets. They're

24:31 Speaker 2: split into VPUs and that's how we manage the data.

24:36 Speaker 2: So you can download the actual stream center lines, and

24:39 Speaker 2: these are the center lines representing the flow path for

24:42 Speaker 2: a river segment, and every stream has a unique nine

24:46 Speaker 2: digit river idea. You might also hear this referred to

24:49 Speaker 2: as a link number, and this is what's going to

24:51 Speaker 2: be on the hydrography data. It's also the link number

24:54 Speaker 2: you'll use when you want to query the retrospective data

24:57 Speaker 2: or the forecast data. This is its identifying number. And

25:00 Speaker 2: so in the hydrography data you can see what the

25:02 Speaker 2: link numbers are. You'll also see that on the hydro

25:05 Speaker 2: viewer as well. And we also in our stream data

25:08 Speaker 2: have some additional attributes that might be useful to some people.

25:12 Speaker 2: So we have the stream order, we have the upstream area,

25:16 Speaker 2: we have what region it came from in TDX hydro

25:19 Speaker 2: and so, and like what the terminal link is the

25:22 Speaker 2: output for that stream.

25:23 Speaker 3: So that is all available in the stream package.

25:29 Speaker 2: And then we also have nexus points and lakes and

25:33 Speaker 2: so nex's points are just the points where the streams connect,

25:36 Speaker 2: so you can see all the points where we have

25:38 Speaker 2: the segments connecting with each other, and it'll mark and

25:40 Speaker 2: it'll tell you what the upstream and downstream rivers are

25:42 Speaker 2: at that point. And then we have the basins that

25:45 Speaker 2: we consider lakes in our model, and well these are

25:48 Speaker 2: all available on the AWS buckets that I'll show you

25:51 Speaker 2: how to access in a minute. And lastly, if you're

25:55 Speaker 2: looking at our hydrography data we do, if you are

25:58 Speaker 2: looking to put it in some sore of GIS software,

26:01 Speaker 2: the easiest way to do it is to take to

26:03 Speaker 2: use the EZRE webmap, and so in collaboration with EZRE,

26:07 Speaker 2: there's a publicly hosted Living Atlas layer. And so this

26:11 Speaker 2: layer is nice because it is done so that the

26:15 Speaker 2: thickness of the streams matches the amount of water in

26:19 Speaker 2: the streams, and then the colors represent if it's exceeding

26:22 Speaker 2: a certain return period. So you can see here that

26:24 Speaker 2: that little area where it's purple there that stream flow

26:27 Speaker 2: is at this point exceeding fifty years. And this is

26:31 Speaker 2: a ten day animated map, and so you can actually

26:33 Speaker 2: see those flow period those flow return period colors are

26:38 Speaker 2: representing what is currently happening because of this living outless layer,

26:43 Speaker 2: and so this is a really nice option to have

26:45 Speaker 2: for GIS. And so we start with our geoglows version

26:52 Speaker 2: two data and that is delivering AWS open data and

26:56 Speaker 2: web service sources. And so this is what I've talked.

27:00 Speaker 2: We've talked about the daily historical data, the daily animated maps,

27:03 Speaker 2: the daily new forecast. We also have our and this

27:06 Speaker 2: is all happening on a supercomputer. And then we're delivering

27:09 Speaker 2: all these products and they're being stored on AWS. They're

27:14 Speaker 2: being done stored as the Living Atlas layer with Ezer,

27:17 Speaker 2: there's an API, and so these are all stored. But

27:20 Speaker 2: then the hope is that all these stored data sets

27:22 Speaker 2: are then taken and powering local adaptable open source code

27:26 Speaker 2: and web applications. So the goal is that all this

27:29 Speaker 2: global data that we've provided can then be taken and

27:32 Speaker 2: applied locally. And so I now want to show you

27:35 Speaker 2: and tell you about how you can take all these

27:37 Speaker 2: things I've told you are available and use them.

27:41 Speaker 3: And so in order to do that, you will need.

27:44 Speaker 2: To access the Geoglosse data. Now, the easiest way to

27:48 Speaker 2: access the data is in the hydrid viewer, and this

27:51 Speaker 2: is going to be demoed in the second half of

27:53 Speaker 2: this workshop, and this is a web application that you're

27:57 Speaker 2: going to again it'll be demoed, but use you open

28:00 Speaker 2: it up, there's a map, you zoom into your stream,

28:02 Speaker 2: you click on your stream, and a bunch of charts

28:04 Speaker 2: and information and things you can download just pop up

28:07 Speaker 2: right there for you. And so it's really nice, it's

28:09 Speaker 2: really easy, and it's also generally easiest for that one

28:12 Speaker 2: if you only have one specific stream. And so let's

28:16 Speaker 2: say you have a whole series of streams and you're

28:19 Speaker 2: trying to query them every every day. The hydro viewer,

28:23 Speaker 2: where you have to click on every single one of

28:24 Speaker 2: the streams if you have two hundred streams, might not

28:26 Speaker 2: be the best option for you anymore. And so there

28:29 Speaker 2: is an option where you can query data yourself from

28:32 Speaker 2: the AWUS bucket. And so this means that you can

28:36 Speaker 2: write Python code to directly access the data and then

28:38 Speaker 2: it'll read into your Python code, which makes it easy

28:41 Speaker 2: from there to be able to manipulate it, include it

28:43 Speaker 2: in web applications, use it for what you need, analyzes, report,

28:47 Speaker 2: it's all available for you. And to help make that

28:50 Speaker 2: one step easier, we have the geoglows Python package that

28:54 Speaker 2: makes it super easy to query the data. And you

28:56 Speaker 2: don't even have to create the a to us bucket directly.

28:59 Speaker 2: You can just align a code to query everything using the.

29:03 Speaker 3: Googles Python package.

29:04 Speaker 2: And this queering the data is generally going to be

29:06 Speaker 2: the fastest way to download the data, especially if you're

29:10 Speaker 2: looking at a large quantity of data. Again, if you're

29:13 Speaker 2: looking at say one hundred different river segments, this is

29:17 Speaker 2: probably going to be your best way. And then of

29:19 Speaker 2: course we have our webmap that I've talked about that's

29:22 Speaker 2: the ten day animated streams layer showing forecast results. And

29:26 Speaker 2: this is free to use or view in any website

29:29 Speaker 2: or GIS, So if you're doing something with GIS software,

29:31 Speaker 2: this is a really easy way to integrate RFS. And again,

29:36 Speaker 2: I know this is a lot of information, and as

29:38 Speaker 2: I go through this, all all of this information is

29:40 Speaker 2: on training that geoglos dot org. So go ahead and

29:43 Speaker 2: take a look at that website to kind of refresh

29:46 Speaker 2: yourself on everything I've talked about. So I've mentioned that

29:51 Speaker 2: all our data is available on AWS buckets and they

29:55 Speaker 2: are available for open use, and this is going to

29:58 Speaker 2: be a really fast way to download large sets of data,

30:01 Speaker 2: and on our AWS buckets we have everything available. So

30:06 Speaker 2: we have metadata, routing, configuration files, we have the hydrography

30:09 Speaker 2: data I talked about, we have forecast data, we have

30:12 Speaker 2: retrospective data. We have some of those derivative products to

30:15 Speaker 2: the retrospective data with different averages, and so these are

30:18 Speaker 2: all available on AWS. And so you can go to

30:22 Speaker 2: the actual AWS bucket like you see here on my screen,

30:26 Speaker 2: and you can click on the subset that you want

30:28 Speaker 2: and you can click through and you can download the file. Now,

30:31 Speaker 2: I will say you will download a lot of data

30:33 Speaker 2: doing this because you're gonna have to download the whole file.

30:36 Speaker 2: So generally it is better to use the geoglos python

30:39 Speaker 2: package or to query it programmatically and then you can

30:42 Speaker 2: query specifically for the rivers you're interested in.

30:46 Speaker 3: And so the geoglows python.

30:48 Speaker 2: Package is a Python package that allows for data to

30:51 Speaker 2: be easily queried, and it defaults to querying the AWS

30:54 Speaker 2: buckets as its source, and so the retrospective and forecast

30:59 Speaker 2: values can be queried there if you want to use this.

31:01 Speaker 2: I'm not going to be able to go over all

31:04 Speaker 2: the options with the geogloss python package, but we do

31:06 Speaker 2: have documentation on how to use it and what commands

31:09 Speaker 2: are available.

31:11 Speaker 3: And one highlight of.

31:13 Speaker 2: The Geoglos Python package is that multiple rivers can be

31:16 Speaker 2: requested at once. You can request all of your rivers

31:19 Speaker 2: and it's also easy to write code to be able

31:22 Speaker 2: to check it again in a couple of days if

31:24 Speaker 2: you need to update it. And so I have a

31:27 Speaker 2: little code snippet here, I just want to show that

31:29 Speaker 2: it actually is like one line of code to get

31:31 Speaker 2: it all. And so you just import the package, you

31:33 Speaker 2: give it all the river ideas you want, and then

31:36 Speaker 2: it's as easy as geoglows dot data dot forecast ensembles

31:39 Speaker 2: with your river list or geoglows dot data dot retrospective

31:44 Speaker 2: with your river list, and that's going to spit you

31:45 Speaker 2: out a data frame with the daytime index and a

31:48 Speaker 2: column per river id, and so again it's a really

31:52 Speaker 2: fast and easy way to be able to get data

31:54 Speaker 2: for a lot of rivers. You also, if you are

31:58 Speaker 2: more comfortable with queering aid to bas directly instead of

32:02 Speaker 2: using the Python package, you can create aws bucket programmatically

32:07 Speaker 2: or use the command line if you are already familiar

32:09 Speaker 2: with this. And there's more examples of this on training

32:11 Speaker 2: dot geoglows dot org, so you can get the URI

32:15 Speaker 2: for the AWS bucket and using x ray you can

32:18 Speaker 2: just open that data set and select the rivers you

32:21 Speaker 2: want and get the data that way. You do need

32:23 Speaker 2: to make sure if you do this option that you

32:26 Speaker 2: need to make sure you create anonymously or it will

32:28 Speaker 2: throw an error.

32:29 Speaker 3: It is open for.

32:30 Speaker 2: Everyone, but if you don't queerate anonymously, it'll say you

32:33 Speaker 2: don't have permission. But as long as you create anonymously,

32:35 Speaker 2: you will have permission to access the data. I generally

32:38 Speaker 2: think the Python package is easier, but if you are

32:40 Speaker 2: more familiar with this way, or you have specific applications

32:42 Speaker 2: where you need to use the AWS buckets directly, this

32:46 Speaker 2: is an option of something you can do. And then again,

32:51 Speaker 2: our other way to access data, probably the easiest, most

32:54 Speaker 2: user friendly is the RFS version two hydro viewer. It

32:58 Speaker 2: has a ton of visualizations available for your river. It

33:01 Speaker 2: will it can do lots of things. You zoom in,

33:04 Speaker 2: you click on your river, you can see old forecast data,

33:07 Speaker 2: current forecast data, retrospective data, a bunch of different graphs.

33:12 Speaker 2: You can see the streams colored and Riley will talk

33:15 Speaker 2: more about this later, but this is a really cool

33:17 Speaker 2: tool we have for data visualization and for data accessibility

33:23 Speaker 2: because you can just download the data straight from there.

33:27 Speaker 2: And now I've kind of talked about our data, I've

33:29 Speaker 2: talked about how to access it, but I wanted to

33:31 Speaker 2: mention one more thing as I kind of start to

33:34 Speaker 2: wrap up this part of the presentation before Riley does

33:36 Speaker 2: the demonstration, and that is I've talked a little a

33:40 Speaker 2: lot about how the goal of geoglows is that well,

33:42 Speaker 2: it is a global model, we hope that it can

33:44 Speaker 2: go to local applications. And now, as a global model,

33:48 Speaker 2: often is bias exists on a local scale when you're

33:52 Speaker 2: going to be using a global model, and some of

33:54 Speaker 2: that bias comes from the runoff data that we input

33:57 Speaker 2: as the input into it arefs, is going to be

34:00 Speaker 2: other sources a bias that may occur based on your

34:03 Speaker 2: specific conditions of where you live. And so we are

34:07 Speaker 2: always working on ways to improve our model to get

34:09 Speaker 2: better numbers, better answers out. However, one of the ways

34:14 Speaker 2: you can get better numbers is a bias correction option,

34:18 Speaker 2: and so there is an option for local bias correction,

34:21 Speaker 2: which then enables better local.

34:23 Speaker 3: Results for the model.

34:24 Speaker 2: So this does require observed data at the stream segment,

34:29 Speaker 2: but then it will give you both forecasts and retrospective data,

34:33 Speaker 2: and it'll be corrected to match the volume of water

34:35 Speaker 2: we're seeing in the observed data. So if we tend

34:38 Speaker 2: to in the simulated data often over predict by bias correcting,

34:43 Speaker 2: we can correct the volume and our range of values

34:45 Speaker 2: will be more similar to what you may see in

34:47 Speaker 2: your observed value. And this is available to use in

34:51 Speaker 2: the Geoglows Python package. And we just have some functions

34:54 Speaker 2: geoglos dot bias dot correct historical and Geoglows dot bias

34:58 Speaker 2: dot correct forecast, and those are available for you to

35:02 Speaker 2: use and to perform your own local bias correction on

35:06 Speaker 2: your own data with your own data on the global data.

35:10 Speaker 2: And so the way this method works is so we

35:13 Speaker 2: take the simulated hydrograph that we are getting out of

35:16 Speaker 2: the RFS, and we're going to get the floatation curve

35:19 Speaker 2: for that simulation data, and then we're going to get

35:23 Speaker 2: the observed floatation curve from the observe data that you

35:26 Speaker 2: upload as part of the function, and we're going to

35:29 Speaker 2: use that observed data to correct the values of a hydrograph.

35:33 Speaker 2: So then the values of the hydrograph, the volume of

35:35 Speaker 2: it matches more similar to what we were seeing in

35:37 Speaker 2: the flowtation curve of the observed data. And so we

35:44 Speaker 2: also acknowledge that observed data, which is necessary to do

35:50 Speaker 2: bias correction, is not available for every river, and we

35:55 Speaker 2: have been working on ways to then try to improve

35:58 Speaker 2: the bias in areas.

35:59 Speaker 3: That may not have the subserved data.

36:02 Speaker 2: And one experimental process we have right now is called

36:05 Speaker 2: Saber stream analysis for Bias Estimation and Reduction. And what

36:09 Speaker 2: this does is it takes the bias correction methods that

36:13 Speaker 2: we have locally and it works to apply them globally.

36:16 Speaker 2: And so we've taken the data that has been given

36:19 Speaker 2: to us, whether and we have permission to use, whether

36:22 Speaker 2: from GRDC USGS, specific organizations or countries, and this is

36:28 Speaker 2: then included in the model. And we've taken the basins

36:31 Speaker 2: globally and they've been clustered and so like basins are

36:36 Speaker 2: similar to each other and so they have similar patterns.

36:39 Speaker 2: And then we're able to take flow duration curves from

36:42 Speaker 2: the gauged networks and the corrections we perform we're able

36:47 Speaker 2: to apply to other gauges, whether in a similar region

36:50 Speaker 2: or in a similar cluster in the same cluster, and

36:55 Speaker 2: so I won't go into the whole algorithm for it,

36:58 Speaker 2: but it's a way to apply the bias correction using

37:02 Speaker 2: regionalized base and data, and so all the streams can

37:04 Speaker 2: benefit from the bias correction happening in some of the streams.

37:09 Speaker 2: And so this data, the global bias corrected data, is

37:12 Speaker 2: available on the hydro viewer and all those time series

37:16 Speaker 2: and things, both for forecast and for retrospective data that

37:19 Speaker 2: we have available for our with our normal data is

37:25 Speaker 2: available for the global bias correction. But I do want

37:28 Speaker 2: to mention that this data is experimental. We are still

37:32 Speaker 2: evaluating it and working on lots of improvements in our

37:35 Speaker 2: model as always, and so just use this data at

37:37 Speaker 2: your own risk. It may improve some of your numbers,

37:40 Speaker 2: but it may make some of your numbers worse. And

37:41 Speaker 2: so just make sure you understand what you are using

37:45 Speaker 2: when you choose to use this data. And so that

37:49 Speaker 2: is going to conclude my presentation. That's kind of overviewing

37:52 Speaker 2: all of the river forecast system. Again, I know it's

37:55 Speaker 2: kind of a lot to take in all at once,

37:57 Speaker 2: and so I encourage you to go check out our

37:59 Speaker 2: way website, our training website at training dot geoglows dot org,

38:03 Speaker 2: where it's going to have all the information I talked

38:05 Speaker 2: about and more in addition to examples and code snippets

38:09 Speaker 2: to help you work through, and you can go look

38:10 Speaker 2: at the parts that specifically interest you. And then I'm

38:14 Speaker 2: going to let Riley take over and he's going to

38:16 Speaker 2: show you in real time how to use the hydro

38:18 Speaker 2: viewer and some of these tools I've talked about so

38:20 Speaker 2: that hopefully you feel a little more comfortable taking them

38:23 Speaker 2: and using.

38:23 Speaker 3: Them for yourselves.

38:31 Speaker 4: Well, thank you Rachel for that introduction to everything there

38:34 Speaker 4: is to know how geoglows RFS.

38:36 Speaker 5: I know that was a lot.

38:37 Speaker 4: I'm going to do my best to provide some concrete

38:40 Speaker 4: examples of everything that we just learned about. I'm going

38:44 Speaker 4: to start by giving another plug for the training sites.

38:49 Speaker 4: This is training dot geoglows dot org. And when you

38:52 Speaker 4: first land here on the left, you'll see and a

38:55 Speaker 4: column of about all the headings and there's River Forecasts

39:00 Speaker 4: system is one of your options. Part one about model

39:03 Speaker 4: formulation is a simplified version of what we just went over,

39:07 Speaker 4: but it does have a lot of extra links to

39:09 Speaker 4: other places where you can read more about the different

39:12 Speaker 4: components and journal articles and things like that. I'd love

39:16 Speaker 4: to draw your attention there if you have further questions,

39:18 Speaker 4: and I'd be happy to take them in the Q

39:19 Speaker 4: and A part as well. However, moving down to two

39:23 Speaker 4: available data where I'm starting to pick up on this

39:27 Speaker 4: this first entry data catalog.

39:30 Speaker 3: Has.

39:31 Speaker 4: If I scroll down just a little bit, there's this

39:35 Speaker 4: table of data sets, section and on.

39:37 Speaker 5: To highlight a couple of things here.

39:39 Speaker 4: First, all of this data is on an AWSS three buckets.

39:44 Speaker 4: There's a ABS has this program that makes it such

39:47 Speaker 4: that you do not need to have a user account.

39:50 Speaker 4: There's no tokens, no put in your credit card, no

39:54 Speaker 4: user name or password, nothing like that is required to

39:56 Speaker 4: access any of the data I'm about to show you.

40:00 Speaker 4: And that second is that there are two locations to

40:04 Speaker 4: find all of this data. So if you're the kind

40:06 Speaker 4: of person that wants to write code that obtain this,

40:09 Speaker 4: if you want to go brows what's there and refresh

40:12 Speaker 4: your memory, this table will give you an entry for

40:16 Speaker 4: every row.

40:17 Speaker 5: Is one of the things that I can provide you

40:19 Speaker 5: what file.

40:19 Speaker 4: Format it's in, what's the URL for you to go

40:22 Speaker 4: find it? The AWS region. If you're using it within

40:25 Speaker 4: the AWS ecosystem. So if if you're interested in obtaining

40:29 Speaker 4: the daily forecasts, you'll see this is in a ZAR

40:32 Speaker 4: formats and that's where you go looking for it. If

40:36 Speaker 4: you're interested in the hourly average flow historical simulation, this

40:40 Speaker 4: is the same thing comes in ZAR formats.

40:43 Speaker 5: Here's the URL to go looking for it.

40:46 Speaker 4: When you go to one of these buckets, all of

40:48 Speaker 4: these links will take you to well zoomed into the

40:52 Speaker 4: correct subdirectory within the buckets, but you'll end up on

40:56 Speaker 4: a page that looks like this. This farthest less column

41:01 Speaker 4: will be the different headings, and these headings will all

41:04 Speaker 4: correspond to something that is in the table on the

41:08 Speaker 4: data catalog portion of the training sites.

41:12 Speaker 5: So feel free to browse these as well.

41:15 Speaker 4: If you're a AWS guru, there should be old hats

41:19 Speaker 4: and feel free to go access this. No tokens or

41:22 Speaker 4: log in or anything like that is required. Moving on

41:27 Speaker 4: to the Hydroviewer. This is hydroviewer dot googlows dot org.

41:33 Speaker 4: There's a two main purposes of the hydro Viewer that

41:36 Speaker 4: I'm going to try to cover with you today. First,

41:38 Speaker 4: it's a visualization tool. We've got a lot of features

41:41 Speaker 4: in here to help you explore data graphically and through

41:45 Speaker 4: maps and all sorts of visual means. And second, it's

41:49 Speaker 4: a way to retrieve data and look at plots for

41:55 Speaker 4: those that are not native English speakers. You're of course

41:58 Speaker 4: welcome to use automatic Translate anywhere you want, but I'll

42:01 Speaker 4: start by up on the top right, we've got official

42:04 Speaker 4: human translated and verified translations in Spanish and in French,

42:09 Speaker 4: so if those that if you speak one of those

42:11 Speaker 4: language and have that preference, we do have a human

42:14 Speaker 4: translated version already available, so that the graphs and the

42:17 Speaker 4: map labels and anything that can be translated will be translated.

42:23 Speaker 4: Next is right next to that is there's this gear

42:27 Speaker 4: icon for settings. Now, this tool is definitely targeting a

42:32 Speaker 4: more scientific user, but even given all of the wealth

42:37 Speaker 4: of the charts and maps and everything that's available, there's

42:41 Speaker 4: some of those things are not on by defaults, So

42:44 Speaker 4: I'll draw your attention to First, we'll be looking at

42:48 Speaker 4: a simplified view of the forecast. By defaults, that's the

42:51 Speaker 4: recommended place I would start. However, there's this option that's

42:55 Speaker 4: off by defaults, which says show additional plots if you

42:59 Speaker 4: want to see the extra feel free to toggle that on,

43:02 Speaker 4: and as well as for bias correction, Rachel mentioned how

43:06 Speaker 4: we've done a lot of analysis to understand that bias

43:09 Speaker 4: is the primary error, and that is that it's correctable

43:12 Speaker 4: using gauge data, where it's a experimental method to make

43:17 Speaker 4: that apply to all river segments in the model, not

43:20 Speaker 4: just the ones where we have gauge data. Feel free

43:23 Speaker 4: to toggle us on, and I suggests the further reading

43:27 Speaker 4: and journal articles on the training sites.

43:30 Speaker 5: If you're going to try experimenting with this option, all.

43:38 Speaker 4: Right, let's start by looking at the visualization options. I'll

43:42 Speaker 4: frust drate your attention to the base map that we're

43:46 Speaker 4: looking at here. This is called the environment map, and

43:49 Speaker 4: it's got a couple of things that helped clarify looking

43:52 Speaker 4: at rivers. The blue lines are the rivers from TDX

43:57 Speaker 4: hydro that we've revise to do hydrologic modeling on. But

44:00 Speaker 4: these brown lines are watershed boundaries from hydro basins. And

44:06 Speaker 4: as you zoom in you'll see more those divisions appear,

44:09 Speaker 4: and there'll be other blue lined references and those all

44:12 Speaker 4: come from hydro sheds. However, if you have a preference

44:16 Speaker 4: for any others, you can of course change to imagery

44:19 Speaker 4: or topographic or any of the other options to be

44:24 Speaker 4: most friendly for all audiences, I'm going to keep it

44:28 Speaker 4: on a simple gray one so that there's the maximum

44:30 Speaker 4: amount of contrast for some of these streams. The first

44:37 Speaker 4: thing I'll point out about the streams themselves is that

44:40 Speaker 4: there's both different thicknesses and different colors that are apparents.

44:45 Speaker 4: As I zoom in and out, will see more streams.

44:49 Speaker 4: That's called a multi scale map. So I'll start by

44:51 Speaker 4: zooming in to this north part of South America, where

44:55 Speaker 4: I see some streams and a few different countries that

44:57 Speaker 4: are colored differently. The legend on the bottom left reminds

45:01 Speaker 4: me that those colors correspond to that's exceeding a return period,

45:06 Speaker 4: or in other words, that the yellow is the beginning

45:08 Speaker 4: of you should be worried about flooding occurring, And the

45:12 Speaker 4: farther you go down this list to exceeding fifty years,

45:16 Speaker 4: the more likely is this is a bigger floods covering.

45:18 Speaker 5: More area, and it is more rare in how big

45:21 Speaker 5: that event is.

45:23 Speaker 4: So I'll zoom in and we'll notice that once I

45:27 Speaker 4: zoom in here more lines where are parents here compared

45:32 Speaker 4: to zoomed out, I'll do that one more time to

45:34 Speaker 4: draw attention. From this level, I can see just the

45:37 Speaker 4: main stems, the biggest portions of the rivers, and when

45:40 Speaker 4: I zoom in, these streams densify. That'll happen two times

45:44 Speaker 4: until you get all the way zoomed in. The Next

45:51 Speaker 4: thing to look at is that this map layer is animated.

45:55 Speaker 4: If I come to the top left, there's a clock icon.

45:59 Speaker 4: That's why of our controls. If I toggle that on,

46:03 Speaker 4: a time slider appears on the bottom. Forecasts are generated

46:08 Speaker 4: every day at midnight, and so that by default you'll

46:11 Speaker 4: see that first time step that corresponding to midnight UTC

46:15 Speaker 4: time zone. Regardless of where you are in the world,

46:18 Speaker 4: it's always going to be midnight UTC is the first

46:21 Speaker 4: time step. Feel free to slide this by hands, or

46:25 Speaker 4: you can use the There's a play button and every

46:27 Speaker 4: couple seconds it will slide through the timesteps. And if

46:31 Speaker 4: I zoom in far enough, we'll see that the thickness

46:35 Speaker 4: and the coloring of the rivers is going to change

46:39 Speaker 4: as the slider progresses down the line. So for here,

46:42 Speaker 4: for instance, if I pick out just one spot at

46:45 Speaker 4: the headwater, there's some yellow here, but as time progresses,

46:49 Speaker 4: the flood is receding, so there's it's leaving that warning

46:54 Speaker 4: threshold and going back to the standard blue, which is

46:57 Speaker 4: within normal flow ranges. If I now pause this and

47:01 Speaker 4: jump ahead, the animation goes out for ten days, and

47:06 Speaker 4: jumping ten days into the future, you can see most

47:08 Speaker 4: of the yellow has disappeared and the thicknesses for some

47:12 Speaker 4: of those segments have gotten narrower.

47:14 Speaker 5: Maybe there's less water in them.

47:21 Speaker 4: Next thing that I will draw attention to on the

47:24 Speaker 4: maps is again on the top left, there's a there's

47:27 Speaker 4: controls to manipulate what streams are visible. This filter icon

47:33 Speaker 4: will prompt you to filter the streams that are visible.

47:37 Speaker 4: Rachel mentions that the one of the intended applications of

47:42 Speaker 4: all of this hydrology data is to support custom applications

47:46 Speaker 4: by country or by watershed, so the map is capable

47:50 Speaker 4: of having filters applied. There's a few that are the

47:53 Speaker 4: most common that I provide you some dropdowns for like

47:57 Speaker 4: for instance, this one river country means if I type

48:00 Speaker 4: in the name of a country like United States or

48:05 Speaker 4: let's say Brazil, I'm closer to Brazil.

48:08 Speaker 5: If I check that box and hit apply.

48:11 Speaker 4: The map will reduce itself to only showing me the

48:14 Speaker 4: river segments that intersect the boundaries of the country. Now

48:20 Speaker 4: you might say, well, if this is in the Amazon,

48:23 Speaker 4: that crosses into some other country boundaries. So if I

48:26 Speaker 4: go back to the filter menu, I might say, show

48:28 Speaker 4: me countries that are either in Brazil or who outlet

48:33 Speaker 4: outlets in Brazil. I hit apply, Then some more of

48:37 Speaker 4: the Amazon becomes visible, even though those segments are outside

48:42 Speaker 4: of Brazil, but they flow into that. Between those two

48:45 Speaker 4: controls you have a lot of ability to display rivers

48:49 Speaker 4: that are trans boundary. So if your application area is

48:53 Speaker 4: in a country that's as most are, where the rivers

48:57 Speaker 4: has trans boundary flow, those two controls here help you

49:02 Speaker 4: create targeted visualizations for whatever application area you have. I'll

49:09 Speaker 4: clear for now to restore that. However, you'll see that

49:13 Speaker 4: there's this enter custom sequel. This is an EZRI web

49:18 Speaker 4: map layer, and I've left you a link to documentation

49:22 Speaker 4: if you are familiar with writing sequel. I won't cover

49:25 Speaker 4: that here, But there are other attributes, like you could

49:28 Speaker 4: provide a bounding box on latitude longitude. You can provide

49:32 Speaker 4: specific river numbers that you want to be filtered this

49:36 Speaker 4: site's the visit the Ezraelving Atlas link here will direct

49:39 Speaker 4: you to the pages with information about what all the

49:43 Speaker 4: attributes are, and you can build your own filters here.

49:50 Speaker 4: The other main components of visualization, we offer some supplementary

49:53 Speaker 4: layers that provide additional context to this.

49:57 Speaker 5: Up on the top.

49:58 Speaker 4: Right of the map area, there's a layer selector and

50:01 Speaker 4: from the most of these come from the Noah twenty

50:04 Speaker 4: satellites veers. You can get the most recent imagery on

50:09 Speaker 4: thermal anomalies. You can get true color reflectants. One that

50:13 Speaker 4: I'll toggle on is this Veer's flood composits. This is

50:17 Speaker 4: using the most recently available Vier's imagery running through a

50:21 Speaker 4: classification algorithm, and the colors corresponds to different land use

50:26 Speaker 4: types or there's.

50:27 Speaker 5: Blue for water, for instance.

50:30 Speaker 4: And use this as an example to say if while

50:34 Speaker 4: looking at anything on this app including in most common

50:39 Speaker 4: perhaps even these supplementary layers like the flood composite. On

50:44 Speaker 4: the top rights there's a question mark icon which will

50:48 Speaker 4: provide you links to both the geoglows RFS data and

50:53 Speaker 4: our various websites for instructions, but also for instance, the

50:57 Speaker 4: Vier's flood composite layer. There's this link to a un

51:00 Speaker 4: University of Wisconsin websites that has the explanation for.

51:05 Speaker 5: What the colors mean on that layer.

51:12 Speaker 4: Another one worth drawing your attention to is hydro SOS

51:15 Speaker 4: Monthly Status. Hydro SOS is the WMO World Meteorological Organization Activity.

51:24 Speaker 4: These polygons that you're seeing correspond to hydro shed's basins,

51:29 Speaker 4: and there's a yellowish hand ish color which represents normal

51:35 Speaker 4: and for that the month that's being visualized, it was drier,

51:42 Speaker 4: slightly drier, or much drier than normal, it gets one

51:45 Speaker 4: of these shades of red, or if it's slightly wetter

51:48 Speaker 4: or much wetter, it gets the lighter blue or dark blue. Correspondingly,

51:55 Speaker 4: These maps are available for every monthly average going back

51:59 Speaker 4: for thirty years, so I see it or sorry. If

52:03 Speaker 4: I direct your attention to the top left, there's a

52:05 Speaker 4: control labeled SOS for hydro SOOS controls, and this one

52:11 Speaker 4: also will animate. It's not quite as responsive as the

52:14 Speaker 4: streams if you animate those, but the slider here can

52:20 Speaker 4: be manually dragged to whatever time period you're looking at,

52:24 Speaker 4: like for instance, I happen to stop on July nineteen

52:27 Speaker 4: ninety nine, and this is what the map.

52:30 Speaker 5: Looked like at this point.

52:36 Speaker 4: That concludes the main components of our visualization parts. I'm

52:42 Speaker 4: going to refresh the page and get back to the

52:45 Speaker 4: defaults view. So if I refresh the page and get

52:55 Speaker 4: back to the default view, we'll start looking at how

53:01 Speaker 4: you use the hydro viewer to extract data and get

53:03 Speaker 4: plots and look at things numerically. But I mentioned already

53:08 Speaker 4: that this map is multi scale, meaning that there's more

53:12 Speaker 4: rivers become visible as you zoom in. So if I

53:15 Speaker 4: start really zoomed out and I try and click in

53:18 Speaker 4: South America, the map is going to drag me into

53:21 Speaker 4: a much higher resolution or a much higher zoom so

53:25 Speaker 4: that I can see the rivers at their full detail

53:29 Speaker 4: being presented to me. And only when you're this far

53:32 Speaker 4: zoomed in will I be able to click on a

53:34 Speaker 4: stream and it will start pulling data for me. After

53:41 Speaker 4: you've clicked on a stream, you're going to see this

53:43 Speaker 4: modal appear. And there's two options at the top right

53:47 Speaker 4: for you to toggle between forecast data, which we're looking

53:50 Speaker 4: at now, and the other button takes me to look

53:53 Speaker 4: at the retrospective simulation data. I will start on the

53:56 Speaker 4: forecasts our main chart here. The top plot this hydrograph

54:03 Speaker 4: shows the predicted flow for the fifteen day period that

54:06 Speaker 4: forecast covers. I'm going to toggle off the return period

54:09 Speaker 4: by clicking on them for a minute so I can

54:12 Speaker 4: comment on the two main lines here, the black one

54:15 Speaker 4: labeled predicted flow, and this blue region labeled uncertainty. So

54:21 Speaker 4: the black line that's called predicted flow. If someone needs

54:26 Speaker 4: just a single number for what flow did we predict

54:30 Speaker 4: for a deterministic application, or you need to put this

54:33 Speaker 4: into an equation somewhere, this black line is labeled predicted flow.

54:39 Speaker 5: Is the median of the ensemble.

54:41 Speaker 4: As Rachel already mentions, this is a fifty one member

54:45 Speaker 4: ensemble and at every time step, so every three hours

54:50 Speaker 4: across the fifteen days, there's going to be fifty one

54:52 Speaker 4: different numbers, and this black line is the median at.

54:56 Speaker 5: Each of the time steps.

54:57 Speaker 4: It's the representative value that we would suggest you use.

55:02 Speaker 4: You can try averaging, you can use other means to

55:05 Speaker 4: reduce the ensemble down, but the black line that we

55:09 Speaker 4: put out is a good starting place is the median.

55:13 Speaker 4: The blue region labeled uncertainty. It corresponds to the twentieth

55:18 Speaker 4: percentile and the eightieth percentile at each time step, so

55:22 Speaker 4: that's sixty percent. So it's sixty percent of the data

55:25 Speaker 4: fall fell within that range. So it's more likely than not,

55:30 Speaker 4: according to our predictions, that the actual flow that will

55:34 Speaker 4: occur at whatever the lead time is will be within

55:38 Speaker 4: that boundary this blue shaded region. It's normal for the

55:45 Speaker 4: blue shaded region to get smaller as if you're looking

55:48 Speaker 4: at bigger watersheds, so there's more upstream drainage area. And

55:53 Speaker 4: that's because the for instance, at this flow right now,

55:57 Speaker 4: we're predicting seven thousand meters per seconds. This may or

56:02 Speaker 4: may not be biased to remember, we haven't done biased

56:04 Speaker 4: correction yet, but we're predicting seven thousand cubic meters per

56:07 Speaker 4: second on June sixteenth at three am. According to this

56:12 Speaker 4: and that, it's hard for there to be a very

56:16 Speaker 4: large window of uncertainty relative to larger and larger flows.

56:21 Speaker 4: But if you use this tool instead to look at

56:23 Speaker 4: a smaller tributary with a much smaller drainage area, it's

56:28 Speaker 4: normal for this blue area to become bigger because there's

56:31 Speaker 4: less certainty, and smaller uncertainties propagate into a larger relative

56:37 Speaker 4: amount of error that might be predicted. I'll turn the

56:42 Speaker 4: return periods back on. Now.

56:44 Speaker 5: These are the same colors that we're used for the

56:48 Speaker 5: map this.

56:50 Speaker 4: They go from yellow to orange, red all the way

56:53 Speaker 4: up to purple. The map stops coloring at fifty, but

56:57 Speaker 4: the chart shows a one hundred year return period as well.

57:01 Speaker 4: The return periods help provide context for how big the

57:07 Speaker 4: flow is. If you don't have a strong concept already

57:11 Speaker 4: of what's normal in a given river, you might look

57:14 Speaker 4: at four thousand cubic meters per second and say is

57:17 Speaker 4: that a lot or a little? And these return periods

57:20 Speaker 4: help you understands. At a two year return period, so theoretically,

57:26 Speaker 4: every other year, perhaps you might expect this river to

57:29 Speaker 4: reach this flow. So it's pretty common, but it's hydrologically

57:34 Speaker 4: we start to think of that as the beginning of

57:36 Speaker 4: If it happens just every other year, that's as a

57:40 Speaker 4: random guess, not knowing specifics about the river, that's when

57:44 Speaker 4: flooding would start to occur. The five, ten, twenty five

57:51 Speaker 4: and as they go up these larger return period values,

57:54 Speaker 4: as the flow is predicted to cross into one of those,

57:57 Speaker 4: it's more likely that the flow is going to cause flooding,

58:02 Speaker 4: and that the extent of how wide the flood will

58:05 Speaker 4: be will get bigger. Another way to look at that

58:07 Speaker 4: same information is in the table just below this. For

58:11 Speaker 4: each of the fifteen days in the prediction, you'll see

58:15 Speaker 4: the line up until about June eighteenth and on the

58:20 Speaker 4: graph is entirely within the Orange region. And this tables

58:25 Speaker 4: tells you specifically the percentage there are more ensemble members

58:31 Speaker 4: being plotted that are rather not being plotted, And this

58:35 Speaker 4: table tells you the percentage that are within each of

58:39 Speaker 4: these thresholds during each of the days in the fifteen columns.

58:47 Speaker 4: The last thing the comment on for the forecast is

58:49 Speaker 4: if you want to go back in time and look

58:52 Speaker 4: at what we projected in the future rather in the past.

58:55 Speaker 4: There's a date picker at the top. Rachel mentioned already

58:59 Speaker 4: that we have or the old forecast. We're in June

59:02 Speaker 4: twenty twenty sixth right now as of the day that

59:05 Speaker 4: we're presenting, the first forecast that we have is July

59:10 Speaker 4: one of twenty twenty four, so we're about to hit

59:14 Speaker 4: two full years of forecasts being stored. And I want

59:18 Speaker 4: to comment on refecasts. The record that you can go

59:23 Speaker 4: back and obtain is whatever we forecasted on that day

59:27 Speaker 4: with no deviations. Even if we found an error in

59:31 Speaker 4: the model later or going back, we determined that the

59:34 Speaker 4: model didn't do very well, and we could have adjusted

59:37 Speaker 4: certain parameters to make it perform better. We have not

59:40 Speaker 4: gone back and done reforecasts, so this past archive perhaps

59:45 Speaker 4: the farther back in time you go. There might be

59:47 Speaker 4: errors in your specific watershed which later we've discovered and addressed,

59:52 Speaker 4: but the forecast that you obtain for those dates in

59:56 Speaker 4: the past will not reflect those edits.

59:58 Speaker 5: There's no refecasting that happened.

1:00:02 Speaker 4: Last thing that I'll show on the charts is there's

1:00:05 Speaker 4: a tab to toggle over to the retrospective. Rachel mentioned

1:00:10 Speaker 4: that there's hourly average is the native resolution, and you

1:00:14 Speaker 4: can obtain that still. To keep the plot simpler, we

1:00:17 Speaker 4: just show the daily average and the monthly averages here.

1:00:20 Speaker 4: You can click on these to toggle them, and it's

1:00:24 Speaker 4: by default zooms into. Only a few years of data

1:00:27 Speaker 4: are visible. The slider on the bottom, you can drag

1:00:31 Speaker 4: the ends to shrink or grow the window that's visible. Normally,

1:00:35 Speaker 4: I'd suggest you keep this shrunk down to five or

1:00:38 Speaker 4: ten years if you want to be able to see

1:00:39 Speaker 4: what's really going on in the data, otherwise it becomes

1:00:42 Speaker 4: really spiky as it's all visible. There are some other

1:00:46 Speaker 4: plots here, and there's several more of these that are

1:00:51 Speaker 4: available if you go back to the settings and toggle

1:00:53 Speaker 4: on the additional plots. The only other one that's on

1:00:57 Speaker 4: by defaults at least at the time that we're in

1:01:00 Speaker 4: our workshop today, as this plots that shows the Hydross categories,

1:01:07 Speaker 4: and if I toggle on each of the individual categories here,

1:01:10 Speaker 4: these shades of blue and red correspond to the same

1:01:13 Speaker 4: thing that the maps show if we toggle on the

1:01:16 Speaker 4: Hydross layer, and by default, just whatever year you're in

1:01:21 Speaker 4: is visible. So that's twenty twenty six currently. But if

1:01:26 Speaker 4: you're studying this watershed and you're using remote sensing from

1:01:30 Speaker 4: twenty years ago or something, you might say two thousand

1:01:33 Speaker 4: and nine is the year i'm interested in, so you

1:01:35 Speaker 4: can scroll down finds that year on the legend entries

1:01:41 Speaker 4: over here, and that will appear on the map. Last

1:01:47 Speaker 4: thing to help you obtain data is on the top right.

1:01:51 Speaker 4: There's a download icon which will prompt you to download

1:01:55 Speaker 4: the CSBS. You can click on that and download either

1:01:58 Speaker 4: of the forecast data or the retrospective data or both

1:02:03 Speaker 4: and if you like the style of plots and you

1:02:06 Speaker 4: want to just borrow the plots that we provide in

1:02:08 Speaker 4: the viewer, every single one of the plots on the

1:02:11 Speaker 4: top rights of when you hover your mouse over, it

1:02:14 Speaker 4: will prompt you to download this plot as a p

1:02:16 Speaker 4: ANDNG and that's for all of them.

1:02:21 Speaker 5: A few other.

1:02:24 Speaker 4: Sort of informational items about the Hydro viewer is up

1:02:30 Speaker 4: on the top right. We have the ability for you

1:02:32 Speaker 4: to save rivers. This bookmark icon will show you by default,

1:02:37 Speaker 4: it's just a dozen or so of the biggest rivers

1:02:41 Speaker 4: in the worlds and you'll see the ID number that

1:02:44 Speaker 4: corresponds to that river and the name, and you can

1:02:48 Speaker 4: shortcut having to browse through the map to find that

1:02:51 Speaker 4: river by just hitting this chart icon and it will

1:02:54 Speaker 4: go load the information for that river, and you'll see

1:02:58 Speaker 4: that when opening up one of the rivers from that menu,

1:03:02 Speaker 4: the heart is is toggled here.

1:03:06 Speaker 5: If for.

1:03:08 Speaker 4: Your application, there's a river that you're gonna want to

1:03:11 Speaker 4: visit all the time, perhaps it corresponds to a gauged location,

1:03:15 Speaker 4: or it's a monitoring point, an infload, or a reservoir,

1:03:19 Speaker 4: something like that. I just clipped on another new river

1:03:22 Speaker 4: and I can hit the heart icon and it will

1:03:26 Speaker 4: auto populate the river ID and give me the chance

1:03:28 Speaker 4: to save it with a name, so I could say,

1:03:31 Speaker 4: maybe this corresponds to gauge one, two.

1:03:34 Speaker 5: Three, four, of course that's the name.

1:03:37 Speaker 4: Can be anything that's that's meaningful to you here and

1:03:39 Speaker 4: I hit add bookmark and the heart turns red. And

1:03:43 Speaker 4: now that's something that I can frequently go back and

1:03:47 Speaker 4: find without needing to browse through the map. If there's

1:03:49 Speaker 4: just a list of, you know, twelve rivers that you

1:03:53 Speaker 4: always want to check on, or that you're going to

1:03:55 Speaker 4: visit a couple times throughout the frequency of your projects,

1:03:59 Speaker 4: this is a good way to convenience yourself on not

1:04:03 Speaker 4: needing to browse the maps as many times and speed

1:04:05 Speaker 4: things up. That concludes my presentation on the Hydroviewer. Thank

1:04:13 Speaker 4: you so much for being with us. I'm looking forward

1:04:16 Speaker 4: to taking your questions and I'll turn the time back

1:04:18 Speaker 4: over to our hosts.

1:04:24 Speaker 1: Thank you so very much Rachel and doctor Hale's for

1:04:27 Speaker 1: your presentations and demonstration of Hydroviewer. This is going to

1:04:32 Speaker 1: be an extremely useful tool for flood prediction and preparedness.

1:04:37 Speaker 1: This brings us to the end of today's session. Just

1:04:40 Speaker 1: to summarize what we saw today, we learned about geoglows.

1:04:44 Speaker 1: It's a partnership initiative among one hundred plus national government

1:04:48 Speaker 1: and organizations to deliver accurate, open and accessible hydrological predictions

1:04:54 Speaker 1: on a global scale based on the ECMWF ERA five

1:04:59 Speaker 1: and integrated for class system data. The ERA five models

1:05:02 Speaker 1: include insitute and multiple satellite data assimilation. Geogloss uses a

1:05:07 Speaker 1: hydrologic model with inputs from models, gages and satellites to

1:05:11 Speaker 1: calculate runoff. Distributed vertically uses a river routing model to

1:05:16 Speaker 1: get river discharge and provides retrospective and forecast hydrographs for

1:05:21 Speaker 1: global river segments. Geoglose provides written periods, monthly and annual

1:05:26 Speaker 1: averages and maximums and flow duration curves from the retrospective data.

1:05:34 Speaker 1: Geoglos data visualization can be done through hydro Viewer and

1:05:38 Speaker 1: RGIS webmap and query and download data from Amazon Web service.

1:05:46 Speaker 1: This brings us to the conclusion of this training on

1:05:49 Speaker 1: monitoring and predicting floods using our third durations for planning

1:05:52 Speaker 1: and preparedness. Just to summarize this training, this was a

1:05:58 Speaker 1: three part training that focused on flood detection on land

1:06:02 Speaker 1: from remotely sensed optical and microwave data, imagery and flood modeling,

1:06:08 Speaker 1: and we were discharge prediction using weather, hydrology and river

1:06:11 Speaker 1: routing models. In part one we talked about global flood product.

1:06:16 Speaker 1: Second part focused on operodynamic surface water extent and today

1:06:21 Speaker 1: we talked about geogloss river forecasts system. So part one

1:06:28 Speaker 1: Global flood product was summarized by doctor Dance. LABA was

1:06:33 Speaker 1: based on Terra and aquamdis and Noah twenty and twenty

1:06:37 Speaker 1: one vers optical data. It uses red infrared and shortwave

1:06:41 Speaker 1: infrared reflectancies to detect water. Identification of recurring floods is

1:06:48 Speaker 1: based on twenty two year flood mask historical data. Also

1:06:52 Speaker 1: its global two hundred and fifty meters respecial resolution and

1:06:56 Speaker 1: near real time and twenty three years of archived data

1:06:59 Speaker 1: are available from Terra and AQUA multilook composites of one,

1:07:05 Speaker 1: two and three day to assign threshold based water detection

1:07:08 Speaker 1: is used. Terrent shadow and cloud shadow corrections are applied

1:07:13 Speaker 1: to remove false positive in water detection and composite product

1:07:17 Speaker 1: recommended by PERSIST for persistent cloudiness and longer lasting larger floods.

1:07:24 Speaker 1: We saw an example or several examples of how to

1:07:28 Speaker 1: look at worldview for data visualization of global flood product

1:07:33 Speaker 1: from different composites, and also how to access data from

1:07:37 Speaker 1: NASA Earth Data and Large Deck, as we reviewed earlier. Today,

1:07:43 Speaker 1: Part two was about opera dynamic surface what extent, presented

1:07:47 Speaker 1: by doctor Renato Franso. He described how operodynamic surface water

1:07:52 Speaker 1: Extent or DSWX has two products, one from optical imagery

1:07:57 Speaker 1: harmonized land set and Sentinel two data and based on

1:08:01 Speaker 1: SAR Centinel one sr set. So these two are DSWXHLS

1:08:06 Speaker 1: and DSWXS one products. Both have relatively high resolution thirty

1:08:12 Speaker 1: meters their near global Both DSWXHLS and S one products

1:08:17 Speaker 1: are available from twenty twenty three to present, HLS from

1:08:21 Speaker 1: April and S one from December. Temporal resolutions are also

1:08:27 Speaker 1: given here a three day for HLS and six to

1:08:30 Speaker 1: twelve for S one. We saw that DSWXHLS data should

1:08:35 Speaker 1: be composited over multiple days or combined with DSWXS one

1:08:40 Speaker 1: data when there are persistent clouds, and we also saw

1:08:44 Speaker 1: how to merge DSWX data layers in GIS for improved

1:08:50 Speaker 1: flood detection capability. There was a demonstration of how to

1:08:54 Speaker 1: visualize DSWX data in NASA Worldview and data Access to data. Finally,

1:09:01 Speaker 1: we had presentation and demonstration from Rachel Megafin and Riley Hales.

1:09:07 Speaker 1: They talked about GEOGLOWS, which is based on the acmw

1:09:11 Speaker 1: ERA five and Integrated Forecast System, uses in situ and

1:09:15 Speaker 1: multiple satellite data assimilation in these models and also uses

1:09:20 Speaker 1: hydrologic model and a river routing model. GEOGLOWS combines models,

1:09:25 Speaker 1: gauge data and satellite data and provides retrospective and forecasts.

1:09:30 Speaker 1: Hydrographs for fifteen day forecasts, written periods, monthly and annual

1:09:34 Speaker 1: averages and maximums and flow duration curves from the retrospective

1:09:38 Speaker 1: data also available from a hydro viewer. There's an example

1:09:44 Speaker 1: I've picked up. This is the cares in Ukraine. This

1:09:47 Speaker 1: is the same flood that we saw in Global Flood

1:09:52 Speaker 1: Product and this is the hydrograph or discharged time series

1:09:56 Speaker 1: for that and you can see from hydro viewer that

1:10:00 Speaker 1: you do see increased discharge during that period when we

1:10:04 Speaker 1: saw flooding in this Careson area. We also found out

1:10:10 Speaker 1: that RTIs webmap can also be used for visualizing this

1:10:15 Speaker 1: data and data can be accessed through Amazon Web Service. Overall,

1:10:20 Speaker 1: there are several benefits and limitations to what we learn

1:10:23 Speaker 1: in this training. For example, combining multiple satellite and incituo

1:10:29 Speaker 1: observations with models provides predictive capability for stream flow enabling

1:10:34 Speaker 1: risk assessment, planning and preparedness for floods, such as using

1:10:39 Speaker 1: weather forecast, ideology and river outing models. In geoglows in

1:10:42 Speaker 1: which multiple observations are used, then remote sensing observations both

1:10:48 Speaker 1: active microwave and passive optical from multiple satellites and sensors

1:10:53 Speaker 1: such as Terra and Aqua Modis twenty twenty one years,

1:10:57 Speaker 1: lands At eight and nine only Sentinel to MSI and

1:11:01 Speaker 1: sentinel ones are These are satellites and sensors used in

1:11:04 Speaker 1: Global Flood Product and Opera DSWX. They enable the detection

1:11:09 Speaker 1: of inundation and water logging during and after flood events,

1:11:13 Speaker 1: supporting response and recovery activities. SAR observations provide all weather

1:11:19 Speaker 1: capability to detect surface water, while optical observations cannot see

1:11:23 Speaker 1: through clouds and are therefore unable to detect surface water

1:11:26 Speaker 1: in overcast conditions. The special and temporal resolutions of these

1:11:31 Speaker 1: flood products vary by satellites and sensors, such as we

1:11:35 Speaker 1: saw that opera products are thirty meters, global flood product

1:11:39 Speaker 1: are two one hundred and fifty meters and resolution Temporal

1:11:43 Speaker 1: resolution varies from one to twelve days, and they carry

1:11:47 Speaker 1: uncerties due to various factors such as orange shadow, cloud shadow,

1:11:52 Speaker 1: and surface types. Monitoring flood potential using jeoglos alongside merged

1:11:58 Speaker 1: flood products from multiple optical inside data sets would greatly

1:12:02 Speaker 1: enhance flood related decision making activities. So the message is

1:12:07 Speaker 1: that we can use beneficial information from different satellites and

1:12:12 Speaker 1: sensors and models and combine them to be better prepared

1:12:16 Speaker 1: for flood related decision making. The homework assignment is posted

1:12:23 Speaker 1: today on our training web page and answers must be

1:12:26 Speaker 1: submitted via Google Forms and the homework is due on

1:12:29 Speaker 1: ninth of July and you will receive a certificate of

1:12:33 Speaker 1: completion approximately two months after the completion of this course

1:12:38 Speaker 1: if you attended all live webinars and if you complete

1:12:41 Speaker 1: the homework assignment by the due date. So we did

1:12:47 Speaker 1: exercise one and two after our previous sessions. Today the

1:12:51 Speaker 1: exercise will focus on using geoglows outfestream flow prediction using

1:12:56 Speaker 1: hydro Viewer. We will be looking at the same flood

1:12:59 Speaker 1: caase in South Sudan between fifteenth of May and thirty

1:13:03 Speaker 1: first of May, so you will be looking at streamflow

1:13:07 Speaker 1: or discharge using geoglows RFS. So about the homework exercises,

1:13:14 Speaker 1: we had three exercises counting today's exercise, first one on

1:13:19 Speaker 1: global flood product then a dynamic surface water extent. Both

1:13:22 Speaker 1: these were through NASA Worldview and today you'll be using

1:13:26 Speaker 1: hydro viheer to look at this South Sudan floodcase and

1:13:30 Speaker 1: some of the homework questions you will find based on

1:13:33 Speaker 1: these exercises. Once again, we want to thank all our

1:13:39 Speaker 1: guest speakers for their excellent presentations, demonstrations and contribution to

1:13:44 Speaker 1: this training and doctor Daniel Slayback, Doctor Renetto Frasso, Rachel

1:13:49 Speaker 1: mcaffin and doctor Riley Hales. We thank you all for

1:13:53 Speaker 1: your contribution to this training. It's the contact information for

1:14:00 Speaker 1: Rachel and Riley if you have any specific questions about Geoglows,

1:14:04 Speaker 1: and you can always contact us at our set with

1:14:08 Speaker 1: any questions that you have. Here is our set website

1:14:12 Speaker 1: with a lot of information you can find about different

1:14:15 Speaker 1: trainings and our set YouTube has recordings of all our trainings.

1:14:20 Speaker 1: For questions, comments, or to share how you have applied

1:14:24 Speaker 1: our trainings to your work or studies, please email NASA

1:14:29 Speaker 1: dot r SET at gmail dot com and join our

1:14:33 Speaker 1: mailing list to stay up to date on our latest trainings,

1:14:36 Speaker 1: and visit our contact page to subscribe. These are the

1:14:41 Speaker 1: resources we used in this training and share your thoughts.

1:14:48 Speaker 1: Within the day or two, we will send you an

1:14:50 Speaker 1: invitation to complete a short online survey and your feedback

1:14:55 Speaker 1: helps us improve the RSET program. Participation is optional and

1:15:00 Speaker 1: all responses are confidential. Survey data are crucial to helping

1:15:05 Speaker 1: us understand how to better meet your needs, and we

1:15:07 Speaker 1: want to help you use our observation data more effectively.

1:15:11 Speaker 1: We read every survey comment, so please share your thoughts.

1:15:15 Speaker 1: The survey will come from no Reply at Alchemera dot com,

1:15:20 Speaker 1: so this is not a spam email that you will

1:15:23 Speaker 1: receive from this address, so please we request that you

1:15:26 Speaker 1: complete the survey and your feedback is very important to us.

1:15:31 Speaker 1: With that, we want to thank you all for attending

1:15:34 Speaker 1: this training session and now we will go to our

1:15:37 Speaker 1: question and answer session with Rachel and Riley for geoglows. Yeah,

1:15:51 Speaker 1: thank you everyone for attending this training series and also

1:15:55 Speaker 1: our speakers for today, Rachel Meguffin and Riley Hale. We'll

1:16:01 Speaker 1: go to the question and answer session now, I'll read

1:16:07 Speaker 1: the questions and our guest speakers will help. So the

1:16:14 Speaker 1: first question is does you close provide data in coastal

1:16:18 Speaker 1: area areas or island nations and you can unmute and

1:16:24 Speaker 1: answer the question, thank you.

1:16:29 Speaker 4: Yes, we do have coverage on some islands, not all.

1:16:34 Speaker 4: The question or the answer that's visible here is that

1:16:38 Speaker 4: some of the foreseeings and data sets we use have

1:16:41 Speaker 4: varying resolutions. One additional clarification I can add on to

1:16:46 Speaker 4: that is that the dem products, the elevation products and

1:16:52 Speaker 4: things that are used to derive where the catchment boundaries

1:16:54 Speaker 4: are just we're not available on all islands. It is

1:16:57 Speaker 4: most places, but not all of them. I direct you

1:17:01 Speaker 4: to look around on the web viewer to see if

1:17:04 Speaker 4: the islands that you're interested in is covered.

1:17:08 Speaker 1: Great, thank you. A second question is can this geoglosse

1:17:11 Speaker 1: modeling be applied to a single basin.

1:17:18 Speaker 4: Yes, it can. The products that we use are all

1:17:22 Speaker 4: open source. The streams we provide the net forcings are

1:17:28 Speaker 4: from ECMWF and those don't require a license to obtain.

1:17:32 Speaker 4: You can either take our pre computed answer using these methods,

1:17:40 Speaker 4: or if you want to use the if you have

1:17:43 Speaker 4: some reason you need to run that yourself, or you

1:17:46 Speaker 4: want to apply some modification to any of the steps,

1:17:50 Speaker 4: then that you can just download the data from the

1:17:55 Speaker 4: relevant places. You'll have to do some additional work on

1:17:58 Speaker 4: your own to modify that odes to work on your

1:18:01 Speaker 4: area of interest, but everything you would need is open

1:18:04 Speaker 4: source to do that.

1:18:07 Speaker 1: Great, thank you. Question three is I'm wondering about the

1:18:11 Speaker 1: hydrological model used in geoglows WIF hydro lease flood.

1:18:21 Speaker 4: Yeah, so this question is asking, I believe about how

1:18:25 Speaker 4: we calculate the runoff volumes from the meteorology. This we

1:18:30 Speaker 4: use products from ECMWFS Integrated Forecast System i f S,

1:18:36 Speaker 4: and the land surface component of that is done with

1:18:39 Speaker 4: the model called h TESSEL and that is we take

1:18:47 Speaker 4: the answer straight from that. We apply modeling downstream of it,

1:18:50 Speaker 4: but our main source of truth for what the runoff

1:18:54 Speaker 4: depth is comes from that h TESLIL model.

1:18:59 Speaker 1: Next question is where we can download the hydrography shape file.

1:19:06 Speaker 4: Yeah, so there are several questions asking about where defines

1:19:09 Speaker 4: data for this one in particular, but also the others.

1:19:13 Speaker 4: I'll direct you to our training websites. That's training dot

1:19:17 Speaker 4: gogos dot org, and that's that'll be in this answer documents,

1:19:21 Speaker 4: and it's in the slides and in various places, and

1:19:25 Speaker 4: you'll find tables that direct you for whichever product you're

1:19:29 Speaker 4: looking for. We'll give you a description of the name

1:19:32 Speaker 4: and what file formats and so on and so forth

1:19:36 Speaker 4: for where you can download.

1:19:37 Speaker 1: That question file. Is there any bias correction too available?

1:19:46 Speaker 4: Yeah, So this can happen in two different ways. One,

1:19:52 Speaker 4: we provide code, and we've have papers describing a method

1:19:56 Speaker 4: where if you have your you want to bring your

1:19:58 Speaker 4: own gauge data, you can apply the method that we

1:20:02 Speaker 4: use to get our model data and your gauge data

1:20:07 Speaker 4: that you provide and calculate what the bias corrected flow

1:20:10 Speaker 4: would be. Alternatively, option two, there are some restrictions on

1:20:17 Speaker 4: what data we are and are not allowed to use

1:20:21 Speaker 4: for a broad global application, but we do have in

1:20:27 Speaker 4: the hydro viewer there's an option to toggle turning on

1:20:31 Speaker 4: using bias correction. This is also available in code, but

1:20:34 Speaker 4: go to the hydrid viewer and in the settings you

1:20:37 Speaker 4: can toggle on that you want to use the experimental

1:20:39 Speaker 4: bias correction and we will do our best guess of

1:20:43 Speaker 4: what the bias corrected flow is everywhere, using what gauge

1:20:47 Speaker 4: data we do have that is appropriately licensed and so on.

1:20:54 Speaker 1: Thank you. Question six, can you download the data from

1:20:57 Speaker 1: the living at last or is it just a visualizer?

1:21:03 Speaker 4: Yeah, so the living atless layer is really meant to

1:21:05 Speaker 4: just be for visualization. You can go browse all the

1:21:09 Speaker 4: attributes that are applied to the streams that's used for

1:21:13 Speaker 4: the styling and animation, but really it's meant just for visualization.

1:21:17 Speaker 4: There are plenty of other code tools and notebooks other

1:21:24 Speaker 4: things in the training to help you download if you

1:21:26 Speaker 4: need to get large volumes of data downloaded.

1:21:31 Speaker 1: Question seven, is there any approach in this method that

1:21:34 Speaker 1: estimates regulated basins?

1:21:39 Speaker 4: So this question also, I think has kind of two answers.

1:21:42 Speaker 4: Some of the most impactful regulations could be extracting water

1:21:49 Speaker 4: for irrigation or running a canal or something. It's really

1:21:52 Speaker 4: difficult to get information about what volumes of water are

1:21:57 Speaker 4: used there. We do not have a way to to

1:22:00 Speaker 4: make that, but there are other ways, like how lakes

1:22:05 Speaker 4: are managed or reservoirs or dams, those are sometimes able

1:22:10 Speaker 4: to or do have the information that we would need

1:22:13 Speaker 4: to make an estimate on that. The current model version

1:22:16 Speaker 4: does not do an especially good job of accounting for

1:22:20 Speaker 4: all those things, but Rachel and I are collaborating on

1:22:25 Speaker 4: doing better in the next model version. How will make

1:22:27 Speaker 4: the model accounts for the effects of reservoirs? So that's

1:22:32 Speaker 4: some some yes, and some know that we that we

1:22:36 Speaker 4: whether or not we're able to account for regulations.

1:22:40 Speaker 1: Great question eight is, since the ensemble forecasting model generates

1:22:46 Speaker 1: multiple possible forecast outcomes, is there a mechanism for incorporating

1:22:51 Speaker 1: subsequent observational data to update and define earlier forecasts, thereby

1:22:57 Speaker 1: improving prediction accuracy over time.

1:23:01 Speaker 4: Yeah, this is a really good question. That this is

1:23:04 Speaker 4: getting a data assimilation essentially. The biggest problem is that

1:23:10 Speaker 4: countries that have often the greatest need for this kind

1:23:15 Speaker 4: of supplemental global hydrology modeling often don't have very many gauges,

1:23:21 Speaker 4: and those that do typically don't have their data publicly

1:23:26 Speaker 4: accessible on the Internet, so that the lag time between

1:23:30 Speaker 4: when the measurement is made and when it would be

1:23:33 Speaker 4: accessible to me to try to do some sort of

1:23:36 Speaker 4: ingestion run A common filter or something like that is

1:23:40 Speaker 4: pretty long, so we don't primarily because it's just not

1:23:45 Speaker 4: feasible the computationally, the data doesn't become accessible on a

1:23:49 Speaker 4: short enough lead time to try doing that. However, as

1:23:53 Speaker 4: I said before, the code and data sets and everything

1:23:57 Speaker 4: you would need to go back in time and do

1:24:00 Speaker 4: a re forecast is all open source and available, So

1:24:03 Speaker 4: even though we don't do that at the global scale,

1:24:05 Speaker 4: that is possible, and you can check out our code

1:24:10 Speaker 4: and send some emails to chat more about that if

1:24:13 Speaker 4: you would like.

1:24:16 Speaker 3: Nice.

1:24:16 Speaker 1: Question nine is has geo gloves been evaluated or benchmarked

1:24:21 Speaker 1: against other operational flood forecasting platforms such as IFAs or

1:24:26 Speaker 1: Google flood up.

1:24:30 Speaker 5: Yes we have.

1:24:32 Speaker 4: Sometimes those are published, and I gave you a link

1:24:35 Speaker 4: to the publications there. That's again on trading that geobos

1:24:40 Speaker 4: you'll find a publication section.

1:24:43 Speaker 5: Also.

1:24:44 Speaker 4: We do similar kinds of benchmarkings and comparisons whenever we're

1:24:49 Speaker 4: participating in w MOO activities, including some that are ongoing

1:24:53 Speaker 4: right now, so I'll direct you to those publications to

1:24:57 Speaker 4: see if that gives you the answer that you're looking for.

1:25:02 Speaker 1: Question ten is does geoglows aim to serve as a

1:25:05 Speaker 1: flood forecasting tool or stream flow stetus and forecasting data provider.

1:25:13 Speaker 4: Primarily our use cases and that we've collaborated with country

1:25:20 Speaker 4: hydromed agencies has been for flooterly warning applications. We do

1:25:25 Speaker 4: have examples of applying this data and building tools or

1:25:29 Speaker 4: apps or other things around other applications such as water

1:25:35 Speaker 4: quality or irrigation or reservoir management, but those have are

1:25:41 Speaker 4: more in the minority to the majority of our collaborations

1:25:46 Speaker 4: have revolved around flood forecasting. We are looking to expand

1:25:50 Speaker 4: the number of pre made things that we provide on

1:25:53 Speaker 4: that on those other application areas. Though.

1:25:58 Speaker 1: Interesting elemon is how are the RFS simulations validated? Are

1:26:04 Speaker 1: the calibrator or validated using hydrometric stations?

1:26:10 Speaker 4: Yeah, so we have I've already alluded to this slightly

1:26:13 Speaker 4: that we have a lot of gauge data. It's growing.

1:26:17 Speaker 4: But the some of the problems in publishing about all

1:26:21 Speaker 4: of those gauge measurements is that they're not all permissively

1:26:24 Speaker 4: licensed for us to publish validation against that, that would

1:26:28 Speaker 4: necessarily mean we're giving away some of those gauge measurements

1:26:32 Speaker 4: and they're being put on plots and things, and that's

1:26:35 Speaker 4: not always allowed, So we do validate against a very

1:26:38 Speaker 4: large set of inciti measurements. Not all of them make

1:26:44 Speaker 4: it into publications, and not all of them can be shared,

1:26:46 Speaker 4: so there's some difficulty in how those gets share or

1:26:51 Speaker 4: how we're able to share that with other people.

1:26:55 Speaker 1: Question twelve, is there any thought to creating our library

1:27:00 Speaker 1: for jyo glows?

1:27:05 Speaker 4: Yeah? This I get this question occasionally. The geoglows main

1:27:10 Speaker 4: team does not have any our developers on it, so

1:27:15 Speaker 4: we're not in a great position to create that ourselves.

1:27:18 Speaker 4: We have a Python a job script job excuse me,

1:27:21 Speaker 4: job a script version. If you would like to take

1:27:24 Speaker 4: the lead in creating in our version, I'm definitely open

1:27:27 Speaker 4: to having that conversation.

1:27:32 Speaker 1: Question thirteen, is there any opportunity to get hand to

1:27:36 Speaker 1: hand training on setting the geoglows model model?

1:27:43 Speaker 4: So I would say yes, the I would direct you

1:27:48 Speaker 4: first to training dot geoglows, and there's some links here

1:27:51 Speaker 4: to a Google group and we have a YouTube channel

1:27:54 Speaker 4: with some webinars, So I'd hope that this covers a

1:27:58 Speaker 4: lot of the the questions and upfronts kind a direction

1:28:05 Speaker 4: that you're looking for, but you would need to reach

1:28:08 Speaker 4: out to us to discuss if you're looking for live

1:28:11 Speaker 4: trainings and things like that.

1:28:15 Speaker 1: Thank you. Question fourteen, how does geoglows differ from glow fasts?

1:28:24 Speaker 4: So this is a really good question, getting right to

1:28:26 Speaker 4: the core of how our model operates. What Rachel was

1:28:30 Speaker 4: presenting at at the beginning. We in rfs and glow

1:28:35 Speaker 4: fasts share the same ECMWF MET predictions that ensemble prediction

1:28:41 Speaker 4: the core. We're both built on the same meteorology predictions.

1:28:46 Speaker 4: As the farther you go down the line, though, we

1:28:49 Speaker 4: start to diverge on the different additional modeling steps that

1:28:54 Speaker 4: are applied. Because of that, that are met forcings at

1:28:59 Speaker 4: the core are the same, so we'll often show very

1:29:02 Speaker 4: similar signals. It's sometimes called or shapes on the graphs.

1:29:07 Speaker 4: If glow fast predicts it, we will probably also predict it,

1:29:11 Speaker 4: and vice versa because we share that same core, but

1:29:15 Speaker 4: because we use different modeling methods downstream of the mets

1:29:20 Speaker 4: and the land surface parts, they won't be identical, but

1:29:23 Speaker 4: they'll often be very complimentary.

1:29:29 Speaker 1: Question fifteen, incheoglows? Can we searched for a river by name?

1:29:36 Speaker 4: Yeah, this is another really common question we get. There's

1:29:40 Speaker 4: some difficulty in getting a really good authoritative source for

1:29:43 Speaker 4: what river names are. So in general, I would say no,

1:29:47 Speaker 4: you should use the ID numbers because that is guaranteed

1:29:51 Speaker 4: to be unique everywhere. For instance, there could be multiple

1:29:54 Speaker 4: rivers names the Colorado River or the fill in the blank.

1:29:59 Speaker 4: There's probably multiple rivers' names are sharing the same name.

1:30:04 Speaker 4: We are exploring ways to try to get some river

1:30:08 Speaker 4: name attributes applied to the streams so that you could

1:30:11 Speaker 4: do some searching and filtering, but that's not available just yet.

1:30:18 Speaker 1: Question sixteen Real Geo Clowse hydro Viewer and the Hissory

1:30:22 Speaker 1: Living Atlas eventually visualized trout conditions or below normal flows.

1:30:28 Speaker 1: I assume that pulling the data and Python would allow

1:30:31 Speaker 1: for this type of visualization in addition to flooding in

1:30:35 Speaker 1: local applications.

1:30:41 Speaker 4: So I think the answer is probably yes eventually. This

1:30:45 Speaker 4: is an ongoing area we're exploring right now. Actually, so

1:30:48 Speaker 4: I don't have an ability to promise specifically wet maps

1:30:53 Speaker 4: or charts or the timeline or anything like that, but

1:30:57 Speaker 4: I would say yes, we do want to have more

1:31:00 Speaker 4: on that front for now, though you'll have to download

1:31:04 Speaker 4: the data yourself and generate the visualization you're looking for.

1:31:10 Speaker 1: Wonderful, We'll take one more question, because we are almost

1:31:15 Speaker 1: at the at the end of our session time, how

1:31:20 Speaker 1: is the uncertainty computed?

1:31:26 Speaker 4: So the statistics for meteorology can get kind of complicated.

1:31:31 Speaker 4: The main thing for you to know is that the

1:31:34 Speaker 4: meteorologists would advise you to view the fifty one different predictions,

1:31:40 Speaker 4: which in turn we use to make fifty one different

1:31:43 Speaker 4: discharge predictions. Those should be treated as if they're equally probable,

1:31:49 Speaker 4: so that you can then do many many directions on

1:31:54 Speaker 4: how you want to analyze that and how you calculate uncertainty.

1:31:58 Speaker 4: I think this question is getting more at when on

1:32:01 Speaker 4: the graphs that we present, what's the blue shaded region?

1:32:05 Speaker 4: How is that uncertainty estimated? This is very unsophisticated, but

1:32:10 Speaker 4: it was done this way intentionally, so it's very simple

1:32:14 Speaker 4: and most readily understood by a big, wide array of

1:32:19 Speaker 4: audiences that have sometimes a great deal of statistical training

1:32:23 Speaker 4: and sometimes not very much at all. We take on

1:32:28 Speaker 4: every kid step, the values are sorted from max to minimum,

1:32:32 Speaker 4: high to low, and the middle sixty percents is that

1:32:37 Speaker 4: considered the uncertainty and that that checks out because the

1:32:41 Speaker 4: most extreme values tend to have very few of the

1:32:47 Speaker 4: members predicting that they happen, and the most common range

1:32:54 Speaker 4: of values sit in the middle. There are, of course,

1:32:57 Speaker 4: many other ways that you could go about estimating certainties.

1:33:00 Speaker 4: This one has chosen specifically because it's simple and it's

1:33:06 Speaker 4: understandable by a wide range of people.

1:33:10 Speaker 5: And the hydro viewer.

1:33:11 Speaker 4: Remove that option and not view the simplified forecast and

1:33:15 Speaker 4: view all the ensemble members and see different ways that

1:33:18 Speaker 4: uncertainty you can be presented. That's just not the default

1:33:22 Speaker 4: for the communication reasons.

1:33:27 Speaker 1: Great, thank you so much. So we have a few

1:33:32 Speaker 1: questions that we will be posting. We will be answering

1:33:35 Speaker 1: and then this question answered document will be posted on

1:33:39 Speaker 1: our training website in a week or so. With that,

1:33:43 Speaker 1: we want to thank our speakers once again, Rachel Megafwin, Ralei,

1:33:49 Speaker 1: thank you so much for your contribution, excellent presentations and demonstrations.

1:33:54 Speaker 1: This is going to be a very useful tool. So

1:33:57 Speaker 1: this brings us to the end of this training and

1:34:01 Speaker 1: we hope to see you at future our set training.

1:34:05 Speaker 1: So please subscribe to our listener or just go through

1:34:10 Speaker 1: the link and you will can subscribe to all the

1:34:15 Speaker 1: future upcoming trainings and stay up to date. At this point,

1:34:20 Speaker 1: I also want to thank our our set team, my

1:34:24 Speaker 1: colleague Sean McCartney, coordinator Natasha Johnson Griffin, and editor Maria Marabito.

1:34:31 Speaker 1: We also have instructional designer Kevin Fuel and our other

1:34:35 Speaker 1: coordinators brock Levin, Salvon, Salvin Hudson or Doy We thank

1:34:41 Speaker 1: you all and Sharry Morris so thank you for your

1:34:44 Speaker 1: help and thank you all for attending this training and

1:34:48 Speaker 1: we hope to see you soon. One last reminder, we

1:34:52 Speaker 1: have an exercise posted today using geoglows height of viewer

1:34:57 Speaker 1: and then some of the questions will.

1:34:59 Speaker 3: Be based on that.

1:35:01 Speaker 1: So you will have free question three exercises and one

1:35:04 Speaker 1: homework assignment and once you finish that and submit the

1:35:09 Speaker 1: homework by dud it, you will receive certificate in about

1:35:13 Speaker 1: eight weeks or so after the after today. So thank

1:35:18 Speaker 1: you all, and please do not forget to take the

1:35:20 Speaker 1: survey when you receive a survey link. We really look

1:35:25 Speaker 1: forward to your comments and thoughts for our future trainings,

1:35:29 Speaker 1: so thanks all once more,