Red River Basin Disaster Information Network

RECENT DEVELOPMENTS IN REMOTE SENSING
FOR DISASTER MONITORING

April 29, 1999 Online Workshop

Featuring

Julie Cranton
Environmental Scientist
Canada Centre for Remote Sensing

David Greenlee
Physical Scientist
U.S. Geological Survey's EROS Data Centre

Amy Sebring, Moderator
EIIP Virtual Forum

Amy Sebring: Welcome to the Red River room of the EIIP Virtual Forum!

If you did not see the email, Gene Krenz has had a family emergency and cannot be with us as previously scheduled. We will reschedule the workshop on the plan development by the Red River Basin Board when we know more about the situation.

We are very grateful to a couple of very good sports, who have "volunteered" to present an alternate program today on some of the recent developments in the area of remote sensing for disaster monitoring and its significance for the Red River basin. I will introduce them in just a moment, but first some quick instructions.

When you see a blue Web address, you can click on it and the referenced Web page should appear in a browser window. After the first one, the browser window may not automatically come to the top, so you may need to bring it forward by clicking on a button at the status bar at the bottom of your screen. BE CAREFUL not to minimize or resize the browser window that has the Red River page in it, or it may cut you off.

Please do not use Direct Messaging to our guests or the Moderator, because it makes it more difficult to follow the discussion. You will have an opportunity to ask questions during this session, afterwards in open discussion, and if we STILL don't get to it, we will follow up via email.

If your system or ISP has an inactivity timeout, you may want to keep your connection alive by sending a Direct Message, such as a "hello" to one of the other members here!

Background information for today's session may be found at http://www.emforum.org/redriver/workshop/wk990429.htm and there are links on that page to some of the information we will be discussing today. There is also an agenda we will try to follow for the layout of the session at http://www.emforum.org/redriver/workshop/ag990429.htm .

Right before we begin the first Q&A portion I will review how to submit questions or comments.

Now it is my pleasure to introduce our participants for today. Joining us from Ottawa, is Ms. Julie Cranton, Environmental Scientist with the Canada Centre for Remote Sensing. She has previous experience with RADARSAT data for ice and ocean applications, and is currently working in the Application Development Section with the Disaster Monitoring group.

And from Sioux Falls, South Dakota, Mr. David Greenlee, Physical Scientist with the U.S. Geological Survey's EROS Data Centre. Dave has had some intensive work experience with flooding, and is currently involved with a project to develop seamless Internet delivery of raster data.

Welcome to you both, and thank you for pitching in for us today. Julie will start us off with an overview. Julie is also a new network member, and we are very pleased to have her. Julie, if you please.

Julie Cranton: Welcome, everyone. I was asked to speak / present about Disaster Monitoring Remote Sensing in Canada. I have put together a brief overview of both the Disaster Response Management team of the Canada Centre for Remote Sensing (CCRS) and the Red River project we are working on with the EROS Data Center.

CCRS is internationally recognized as a leading centre of excellence in the use of earth-observation data, and supports an expanding industry sector including the world leaders in global ground station, image analysis and radar mapping markets. CCRS is responsible for the reception, processing, archiving and dissemination of remotely sensed data for Canada. In conjunction with the private sector, it develops remote sensing technology and applications.

Within CCRS there are 4 divisions, one of which is Applications Division (AD). AD is responsible for the research, development and demonstration of the use of remote sensing for resource management and environmental monitoring. Working in close collaboration with the Canadian Space Agency, it validates and evaluates the potential of RADARSAT in areas such as ice, ocean, hydrology, land use, forestry, geology, mapping and environment, and transfers the resulting techniques to industry and operational users.

Within the last year an AD / Disaster Response Management team was created. The team, led by Terry Pultz (613-947-1316) is made up of a group of experts within Disaster Monitoring related applications. The team consists of: Terry Pultz, Project Leader/Floods, Vern Singhroy, Geohazards, Paris Vachon , Oil Spills, Julie Cranton, GIS.

The team objectives are:

1) Establish collaborations with National and International Agencies to promote of the use of RS data in disaster management.

2) Rapid response to Canadian disasters by producing pertinent products from remotely sensed data, distributing them to authorities and working with the Centre's communication team to ensure proper visibility.

3) Establish and implement methodology of disaster response management including information requirements and data delivery methods.

4) Assess the use of future satellite platforms, such as RADARSAT II in disaster management activities.

As well as leading the Disaster Response team Terry Pultz leads the Hydrology team. The CCRS hydrology project was established as part of the Radar Data Development Program in 1988. Its activities are directed towards the utilization of remote sensing technology for the estimation of the various components of the hydrological cycle, and to provide the hydrology community with tools for water resource management. The principle objectives of the program are:

i) to evaluate the feasibility of extracting information from RADARSAT SAR data on soil moisture, extent of snow cover (wet and dry), snow water equivalent (SWE), river ice dynamics, flood extent and damage and wetland/landcover conditions

ii) to initiate and support research in the development of distributed hydrological models that effectively use spatial data and to develop techniques to use these data to improve runoff forecasts

iii) to demonstrate the operational use of geospatial data for hydrological applications.

An important goal is to transfer the techniques developed to potential users. To meet that requirement, the program concentrates on development of joint research efforts with other agencies, universities, industry, and on providing financial and technical support to these groups for the development of tools and techniques that will encourage the use of remote sensing in hydrology.

I'm including a link to the Hydrology portion of the CCRS Website for your records. You will find examples of projects the Hydrology team has worked on in the past. I would encourage you take a look (at a later time). http://www.ccrs.nrcan.gc.ca/ccrs/tekrd/rd/apps/appse.html

One area of interest with both the CCRS / Hydrology & Disaster Response teams, has been the Red River area in Manitoba. Here's an example from the CCRS Website which illustrates the use of RADARSAT imagery for Flood Mapping in the Red River area.

http://www.ccrs.nrcan.gc.ca/ccrs/tekrd/rd/apps/hydro/redrvr/redflde.html Please take a look now.

CCRS and the EROS Data Center are currently integrating a number of Canadian and U.S. spatial data sets to create a seamless international GIS database for the Red River. The data which will serve as the first level geospatial data base required for the development of a system to monitor and predict floods within the Red River watershed. The system will have a requirement to integrate raster, vector and point data at regional and local scales. To date CCRS has compiled a Red River database with imagery and topographic data over Canada.

We have a total of 16 RADARSAT images. Eight from 1996, one is during the flood and 7 are pre and post flood times. The remaining eight images are from 1997. Six during the flood period and 2 pre and post flood. By using a threshold procedure, we have delineated flooded and non flooded areas in each of the images acquired during the flood. The delineated flooded and non flooded areas have been saved as bitmaps.

We have joined together 16 National Topographic Data Base (NTDB) 1: 50000 mapsheets of points, lines and areas data. The point data we have consists of barriers, buildings, campgrounds, cemeteries, dams, depots, historic sites, liquid dump, sites, silos, tanks, and towers.

The line data we have consists of contours, dams, dyke/Levees, footbridges, limited use roads, NTS territorial limits, pipelines, pond partitions, railways, roads, sports track, trails, transmission lines, watercourses and wharves.

The area data we have consists of buildings, built-up areas, campgrounds, mining areas, park/sports field, solid depot, transformer station, vegetation, water bodies and wetlands.

We are missing one important component for this database, soils data. We have ordered the data from the University of Manitoba and expect to receive it some time in the next few months.

The next step for this project is to join the CCRS compiled dataset with the EROS Data Center dataset. This will be a challenging task as our datasets consist of different themes and are in different scales and projections. We look forward to working together with the EROS Data Center to perform this task.

Amy, I'd be happy to answer any questions now.

Amy Sebring: Thank you Julie. We will now take about 10 minutes for questions/comments from the "floor". Audience please send in just a question mark (?) to indicate you wish to be recognized, go ahead and compose your comment or question, but wait for recognition before hitting the enter key or clicking on Send. We are ready now for your questions or comments for Julie.

While you are thinking, I would like to mention that the CCRS has put together an excellent introductory tutorial on the subject of remote sensing, and there is a link to it on the background page for today's session.

Question:

Lee Klapprodt: How about mapping watersheds?

Julie Cranton: To answer Lee, no we have not concentrated on watershed delineation.

Question:

Slobodan Simonovic: How can this data (CCRS) be obtained?

Julie Cranton: To answer SS, please contact CCRS, however there are still some copyright issues to be addressed. Contact: julie.cranton@ccrs.nrcan.gc.ca or 613-947-1279.

Question:

Terry Birkenstock: Two questions: As part of the IJC Red River Study we have been assessing data sources and creating Metadata for those data so they can be searched and the metadata reviewed. Do you have metadata for those data sets you have assembled? Also, how do you plan on using this data via the web - as a base for generating maps as requested by users, or some more grand plan?

Julie Cranton: Metadata for the NTDB data can be found on the Centre for Topographic Information Website (I don't know if off hand but if you can't find it, email me). 2nd question: For this project we do not have plans to post on the web but CCRS has a project called CEONET which will address this issue.

Question:

Amy Sebring: Julie, can you give us an example of how this type of data collection has been used in the past or may be used in the future for disaster response or recovery?

Julie Cranton: Amy, can you clarify your question? Do you mean CEONET?

Amy Sebring: How are the images used in disaster response? Are they disseminated to response agencies e.g.?

Julie Cranton: Yes we disseminate the data as required, however we are sensitive to competing with private industry. For example, RADARSAT images were provided to the Transportation Safety Board for the Swiss Air Crash. .

Question:

Slobodan Simonovic: I am interested to learn how you deal with dynamic characteristics of the area like Red River basin? How often do you update your datasets? New structures are being added, new land use is being introduced; how do you capture and keep up with these changes?

Julie Cranton: We are working on a demonstration project. CCRS is not an operational agency, we are a research and development agency.

Question:

Ted Bailey: I assume that you work on cost recovery. Who pays during an emergency or disaster situation?

Julie Cranton: No, we do not work on cost recovery; however, data are sold at commercial rates by RSI (RADARSAT International.)

Amy Sebring: Thank you very much Julie. Next to give us the latest from the U.S. side is Dave Greenlee. Dave is also a network member and we look forward to continued involvement with both of our presenters. Dave, if you please.

Dave Greenlee: Thank you, Amy. I have 4 topics to share with you today:

1) Landsat 7 Status

2) Open Skies Program - Radar data for Red River and Devils Lake

3) NE South Dakota Flooding Information

4) Cooperative Research w/ ESRI - Seamless Data Browse and Delivery

1) Landsat 7 - First, a quote from our newest web site: "The Landsat 7 mission began after a spectacularly successful launch on April 15, 1999. Since then, in addition to the normal post-launch checkout activity, a great deal of effort has been devoted to extracting and publicizing a few scenes for media and public relations purposes."

The web site is http://landsat7.usgs.gov --- you might want to take a peek right now.

As you check it out, notice the cute "animated GIF" in the "Where is Landsat 7" located under Satellite and Sensor. Our web team gets extra points for that one!

Let me also direct you to the imagery (click on "first images" in upper right) and especially the enlargement of Sioux Falls, which almost does justice to the quality of the image for the new 15m ETM sensor. This is really excellent imagery, and I'm certain that we will be able to put it to good use in flood extent delineation and cover type mapping. I've been trying to shoe-horn some of the engineering data for the area in NE South Dakota (happens to be in the same swath as the Sioux Falls image). They tell me the spacecraft is not positioned at the correct altitude and they don't want to release data that aren't fully calibrated and that are under-sampled - high speed / low orbit...

They will be positioning directly over Landsat 5 for some important calibration and eventually it will be positioned 8 days out of synch with Landsat 5, to minimize the worst case repeat cycle. We hope it will go operational about July 1. To see more Landsat 7 imagery and higher resolution "jpegs", check out NASA's collection at: http://ltpwww.gsfc.nasa.gov/LANDSAT/CAMPAIGN_DOCS/DATA/Browse/Landsat7_Gallery.html

2) Open Skies Program - After an 11th hour invitation to acquire some Open Skies imagery, some of you may know that we were told that we did acquire aircraft Synthetic Aperture Radar (SAR) imagery over the Red River and Devils Lake on April 15, 1999. It was too cloudy for panchromatic imagery, but we should be getting the radar on 8mm tape any day now. Several people recommended we add the Devils Lake area to the mission, and that it is what they reported did.

Some background is in order. The Open Skies Program is a military program that was originally intended for verifying arms reductions for all signers of an Open Skies Treaty. On April 14, we were offered the opportunity to get data acquired as a training flight and to check out a recently installed radar acquisition system. If we like what we get, we may be adding some sites to their "target encyclopedia" so that we can get more low cost data from their training missions.

More information on the Open Skies program can be found at: http://www.nawcwpns.navy.mil/~treaty/Open_Skies.html

If you are interested in the SAR, I just discovered another helpful web site at Sandia: http://www.cmc.sandia.gov/facts/saros.html

3) Northeast South Dakota Flooding - In NE South Dakota, the headwaters of the Red River begin on the north flank of a glacial feature called the Coteau des Prairie. This coteau area is similar in many ways to the Devils Lake area, with prairie potholes and many "relatively closed" basins, which in recent years have been growing steadily and causing flooding of agricultural areas and considerable property losses. There is almost no groundwater movement or surface water runoff, so evaporation is about the only way out! Above average precipitation and below normal evaporation conditions have caused the area of the lakes to triple since 1992, and the volume of water has increased even more.

FEMA declared a Federal disaster in June, 1998 (FEMA-1218-DR) and sponsored a science team to document the history of inundation, create models, and investigate options for mitigation. The team includes: FEMA, USGS Water Resources and Mapping Divisions, USACE - Omaha, SD Department of Environment and Natural Resources / Water Rights, Natural Resources and Conservation Service, SDSU State Climatologist.

Our first job at EROS was to document the recent history of inundation using the Landsat archive to form a time-series. We calculated water areas from 4 dates. Now we are getting the data ready to make a "movie" of over 20 images that we intend to put on the web as an mpeg or avi. It should be very interesting and instructive.

A second job was to improve on the delineation of basins using USGS Digital Elevation Models that have been given special treatment and reinterpolated to 10m posting interval. This is about as good as DEMs can get without the added expense of flying the area with LIDAR. In the last few years, the USGS has cost-shared with many other agencies (in this case the State of SD) to get 10m data for low-gradient areas that require this treatment.

As a by product of the basin delineation, Glenn Kelly was also able to calculate the "areas subject to inundation" by filling all depressions to their theoretical maximum. The results of the basin delineation has been quite gratifying, as the other team members (WRD, USACE, SD DENR) seem to be making very good use of these data in their work.

For more details on NE SD flooding, see our web site at: http://gisdata.usgs.gov/nesdflood

4) Cooperative Research - ESRI Seamless Data Browse and Delivery: We have begun an ambitious and exciting 3 year Cooperative Research and Development Activity (CRADA) with the Environmental Systems Research Institute (ESRI) to develop a web based browse and download capability for some truly large datasets. We intend to provide several raster datasets for the conterminous US at 30m resolution.

The featured data include MRLC (e.g. MultiResolution Land Characterization and georeferenced TM imagery), NED (National Elevation Dataset - seamless, best-available DEM), and the NHD (National Hydrologic Dataset - a conflation of EPA's river reach data and USGS's 100K scale hydrology).

We intend to demonstrate our first results to over 10,000 people at the ESRI User Conference (end of July). This involves prepping and loading about 600 Gbytes of data into Oracle to be accessed by ESRI's version 4 Spatial Database Engine (SDE4), and served using some newly developed server "middleware" and applets that will bind with a web browser to provide full resolution browse and download capabilities. The web sites for NED, MRLC, and NHD aren't yet public.

When they are ready, I will provide links to them at: http://gisdata.usgs.gov Amy, I am ready for some questions.

Amy Sebring: Thank you Dave. We are now ready for questions/comments regarding the U.S. efforts. Remember to send in a question mark (?) please.

Question:

Paul Bourget: Not a question really, Dave. I just received a document from the Open GIS Consortium and they plan to use the Red as a pilot area for their Disaster Mgmt Interoperability Initiative. Seems like a wonderful leveraging opportunity between your work, the IJC and GDIN.

Dave Greenlee: Yes, I think that's great news. The OGIS spec needs to be tested, and I think we have several datasets and systems being used in our study. If we can show how the OGIS interoperability can work here, that is really something.

Question:

Amy Sebring: Given the current bandwidth limitations of the Internet, what is your strategy for delivering large raster files via Internet?

Dave Greenlee: The short answer is divide and conquer but this can take many forms. We need browsers that work efficiently through all scales of data. Also, we need to understand that mirror sites and redundant delivery are good things. For example, the RRB datasets might be served up from a National data base, also served from some RRB site.

Question:

Terry Birkenstock: First - Dave, I hope we can work closely with you in coordination with the OGC initiative - it should give us all some critical pieces of software we really need. Second, is the LANDSAT 7 data going to be priced similarly to LANDSAT 5 data --- I assume with higher resolution data, scene size may be different, or not?

Dave Greenlee: Second question first. I think the data will be cheap. I've heard $200 to $500 per scene. First question. I think we all need to be involved to test the OGIS spec and we would want to be able to transfer data in forms that each agency intends to use them.

Question:

Slobodan Simonovic: Are you going to limit your datasets to US or basin (I mean Red River) will be covered as a whole?

Dave Greenlee: Julie is coming down next month to work on the DB, so we don't really have it together yet. USGS is working on more of the regional data including our 1 arc-second (~30m) elevation data, 100K hydro and transportation. I guess the answer is basinwide.

Amy Sebring: Julie, would you like to comment on the US/Canadian cooperation? Joint efforts?

Julie Cranton: Yes I will be meeting with Dave.

Question:

Lee Klapprodt: Amy, I think this is relevant. Why doesn't the this effort include the Devils Lake Basin as it is part of the Red?

Amy Sebring: Dave, you have mentioned the Devils Lake area; would you like to elaborate on that?

Dave Greenlee: I would defer to Terry B.

Terry Birkenstock: I'll try and Slobodan can add too. When you say "this effort", you are talking about a USGS and CCRS initiative. If you are really thinking the IJC study, that decision was made at the outset of the study. Slobodan?

Lee Klapprodt: Actually I'm referring to the Red River Emergency Net

Slobodan Simonovic: IJC has made a decision that RRB Task Force should focus on the Red River without Devils Lake in US and Assiniboine in Canada. There are some reasons for that decision: political, international, etc.. I am not able to elaborate more in this moment.

Amy Sebring: (We have also tried to relay Devils Lake info via network at least some.)

Lee Klapprodt: Devils Lake is one MPF away from contributing.

Terry Birkenstock: Lee, I think that the Red River Basin Disaster Information Network could (should?) certainly evolve to include the Devils Lake Basin, as well as the Assinniboine in Canada. It just isn't part of the specific study at this point.

Lee Klapprodt: OK, should consider it in future.

Question:

Amy Sebring: Dave, you did say you were expecting some imagery from that area soon?

Dave Greenlee: Yes, I was hoping to get Landsat 7 data from last week and we did get some SAR imagery flown a couple weeks ago. It isn't here yet.

Amy Sebring: Do you expect it will be posted on Landsat 7 site?

Dave Greenlee: No. Not for awhile. The engineers are not going to free up data until July.

Question:

Paul Bourget: I keep hearing this Devil's Lake argument and believe we should, in fact, include it within the RRBDIN region of interest

Dave Greenlee: We certainly have some related work in USGS going on that should be linked. Contact Russ Harkness or Greg Wiche for details.

Wrap Up:

Amy Sebring: We are just about out of time but Julie has looked up a couple of the web addresses for us that she mentioned earlier. Julie please.

Julie Cranton: www.ccg.rncan.gc.ca (this is the CTI Website) Center for Topographic Info

http://ceonet.ccrs.nrcan.gc.ca/api?Language=English&Request=DisplayURL&Target=http://ceocat.ccrs.nrcan.gc.ca/cgi-bin/client_acc/ceocate/holdings.phtml (this is the CEONET Website)

Amy Sebring: Thank you very much Julie. Again, our thanks to both of you, and thanks to our members for joining us today. We will officially close the session, but you are invited to remain a few minutes longer for some off-the-record, open discussion.