Minutes
UCIME Group Meeting
October 6, 2000
*** Important Dates and Announcements:
- UCIME group meetings will be at 10:00 am in Keith’s office on the first Friday of every month. The next UCIME group meeting will be November 3, 2000.
- Update Reports from each team member:
- Noah:
- He is waiting for some basic data from the USGS folks who supplied him the Shootout data of Soux Falls in order to finish the readme file.
- He wants to modify the Shootout web page to make room for people to get in touch with each other. This may just be an "add my name to the Announcements" button which they would email to him.
- He is working on creating the "pseudo census" for Santa Barbara in order to get accurate demographic changes over time.
- He is getting started on coupling the Urban & Fire Models.
- He is working on the write up of the historical, "backcasting" paper with Max Moritz & Jeannette and is exploring more fire modeling materials in Santa Barbara with Max.
- He is talking with Steen about funding the web-based work that was begun while in Los Alamos last month.
- He will be talking to Helen & Ray Smith about funding and logistics for a web-based survey his work in disaster information systems.
- He wanted to remind Keith to get familiar with Germana's translation of the UGM model. She is returning at Christmas time.
- We need to formally invite Steen & Claes to UCSB to talk about their modeling efforts and publications. Keith is going to work on this when he and Jeannette go to LANL before the Kuwait trip.
- Keith
:
- The Shootout was announced at the Land Use Modeling Workshop at Banff and there seems to be a lot of interest. The guidelines and rules will be the same as what was laid out during the General UCIME meeting last Spring. Dates and a schedule of events have not been determined yet. By next month this should be on line. There are 5-6 groups of modelers interested in participating. Once the web page is completed, an event announcement will be sent out to all participants of the Banff Conference suggesting they forward it to any others with interest. This might be a part of a special session that Brian Pijanowski is planning on Landuse Modeling at the IALE Conference next year.
- Shootout Description
: We would put data on a web page of an urban area and its surroundings. You would then have to model the site, in this case Soux Falls. You will have a year to do it. Any additional data that you find or are going to use you must share it with everyone (e.g. make it available through the Shootout web page). The outcome of the modeling effort would be a meeting at a conference to look at model results to compare and contrast. There would be a second meeting in Soux Falls to present to the results to city planners and then run the models out to 2010. The final meeting would be in 2010 to see which model performed the best. A book would be published with the results from each of the 3 meetings.
- David Stoms has been awarded a grant entitled "Integrated Modeling for Region Biodiversity Conservation and Urban Growth" that will use the landuse data the Melissa is putting together in a modeling effort between landuse change and species diversity. The project was funded through a program called Research ACROSS the Discipline.
- He has been working on additional code for the UGM model to address multiple landuse layers, which has been difficult/challenging in the area of building a transition matrix for each year long time step. He is also looking at the stability and transition probabilities.
- Still lots of interest his model and requests for the code.
- He finished a paper with Elizabeth De Silva on the Portugal (Lisbon and Oporto) model application. She will be back in the winter to continue her work.
- He will be in Kuwait for two weeks in December.
- Jeannette
:
- She is still working on Version 3.0. Tommy Cathy is technically off the project but is still working on it on the side. She has been working on the MPI (Multiple Parallel Interface) here on the 12 machines of the Descartes Lab. Since the model runs in a linear fashion (embarrassingly parallel processing), addition of a CPU will double the speed, which is good. With this up and running we will not be dependent on the EPA.
- She has added documentation in the code, which is under review with Tim and should make the code easier to operate.
- She worked with Paola Gomez on the Mexico City application but is waiting to get Version 3 working before continuing on with her.
- She just started a new fiscal year with the USGS. She might be working on urban growth in the Bay Area on a five area based study, looking at regional smart growth.
- She, Keith and Xiaohang are going to GIScience 2000 at the end of the month.
- She will be attending the Aurora Conference in December, a cross-governmental agency gathering (Civilian Application?)
- Jeff:
- He finished the porting of the model (See the Future / You Grow, Will Orr’s model) to Stella. He is now coupling Sleuth with See the Future, working on input/out tables between the models.
- ECP wants them to put this on the web and they are wanting for funding to do that. ECP (Economic Community Project) has been struggling this summer, as it did a rotation of board members. Now looking better and we should be able to move forward with new money on line and Keith on one of their sub-committees. Melissa’s effort will plug right in here. Plus Ryan’s focus now is on CalTran funding to develop a virtual environment of this area (campus, Hwy 217 area, etc.). This is in ArcView 3D with actual digital photos of the sides of each building. Campus is the test site with the rest to come.
- He and Keith are also talking with HPS (producer of Stella) to acquire software and technology discounts. The president has made an offer of $4 k and Keith would like to go in that direction (an interface to run with Stella and can run with C++ and Java for Unix or NT platforms). Will need a Java programmer who might be on line next year.
- Helen:
- Helen brought a new reference for all to look at, Review and Assessment of Landuse Models (Martin Grove). It’s a comparison of 19 landuse models put together by the Center for the Study of Institutions Population and Environment Change (CSIPEC) (done separately from the EPA study). It is not published yet but coming soon. This could be good material for a graduate seminar. The project is connected with the Forest Service.
- She finished the bibliography list from the graduate seminar given last year. These will be put directly into the UCIME web page.
- She has been working on two papers; one for GIScience 2000 and AAG, the February session on Integrated Modeling.
- She is heading to Ottawa, Canada, to give a talk to government agencies on urban modeling.
- Melissa:
- She is working hard on the landuse layer for Anderson Classification Level 3. Remaining for this first year (1998) is the extremes outside of the urban area. Keith has a student that is interested in helping her out. The project will be done very soon and all are going to be very excited (UCSB, City and County).
- Xiaohang
:
- She went to LANL for a short period of time.
- She is working on a couple of paper ideas. First is working with Steen’s model to compare discrete and continues time. If you have interest in Steen’s model, get in touch and she can get you a copy of a paper on the model. Model runs in Pascal with Delphi and she is having problems running it here at UCSB because of lack of software access. The model isn’t calibrated as it is based on mathematical assumptions and is not designed to be calibrated. The second paper will look at model comparison that she and Noah are just beginning to focus on. It is a long-term project and could be a nice combination with the Shootout. She did a few diagrams on the board on how discrete and continues time works.
- She is beginning to concentrate on her paper for GIScience 2000, which will be on the uncertainty associated with each paper, running some indicators on the data.
- Tim
:
- He has been working on the UCIME web page and learning DreamWeaver. A draft copy of the page was passed around for editions.
- He has begun work with Jeannette on updating the Gigalopolis Project web page, specifically in reviewing the Version 3.0 updates (downloads and readmes). This has required installing the model on his machine, which has been a process in itself.
- 3 weeks of his summer were spent at LANL where he looking into watershed modeling and model coupling tools. He is working with two groups; Legacy for Future Framework (L2F) and Modular Modeling System (MMS). L2F was put together by Denise and Jim George who created their own grammar linking with Visibroker to make a web-based coupling tool. He has been setting up a simulation with this tool that will be run from UCSB and linked with various different servers. Since it is a beta product there are lots of bugs and pit-falls, so we will see how far we can get. The idea is to test its validity and see if moving forward with their software is worthwhile. MMS is a USGS project out of Denver under the guidance of George Leavseley. The included models are hard coupled but the next version will have a different architecture that will have the models wrapped in Java and will have a web-based access, plug and play environment. We are hoping to have Leavseley come out and give a workshop about MMS.
- The watershed modeling that he will be working on is watershed dependent and he is hoping to work on the LTER for the south coast and model the three watersheds that will be part of that large study. Keith mentioned that the Hudson River Project has linked their model with UGM and it would be interesting to be in touch with those guys. It is hoped that part of this modeling effort can be done through his involvement with the watershed analysis group that ESRI and David Maidment are forming.
- He did make a visit to NASA Ames and saw our team members up there.
- Matt:
(absent)
B. Review of the UCIME web page:
Everyone reviewed the first draft and made comments for addition. Keith is going to give write permission to Tim on the directory where the web page sits to facilitate web page maintenance. We need to change the picture and get a project logo (talk to Susan).
C. Conferences up and coming:
- GIScience: 28-31 October. Savannah, Georgia. (http://www.giscience.org/)
- Aurora Conference in Charleston, South Carolina, in December.
- IALE Conference next year.