Date: Fri, 13 Oct 2000 23:54:20 -0700
From: Ramon Arrowsmith 
Subject: [Fwd: Phoenix2100-NSF ITR proposal]

1)  Here is the url for the NSF Information Technology Research:
http://www.itr.nsf.gov/
Here is the RFP:
http://www.nsf.gov/cgi-bin/getpub?nsf00126

2)  The background of my involvement and what little I know:
I was invited to be on an interim steering committee on Geoinformatics that was charged by Herman Zimmerman (head of NSF-GEO-EAR; the earth sciences part of NSF--to where I submit my disciplinary proposals, and I think within which Jon was a program director for a bit) to explore the issues of databases and geospatial data in earth sciences and write a white paper.  I have been a little soldier in the operation, but I was involved in all of the discussions (three meetings:  2 in the spring in Washington DC and one this fall in Houston) and I manage the website:
http://activetectonics.la.asu.edu/geoinformatics/

In our meetings, Zimmerman and some of the other NSF colleagues (including Alan Gaines--I have lots of notes from discussions with him and I think that he was the visionary behind much of this) encouraged us to also take advantage of the NSF-ITR proposal opportunity that was coming along. The possibility of $ focused initially, but it ended up fragmenting us at the end, so now we will write one Large proposal with the San Diego Supercomputer Center and a bunch of small ones.  The history of ITR is that it grew out of KDI (Knowledge and Distributed Infrastructure).  Last year was the first year of ITR and the response was very large, but no new staff were put on the job by NSF.  It taxed them physically and financially.  THe funding rates were terrible:  ~5%.  I got the feeling that many non-computer science/IT people felt like they got burned because they put a lot of effort into proposals, but got shot down because it was dominated by CISE (Computer and Information Sciences part of NSF).  The degree to which this will be an issue for this coming year is not clear to me, but for sure, we have to identify some IT Research challenges.

I don't know the exact level of NSF funding this next year.  I guess the VA_HUD-WATER_ENERGY bill passed yesterday.  According to my notes, last year, the ITR was $90M for all IT Research.  Because of the overwhelming response, the NSF and the President put in a major increase for the next year request:  +$180M, so if the NSF were fully funded, the total ITR would be $270M.  However, some (much?) of the addition will be distributed Foundation-wide.  For example, the GEO program expects to get about $16M of new IT$ next year.

With regard to the proposals, as I understand, the small ones go to the appropriate disciplinary directorate.  So I hope to write a small one about the San Andreas Fault system.  That will go directly to EAR and if it is funded it will come out of their piece of the distributed bits.  It would also be reviewed only be EAR people.  As far as I understand, the Group and Large proposals will compete Foundation-wide, and so we have to watch for the CISE-driven IT aspect and address that.

3) Here is something from Jon that looks interesting:
http://www.nap.edu/books/0309050820/html/index.html

Ramon,
The above report is five years old but deals with combining various environmental data sets. It may provide some
historical context for the ITR project we discussed today (or it may not-I've only looked quickly at the executive
summary).

Jon

4) Here is an email from Zimmerman:
Dear Colleague,

I would like to bring to your attention a new funding opportunity under the
National Science Foundation's Information Technology Research (ITR)
Initiative.

The National Science Foundation seeks to fund innovative, high payoff
research, which explores new scientific, engineering, and educational areas
in Information Technology (IT).  The Information Technology Research Program
is entering its second year. This year's solicitation is considerably
broadened to include not only fundamental research in IT, but also new
applications of IT in all scientific, engineering, and educational areas, as
well as innovative infrastructure to support IT research and education. The
Directorate for Geosciences (GEO) is an active participant in the
initiative, which represents a significant opportunity to meet important
needs of the Geosciences community in partnership with computer scientists
and others.  GEO has identified three of this year's five thrusts as
particularly relevant to achieving the long-term scientific goals outlined
in GEO's recently published long-range plan "NSF Geosciences Beyond 2000":

*       Applications in Science and Engineering (AP);
*       Information Management (IM) and;
*       Scalable Information Infrastructure For Pervasive Computing And
Access (SI)

A brief description of these topics can be found on the GEO web page
(http://www.geo.nsf.gov/geo/itr/) <http://www.geo.nsf.gov/geo/itr/)> , with
more details and a copy of the program solicitation available on the ITR web
page (http://www.itr.nsf.gov) <http://www.itr.nsf.gov)> .  Time is
relatively short to respond (late November to early January depending on the
type of proposal) so we ask you to quickly and broadly share this
information with colleagues at your institution and elsewhere.  Researchers
interested in submitting proposals are encouraged to discuss their ideas
with the appropriate GEO staff member identified in the program
solicitation.  In addition, Dr. Eric Itsweire (eitsweir@nsf.gov)
<mailto:eitsweir@nsf.gov)> , the GEO primary representative for this
NSF-wide program, and Mr. Russell Kelz (rkelz@nsf.gov <mailto:rkelz@nsf.gov>
), contact for the Division of Earth Sciences, are available to answer
general questions about the ITR competition.

Thank you for your attention.
 

Herm Zimmerman
 

NSF 00-155
 

Please note new phone numbers
Herman B. Zimmerman
Director, Division of Earth Sciences
National Science Foundation, Suite 785
Arlington, VA  22230
TEL:  703-292-8550
FAX: 703-292-9025
 

5)  So what can we do?
I think that the ideas of Phoenix2100 provide a very interesting hook for ITR in that we have lots of data about the city that are gathered and have been gathered.  We can ask many questions of each dataset, but as we all know and need to argue persuasively, the big questions about how people interact with their environment in a situation such as Phoenix require inquiry of diverse datasets simultaneously.  ITR can support the development of the database infrastructure (the data and organizing it), as well as the tools for the inquiry in which we are interested, and the visualization of the results.  IF we also include some reference to the SimCity or other modeling initiatives, we have it all.

Look at the website of the San Diego Supercomputer Center: http://www.sdsc.edu/.  The tools that they develop are the results of IT research and are of interest to IT researchers.  I do have a contact there with a nice person whose PhD is in ecology if we want to partner with them, but I am unsure if we want to.  That is the uncertainty I have about the degree of IT emphasis.  Do we have enough expertise in house?  Do we want to involve the local colleagues?  The other thing I realized is that some of the IT Research is the (re)invention of the perfect wheel. ESRI (owners of Arc-Info) are the industry standard tool for geospatial analysis, but it is an old program and some say it is built around 1960s technology idea and code-wise.  Dangermond might argue with that, but the SDSC colleagues are looking forward with the 2000s technology.  So there can be a struggle there.  The visualization of diverse and non explicitly spatially related data is a major ability of SDSC, as are transparently distributed databases and the "middleware" that manages the queries of those databases.  If the modeling is very complicated, of course a supercomputer can help as well.

The problem will be the focus.  But, I think that is where the ideas of Phoenix2100 come in.  We want to know what has happened (all of the different parameters about the city (page 2 of the Atlas writeup) with time), what is happening, and what can happen.  To do the history, we need to put together the datasets.  Many many are available, off the shelf from the various stakeholders.  To figure out what is happening, we need to establish a means of maintaining the databases that are built and their connectivity and gather new data, so we have the pulse of the region.  To anticipate the future, we have to train our models on the history, situate them in the present, and send them forward and test the results and visualize the various scenarios.

6)  So what could we do with $1M/year?
We could certainly develop the datasets and the analysis such that a major product would be the Phoenix Metropolitan Regional Atlas.  As a small aside, I go back and forth on the name of that, in that an atlas seems to denote a static feature (which as a coffee table book, as I imagine the one from Tijuana to which Fritz referred would be cool and I would like and so would many others), but an atlas can also change (unless we call it historical)).
THe main $ would probably not be for students.  We should look at the language from the RFP to see how it sounds, but all of the conversations I have had did not include a major student effort.  But I am certain that there would be the RA-type funding in there.  I do not think we would have to buy much data, but the cost is taking the data and making it useful and that is very expensive.  I imagine some kind of a structure like this (numbers are $k):

GIS techs (3)   150
Modeling tech (1)   50
GIS workers (8)   160
Professor/PI summer salaries (4)   32
Graduate students (6)   90
  Fringe 147.64
Data   20
Operations   50
  total direct 699.64
  indirect 367.311
  Total/year 1066.951

I am not much of a budgeter at these levels of $s.  I have no idea what would be appropriate as far as a University commitment.  The budget above basically has 4 professionals (3 divisions of the Atlas + modeler) managing 8 workers (1 for every 3 subjects).  The thing is that salaries in these fields are high, so it will not be cheap to hire someone who know what he or she is doing... I roughed out some PI and other interested colleague support and six grad students worth of research.

ramon
 

__________________
Ramon Arrowsmith ramon.arrowsmith@asu.edu
Assistant Professor Office: (480) 965-3541
Department of Geological Sciences Home: (480) 777-8371
Arizona State University Fax: (480) 965-8102
Tempe, AZ, 85287-1404 USA http://www.public.asu.edu/~arrows