Archive AGMODELS-L: file log9312, part 1/1, size 161652 bytes:

------------------------------ Cut here ------------------------------

Hello everybody,

I am evaluating the soil conservation efficiency of different cultivation
practices with EPIC. I'd like to express the results on a relative scale.
Therefore, I need to calculate potential soil loss (i.e., soil loss from a
clean-tilled bare fallow). Any one who knows how to enter this in the EPIC
file?

Best regards,

Mark Snethlage
m.snethlage@frw.ruu.nl

Howday fellow listpersons-

Introduction:
-------------
Allow me to introduce myself.

name: Ken Rojas
title: Hydrologist / computer specialist
affil: USDA-ARES (for those who haven't heard, the ARS-Agricultural
Research Service has been changed to the
ARES-Agricultural Research and Education Service)
group: Great Plains Systems Research Group
where: Fort Collins, CO

current assignment: I have been involved in the conceptualization and
development of the Root Zone Water Quality Model (RZWQM) since it's meager
beginnings back in '87. For those interested I can fill your e-mail
habits with descriptive writeups and other related material. Before that
I created and modified simulation models for various uses (e.g. snow
hydrology, forest management scenarios, watershed stream routing,
orographic precipitation, and others).

Discussion Topics:
------------------
I am assuming that the people subscribed to this list are either involved
or have been involved as users or developers of agricultural simulation
models. I'd like to start the flow of information between all of us to
discuss problem solving within ag models and hopefully address some of
the conceptual issues nagging me.

Issues that are buggin me:

- Plant nitrogen uptake from the soil. What is the state-of-the-art and
how does observed values relate to what simulation models can predict?

We use a coupled active/passive uptake scheme. Take the current plant
transpiration rate with layered nitrogen concentration and distributed
roots and determine passive uptake. For active we use a Michaelis-Menten
approach. Can we improve upon this? Why?

As for the second point how does observed values relate to simulated
values? With our models we can determine the actual amount of nitrogen
extracted from the soil into the plant. Is there observed information
that we can use to test if any uptake routine is working as expected? I'd
like to get a true partitioning between the roots and the aboveground
biomass to see were all this nitrogen is going. We can create a balance
sheet to satisfy our own curiousity as to where our numerical nitrogen is
being distributed once it enters the plant, this is no problem. Have
studies been done to address this more detailed question of temporal and
spacial distribution of nitrogen in plants?

I realize this would have be extremely costly to get this data, but does
it exist? One can not just go over and hack off the tops of a plant at
harvest time and say "well it looks like the plant took up XXX kg/ha of
nitrogen this year" without accounting for the root storage, dead root
losses, leaf litter, etc. To further complicate this I hear that plants
are even breathing out some of the nitrogen uptaken, not just a little bit
either.

- Reconsolidation of the soil over time after tillage. What is the
state-of-the-art and availablity of validation data.

We use an approach presented by Dr. Dennis Linden in his Phd diss. which
looks at the accumulated kinetic energy produced by impacting rain drops.
Under certain conditions it works fine but this process is highly
spacially dependent and many times it takes too long or too short of time
to completely return to baseline values. Can we improve upon this?

- User expectations of simulation models. What does users expect to see
when they use a simulation model? Should ms-windoze, dos, X-windows, etc.,
be the preferred platform for simulation model presentation and development?
I could go on forever on this topic, it really controls model development
processes and eventual customer satisfaction. One can not just create a
model with basic features and expect users to be satified, users are
getting too sophisticated and want the same look and feel as everything
else they use. Is this a reasonable request to put on the scientist
developing the model?

Well here a few ramblings that will probably strike a cord in most of you,
so contribute to the list and tell me and others what you think. Don't
worry about sounding too stupid, I feel that I've set a new standard for
just how stupid can one person sound. Anything you've to say has got to
be more intelligible than my ramblings without much effort.

later,

Ken

===========================================================================
Ken Rojas
Hydrologist / Environmental Physicist / FORTRAN GOD : Hydrology--
Great Plains System Research Unit :
USDA - Agricultural Research Service . The
301 S. Howes St. P.O. Box E . . Ecologically
Fort Collins, CO 80522 . . Correct
phone: (303) 490-8326 // fax : (303) 490-8310 . . Science
E-mail: KENR@GPSRV1.GPSR.COLOSTATE.EDU . .
ATTmail: A03KROJAS ..... aka: TECS
===========================================================================

I would like to thank my compatriot Ken Rojas for starting a thread which
I hope can answer some burning questions we have had: what is the 'best'
way to model plant nutrient uptake? The problem we seem to have with RZWQM
is that the plant seems to 'clean' the soil of all inorganic nitrogen in
solution resulting in nitrate concentrations in leachate which are lower
than those normally observed. Another spin on Ken's question, from my point
of view, would the use of a convective-dispersive model such as that
developed by Barber provide a better mechanism for driving N uptake
by plants than a simpler approach. My philosophy is that there must be
a definite gain in both understanding and prediction when using a
model of greater complexity. Slapping a convective-dispersive model
into something that is already complicated must be appraoched with caution.
I beleive there are many plant modellers subscribed to this list. Could
any of you please let us know how you are doing nutrient uptake?

Thanks a bunch. I've made Ken look not nearly as ignorant as he thinks
he is 8-)

Jerome Pier
jp@unl.edu

Name: Chris Butts
Title: Agricultural Engineer
Affiliation: USDA, ARES
National Peanut Research Laboratory (We work for peanuts)
Dawson, Georgia

Current Projects: Coordination of national effort to develop and use
expert systems to manage peanut production, harvest, storage and processing.
Main interest is utilization of models to simulate the drying and storage
processes to optimize controls and discrete models to simulate the
harvest and marketing operations.

Simulating nutrient uptake by plants is indeed difficult. The problem
is at least two fold. First, we really do not know what is happening
in the root environment. Plants take up nitrogen, roots die, and the
nitrogen is recycled. Is nitrogen ever exuded from the roots? How
much N is taken into the plant, but never makes it the leaves or shoots?
We really do not know. Second, most experimenters never measure below-
ground production and subsequent N pooling. As a result, the numbers that
are generally reported are only based upon above-ground standing crop.
Therefore, N uptake is not being measured at all. So, in a model, such as
RZWQM, when we report N-uptake we are estimating the amount of N that
moves from the soil into the root. Some of that N is translocated to the
above-ground pool. Some does not make it. So, RZWQM appears to report
N-uptake rates that are too high.

We have been working on this problem, and after accounting for N in live
and dead roots, I think the results will be very pleasing. In response,
then, to Jerome's comment, I would say that plants are very efficient
harvesters of N. Given adequate moisture, we might expect most if not all
of the available inorganic N in the soil to be take in by the plant (and
indeed this does happen on rangelands).

By the way, my name is Jon Hanson and I wrote the plant component of RZWQM.
My interests are in ecosystem modeling with primary emphasis on grassland
systems.

+---------------------------------------++---------------------------------+
| Dr. Jon D. Hanson || Comm: (303)490-8323 |
| USDA, Agricultural Research Service || Fax: (303)490-8310 |
| Great Plains Systems Research || jon@gpsrv1.gpsr.colostate.edu |
| 301 S. Howes, P.O. Box E || FTS2000: a03jonhanson |
| Fort Collins, Colorado 80522 || |
+---------------------------------------++---------------------------------+

On models for penut drying you might want to talk to Dr. Marvin Stone he
did his phd thesis on drying hops. His mail address is marv@master.ceat.okstate.edu.
Gordon

Gordon Couger
Biosystems & Agricultural Engineering
Oklahoma State Universtiy Stillwater OK

It seems like some N should be left in solution if uptake is
primarily passive. The concentration in the soil solution should be
the same as in the water flowing into the roots, and gradually
decreasing if uptake is greater than new supply. Tom

Tom Hodges Cropping Systems Modeler ___ ___
USDA-ARS / \_/ \
Rt. 2, Box 2953-A Telephone: 509-786-2226 | |
Prosser, WA 99350 \______/^\/
USA potato tuber
thodges@beta.tricity.wsu.edu thodges@asrr.arsusda.gov
=====================================================================
...photosynthesis makes the world go around... Mr. Potato Head

>
> From: MX%"agmodels-l@unl.edu" 2-DEC-1993 19:41
>
> It seems like some N should be left in solution if uptake is
> primarily passive. The concentration in the soil solution should be
> the same as in the water flowing into the roots, and gradually
> decreasing if uptake is greater than new supply. Tom
>
> Tom Hodges Cropping Systems Modeler ___ ___
> USDA-ARS / \_/ \
> Rt. 2, Box 2953-A Telephone: 509-786-2226 | |
> Prosser, WA 99350 \______/^\/
> USA potato tuber
> thodges@beta.tricity.wsu.edu thodges@asrr.arsusda.gov
> =====================================================================
> ...photosynthesis makes the world go around... Mr. Potato Head
>

Yea your right. Jerome didn't quite tell all the truth when he said the
model sucks the soil dry of nitrate. There is always alittle left over
from the passive process, but the Michealis-Menten active uptake will work
independent of water movement and gobble up any nitrate left behind. This
is, of course, if the M-M equation parameters are NOT set correctly (ie,
active uptake controlled). For this reason I tend to agree with your comment
and Jerome Pier that a diffusion mediated system will probably give the most
realistic response.

There is a model named DAISY from the netherlands, author is Soren Hansen,
that has a diffusion coupled passive uptake procedure. It's rather
tedious since it try's to optimize for nitrogen demand by varying the
concentration gradient at the root surface. Can we possibly get anymore
iteration solutions into any model? You start stacking these solutions on
top of each other and you turn your computer into a very expensive paper
weight, while your waiting for a simulation to complete.

-------

This brings up another point; what is the maximum amount of nitrate that a
plant can uptake in a single day? I'm talkin ag crop here not your giant
redwoods. And why does this limit exist? What is the physical limitation
imposed on the plant?

Regards,

Ken

===========================================================================
Ken Rojas
Hydrologist / Environmental Physicist / FORTRAN GOD : Hydrology--
Great Plains System Research Unit :
USDA - Agricultural Research Service . The
301 S. Howes St. P.O. Box E . . Ecologically
Fort Collins, CO 80522 . . Correct
phone: (303) 490-8326 // fax : (303) 490-8310 . . Science
E-mail: KENR@GPSRV1.GPSR.COLOSTATE.EDU . .
ATTmail: A03KROJAS ..... aka: TECS
===========================================================================

Nitrogen will flow into the roots with a concentration of the soil solute,
however, the problem we face in RZWQM is that we do not simulate diffusion
to the root surface. Therefore, all nitrogen in the a given layer is
available to the root even though it may not be near to the root surface.

The problem gets very difficult when one attempts to consider the
spatial movement of N across a soil layer as well as through soil layers.

+---------------------------------------++---------------------------------+
| Dr. Jon D. Hanson || Comm: (303)490-8323 |
| USDA, Agricultural Research Service || Fax: (303)490-8310 |
| Great Plains Systems Research || jon@gpsrv1.gpsr.colostate.edu |
| 301 S. Howes, P.O. Box E || FTS2000: a03jonhanson |
| Fort Collins, Colorado 80522 || |
+---------------------------------------++---------------------------------+

Dear Agmodels-l Subscribers,

After only 2 weeks, there are well over 180 subscribers to this list! In
the words of Bikash Kumar Pathak :

"Electronic mail based common interest discussion forums should be
created to facilitate discussion among those in the systems researchers
and users community who have access to global networks. There are
many such forums in agriculture related areas but not in the specific
area of agricultural systems to the best knowledge of authors.
A Listserv based discussion forum for professionals having interest
in agricultural expert systems possesses some value for agricultural
systems research community. "

I was unaware of Dr. Pathaks recommendation when I began this list. All I knew
was I had some questions about simulation models and since modelers are usually
in front of their computers anyway there should be anaudience for such a forum.
I am glad that things have worked out as they have and look forward to
interacting and learning from you all.

The archives for the month of November are available! If you are a new
subscriber (we all are!) and have missed some off the introductions or
want to keep a record of everything that has transpired on the list,
send an email message to listserv@unl.edu containing the text:

get agmodels-l log9311

and you should receive a text file with all the messages posted to the list
from Nov. 16-30.

There is now an option for receiving a digest version of the list so that
you receive one mail message/day with all the messages posted to the list
during the 24 hr period. To receive the digested format send email to
listserv@unl.edu with the text:

set agmodels-l mail digest

I would like to thank those subscribers who have introduced themselves:

B. K. Pathak <pathak@emailhost.ait.ac.th>
Dr. Gerrit Hoogenboom <GHOOGEN@GAES.GRIFFIN.PEACHNET.EDU>
Tom Hodges <thodges@beta.tricity.wsu.edu>
Peter Cox peterco@apsrusg.sth.dpi.qld.gov.au
Jess Lowenberg-Deboer <LOWENBERG-DEBOER@agecon.purdue.edu>
Tim Griffin <tgriffin@umce.umext.maine.EDU>
Jon W. Zeitler zeitler@ariel.tamu.edu
Stefano Pagiola spagiola@FRI-nxt-Pagiola.Stanford.EDU
Gordon Couger gcouger@olesun.agen.okstate.edu
Stephen Binns <sbinns@cix.compulink.co.uk>
Meindert de Jong <VLINDERS@RCL.WAU.NL>
J.Kaduk <kaduk@dkrz.d400.de>
Sharyl E. Walker <guernsey@holstein.age.uiuc.edu>

Please do not be shy about introducing yourself! I have found these
introductions to be fascinating. There is a wide range of experience
and expertise out there available to you. A brief description of your
background and interests will facilitate the transfer of knowledge.
Do not be intimidated by the level of expertise that you have seen
in these introductions. If you don't tell us what you want to know,
we can't help you!

Here is a summary of possible discussion topics that were gleaned from
the introductions (as I see it):

1. How this list can better serve its members.

2. How electronic peer reviewed journals for the publication of
software will affect simulation model development.

3. Models which deal with soil quality in the vadose zone.

4. The integration of Geographic Information Systems with simulation models.

5. The role of models in farming systems research in developing countries.

6. The usefulness of 'simple' spreadsheet models for use in developing countries

7. The combining of several crop models to simulate a cropping system.

8. The integration of meteorological databases and simulation models.

9. How integrate small scale simulation models into a global model.

10. Modeling of nitrate concentrations in tile drainage.

11. What is the state of the art as far as mechanistic models for
plant nutrient uptake?

12. Mechanistic vs. empirical vs. hybrid models

I hope this synopsis is fuel for discussion and that the
introductions will continue.

Sincerely,

Jerome Pier
List Owner
jp@unl.edu

Hello Ken, Soren Hansen is a Dane, not a Netherland. His address
is the Department of Agricultural Sciences, Agricutural Soil and Water
Management, Thorvaldsenvej 40, DK - 1871 Frederiksberg C, Copenhagen,
Denmark with a telephone of +45 35 28 33 86. Alive and well and
living in Copenhagen.

Merry and Happy, Tom Dumper

Dear list subscribers,

With people leaving their terminals behind for the holidays
and since we don't have a ring of geosynchronous communication
satelittes to provide us with email anywhere on earth (sniff!)
many of you will want to suspend the delivery of list mail
to avoid having a run in with your SysAdmin. (What do you mean
I have 2,347 mail messages? I am only subscribed to startrek-l!)
It would appear that there are several flavors of listserv out there
and the commands differ slightly. I am beginning to see people
posting set nomail commands to the list much to their embarassment.
This is not worthy of capital punishment but it does cause confusion
when the subscriber is still receiving mail even though he told]
the server to stop his mail.

Please remember that you send all list commands to listserv@unl.edu
and not to the discussion address. Also, to temporarily suspend
delivery of list mail, send the following command to listserv@unl.edu

set soils-l mail postpone

and substitute agmodels-l where appropriate.

When you get back from utah and are sitting in your chair with your
leg in a cast and want to read your mail, send the following
to listserv@unl.edu:

set agmodels-l

and this should restore everything to default values.

Note: the above was written in a lighthearted manner and no
offense was intended nor was this post directed at anyone
in particular. It is friday afternoon afterall!

Bye,
Jerome Pier jp@unl.edu

How about trying a resistance term for diffusion through the soil layer.
Tom Hodges
> however, the problem we face in RZWQM is that we do not simulate diffusion
> to the root surface. Therefore, all nitrogen in the a given layer is
> available to the root even though it may not be near to the root surface.
>
> The problem gets very difficult when one attempts to consider the
> spatial movement of N across a soil layer as well as through soil layers.

Aloha from Hawaii!

My research involves developing models of phenology and yield of tropical
fruit crops. I also teach a graduate course on crop modeling.

-Kent

>===============================================================<
> Kent Kobayashi, Assoc. Prof. | Dept. of Horticulture <
> kentko@uhunix.uhcc.hawaii.edu | University of Hawaii at Manoa <
> Ph. (808) 956-7876 | 3190 Maile Way, St. John 102 <
> Fax (808) 956-3894 | Honolulu, HI 96822 USA <
>===============================================================<

Hello,
I'm a graduate student with a research focus of examining weed
competition in direct-seeded onions. Part of my study involves studying
the potential for utilizing climatic information, such as would be
available from an automated weather station here in Ontario (specifically,
temperature and rainfall), in the modelling of weed emergence under field
conditions. If anyone has any experience or opinions on the subject.
please drop me a line. Thanks.

Rodger Tschanz
Horticulture Department
University of Guelph
Guelph,Ontario,Canada
N1G 2W1
e-mail address:RTSCHANZ@uoguelph.ca
fax number:(519)767-0755

Claude Dunan was doing a PhD project on weed competition in wheat and
onions at Colorado State Univ. He was developing an onion model. Contact
Frank Moore at horticul@shep.agsci.colostate.edu to see if he is still
there. Tom
Tom Hodges Cropping Systems Modeler ___ ___
USDA-ARS / \_/ \
Rt. 2, Box 2953-A Telephone: 509-786-2226 | |
Prosser, WA 99350 \______/^\/
USA potato tuber
thodges@beta.tricity.wsu.edu thodges@asrr.arsusda.gov
=====================================================================
...photosynthesis makes the world go around... Mr. Potato Head

On Sat, 4 Dec 1993, Roger J Tschanz wrote:

> Hello,
> I'm a graduate student with a research focus of examining weed
> competition in direct-seeded onions. Part of my study involves studying
> the potential for utilizing climatic information, such as would be
> available from an automated weather station here in Ontario (specifically,
> temperature and rainfall), in the modelling of weed emergence under field
> conditions. If anyone has any experience or opinions on the subject.
> please drop me a line. Thanks.
>
> Rodger Tschanz
> Horticulture Department
> University of Guelph
> Guelph,Ontario,Canada
> N1G 2W1
> e-mail address:RTSCHANZ@uoguelph.ca
> fax number:(519)767-0755
>
>
>

Hello
I am writing to introduce myself
I'm a PhD student here at U Tasmania. I'm interested in forest
climates and various modelling approaches to both the water and energy
budgets. Since much progress has come from agricultural modelling
I thought I'd join and watch the fun!
Cheers
Mark

set agmodels-l mail digest

Hello

You may or may not have seen may name when I sent all of you
a request to have messages sent to me in the digest format.

The whole see's when you make a mistake on INTERNET!

I'm Ed Rayburn at West Virginia University. I am interested in
modeling the livestock production from grazing systems. What I'v seen so
far is important and interesting work but are there any others
out there working on animal production models.

Although I joined the group as soon as it was established I did not
introduce myself. I am interested in modelling total farm
systems in developing countries at farm level in which crops are submodels.

Yours sincerely,

GAJENDRA SINGH
Professor and Coordinator
Agricultural & Food Engineering Program
School of Environment, Resources and Development
Asian Institute of Technology
GPO Box 2754, Bangkok 10501, THAILAND

TEL: (66-2)524 5450 * FAX: (66-2)524-6200, 516-2126
INTERNET: singhg@emailhost.ait.ac.th
----------------------------------------------------------------------

INTRODUCTION

My name is Holger Meinke and I am an agricultural research scientist with the
Agricultural Production Systems Research Unit (APSRU) in Toowoomba, Australia
(you know, the part of the world where they spell modelling with two "l"). APSRU is
a multi-disciplinary team with experience in agronomy, plant nutrition, climatology,
crop physiology, economics, soil conservation and communication. While our
primary concern is the grain-grazing system of sub-tropical Australia, we are also
collaborating in projects at a national and international level. Models are our
primary vehicle for communication between disciplines and clients. Model
development is always targeted towards its future use, thus avoiding unnecessary
complexity. APSRU's emphasis is very much on the useful application of models.
For this purpose we have developed a new cropping systems model (APSIM -
Agricultural Production Systems Simulator), a windows based modelling
environment which allows users to easily combine a range of different models or
modules in order to simulate cropping systems.

Within APSRU I am one of the "crop modellers" and have personally been involved
in the development of a sunflower model (QSUN), a peanut and a barley model.
Presently I am testing various wheat models within APSIM in order to improve our
wheat modelling capabilities for our region with emphasis on the water x nitrogen
interactions. I am collecting data sets from around the world for this testing
exercise.

I am looking forward to the discussions going on, particularly in regards to N
uptake and use, but also about the wider aspects of systems modelling and
applications of models. I hope that I will be able (in spite of time constraints) to
actively contribute to the debate.

For any specific questions/responses you can contact me under

holgerm@apsrusg.sth.dpi.qld.gov.au

Holger Meinke

On Sun, 5 Dec 1993, Prof. G. Singh wrote:

>
> Although I joined the group as soon as it was established I did not
> introduce myself. I am interested in modelling total farm
> systems in developing countries at farm level in which crops are submodels.
What models are you interested in using? Mechanistic or more empirical?
What specific models? Tom
Tom Hodges thodges@beta.tricity.wsu.edu

>
> From: MX%"agmodels-l@unl.edu" 4-DEC-1993 07:10
>
> I'm a graduate student with a research focus of examining weed
> competition in direct-seeded onions. Part of my study involves studying
> the potential for utilizing climatic information, such as would be
> available from an automated weather station here in Ontario (specifically,
> temperature and rainfall), in the modelling of weed emergence under field
> conditions. If anyone has any experience or opinions on the subject.
> please drop me a line. Thanks.
>
> Rodger Tschanz
> Horticulture Department
> University of Guelph
> Guelph,Ontario,Canada
> N1G 2W1
> e-mail address:RTSCHANZ@uoguelph.ca
> fax number:(519)767-0755

Rodger-

You might want to talk with Marvin Shaffer in our shop
Shaffer@gpsrv1.gpsrv.colostate.edu, he had a post-doc named Dan Ball
working for him a few years back that did a mutation of the ntrm model.
This contained a multispecies competition algorithm, which I believe
included pigweed, grasses, and other popular natural inhabitants.

later,

Ken

===========================================================================
Ken Rojas
Hydrologist / Environmental Physicist / FORTRAN GOD : Hydrology--
Great Plains System Research Unit :
USDA - Agricultural Research Service . The
301 S. Howes St. P.O. Box E . . Ecologically
Fort Collins, CO 80522 . . Correct
phone: (303) 490-8326 // fax : (303) 490-8310 . . Science
E-mail: KENR@GPSRV1.GPSR.COLOSTATE.EDU . .
ATTmail: A03KROJAS ..... aka: TECS
===========================================================================

>
> From: MX%"agmodels-l@unl.edu" 4-DEC-1993 19:14
> Return-Path: <root@crcnis1.unl.edu>

> You may or may not have seen may name when I sent all of you
> a request to have messages sent to me in the digest format.
>
> I'm Ed Rayburn at West Virginia University. I am interested in
> modeling the livestock production from grazing systems. What I'v seen so
> far is important and interesting work but are there any others
> out there working on animal production models.
>

Give Jon Hanson a line at Jon@gpsrv1.gpsr.colostate.edu, he worked on SPUR-II
which I'm sure you've heard of.

later,

Ken

===========================================================================
Ken Rojas
Hydrologist / Environmental Physicist / FORTRAN GOD : Hydrology--
Great Plains System Research Unit :
USDA - Agricultural Research Service . The
301 S. Howes St. P.O. Box E . . Ecologically
Fort Collins, CO 80522 . . Correct
phone: (303) 490-8326 // fax : (303) 490-8310 . . Science
E-mail: KENR@GPSRV1.GPSR.COLOSTATE.EDU . .
ATTmail: A03KROJAS ..... aka: TECS
===========================================================================

I am working on simulating rangeland production systems. In 1987, we
a model called SPUR. In 1992, we released SPUR2, an upgrade from SPUR.
We have done a lot of work using this model and the papers are just now
beginning to see the light of day. SPUR2 is currently being re-written
so that it works nicely in a MS-Windows environment. That portion of
the project should be done by spring 1995.

If you would like any further dialog, please contact me directly at the
address given below:

+---------------------------------------++---------------------------------+
| Dr. Jon D. Hanson || Comm: (303)490-8323 |
| USDA, Agricultural Research Service || Fax: (303)490-8310 |
| Great Plains Systems Research || jon@gpsrv1.gpsr.colostate.edu |
| 301 S. Howes, P.O. Box E || FTS2000: a03jonhanson |
| Fort Collins, Colorado 80522 || |
+---------------------------------------++---------------------------------+

Hello Folks,

First of all, I would like to congratulate the organizers of this
network conference. It is really exciting to read introduction of and
introduced ourselves to several professional colleagues who desire to
maintain contacts, but 24-hour days do not allow them to do all what they
would like to accomplish. This conference is providing an excellent
mechanism to overcome some of these limitations.

Now, to introduce myself, I am an agricultural engineers (B.Tech and M.
Tech from I.I.T. Kharagpur, India, and Ph.D. from the University of
Florida, Gainesville, FL). My current interests include using site-
specific models (crop growth, sediment, water quality, and pesticide drift,
and others) for landscape analysis and regional planning. At ManTech
Environmental Technology, Inc. (the on-site contractor for the USEPA-
Environmental Research Laboratory, Corvallis, OR) I am working with the
Wetlands Research Program. The program is implementing a risk-based
framework and associated databases for wetland protection and management.
We are developing approaches, methodologies, and case-studies for
ecological risk assessment of agricultural stressors on wetland functions
at thelandscape level. We are currently evaluating and selecting
sediment, pesticide drift, and water quality models for regional risk
assessment; and conducting surveys and compiling data to parameterize and
run these models for our study sites.

Prior to moving to Oregon, at the University of Florida, I worked with two
multi-disciplinary projects for developing approaches, methodologies and
applications of combining site-specific agricultural, water quality, and
economic models and expert systems with GIS for regional risk assessment,
planning and policy analysis, and decision support systems. The first
project, funded by the South Florida Water Management District, developed
LOADSS -Lake Okeechobee Agricultural Decision Support System. This UNIX-
based system stores several layers of spatial and attribute data for the
basin and can access a number of water quality, and dairy
management models. It allows users to assess the risk of current level of
phosphorus loadings to the lake water quality, and also interactively
evaluate the effects and cost effectiveness of different options (Best
Management Practices) for reducing phosphorus loading to the lake. In the
other project, funded by the Caribbean Basin Administrative Group, we
designed and developed AEGIS - Agricultural and Environmental Geographic
Information System, that integrated crop growth and soil erosion models
with the spatial soil, land use and weather databases for three sites in
Puerto Rico. The purpose of this system was to evaluate plans for meeting
regionalproduction goals with minimal soil loss.

As part of my Ph.D., I designed, developed, and evaluated a whole-farm
planning and decision support system (FARMSYS) for evaluating the
effects of management strategies on resource use efficiency, crop
yields, and farm profits.

I am really enjoying to be a part of this conference, and hope every one
else. I would be glad to share of any information on any of above systems.

Thanks and Happy Holidays,

Harbans Lal
ManTech Environmental Technology, Inc.
Corvallis, OR 97333
Ph (503) 754-4666 (Work)
Ph (503) 752-8320 (Res)
Fax (503) 754-4335

>How about trying a resistance term for diffusion through the soil layer.
>Tom Hodges
>> however, the problem we face in RZWQM is that we do not simulate diffusion
>> to the root surface. Therefore, all nitrogen in the a given layer is
>> available to the root even though it may not be near to the root surface.
>>
>> The problem gets very difficult when one attempts to consider the
>> spatial movement of N across a soil layer as well as through soil layers.

I appreciate the brevity of your suggestion, but I, for one, am not so
familiar with the models such that I can follow the full significance of a
cryptic suggestion. I can take a good guess, but I might miss the
significance, and I may not be as imaginative next time. Please coddle us
a little!
Cheers,
R. H. (Dick) Richardson
Zoology Dept.
Univ. of Texas
Austin, TX 78712

email: dickr@utxvm.cc.utexas.edu
office: 512-471-4128
home: 512-476-5131
FAX: 512-471-9651

My name is Ranjan Muttiah. I work here at the Blacklands Res. center
(Texas A&M) on soil & water and remote sensing topics. Lately, I've been
developing algorithms to extract recession constants from stream gauges
readings. The model we work with is called Soil and Water Assessment Tool
(SWAT). The model is the old SWRRB and ROTO (routing) models integrated
with GIS. My previous exposure to a rootzone model was PRZM, and to
groundwater models was MODFLOW, and SAFTMOD. I've been reading through
the "newly" supported EPA's model called FEMWATER.

I liked the discussion on nitrogen uptake. Many of the (theoretical)
discussions on the topic in groundwater has been on using the standard
D-A equation (tho' I haven't seen reasonable estimates on dispersion,
and adsorbtion coefficients -- if the latter is indeed important).

Ranjan

This may be of interest to the list.
Tom Hodges---------- Forwarded message ----------
Date: Mon, 6 Dec 1993 12:34:59 -0500 (EST)
From: Gabriel Hegyes <ghegyes@nalusda.gov>
To: SANET-mg <sanet-mg@ces.ncsu.edu>
Subject: Peer Reviewers for NRI's Ag Systems Program (fwd)

---------- Text of forwarded message ----------
Date: Mon, 6 Dec 1993 11:27:00 EST
From: Peter Johnson 202-401-1896 <JOHNSON@a1.darth.esusda.gov>
To: irm@ag.srnr.Arizona.EDU
Subject: Peer Reviewers for NRI's Ag Systems Program

The USDA National Research Initiative Competitive Grants Program is seeking
the names of researchers to be ad hoc reviewers and/or panelists for its
new "Agricultural Systems" program. Ad hoc reviewers would be sent no more
than 2 or 3 proposals to evaluate between April and June of 1994.
Responses can be e-mailed to: Dr. Peter J. Johnson
Johnson@CSRS.ESUSDA.GOV
A follow-up questionnaire will be mailed to each respondent.
Information on proposal submission can also be solicited. The submission
deadline is Feb. 22, 1994 . Thank you.

If you have been following this thread you'll remember that nitrogen
uptake can be handled with either/both active and passive uptake
processes, in my conception. I'd like to address two issues in this query:

1) Is there a preference within the plant uptake process, for NH4 or NO3?

2) What percentage of total daily uptake of N would be from active versus
passive uptake?

The obvious limitation to the preference for nh4 or no3 is the adsorption
of nh4 to soil colloids and organic matter. Not to mention that the nh4
is not readily available in large quantities relative to no3. Could you
soil fertility guys staighten me out on this?

In reference to this passive/active uptake percentage, we use an estimate
somewhere around 90% passive to 10% active. Any refinement on this would
be appreciated.

later,

Ken

===========================================================================
Ken Rojas
Hydrologist / Environmental Physicist / FORTRAN GOD : Hydrology--
Great Plains System Research Unit :
USDA - Agricultural Research Service . The
301 S. Howes St. P.O. Box E . . Ecologically
Fort Collins, CO 80522 . . Correct
phone: (303) 490-8326 // fax : (303) 490-8310 . . Science
E-mail: KENR@GPSRV1.GPSR.COLOSTATE.EDU . .
ATTmail: A03KROJAS ..... aka: TECS
===========================================================================

Hi, I am a grad student in resource economics working on a model of corn
production in the U.S. It is a bit different from the kind of ag models
you folks have been discussing but my interest is in runoff of excess N.
At the moment I am modelling this as (N applied)-(quantity of corn
harvested * K) where K is the minimum amount of N applied per ton of corn
harvested in the region. The model has four regions in the continental
U.S. Each region includes categories for dryland-irrigated and
participation in commodity programs or not. It is an offshoot of the
USDA/ERS/RTD USARM model, which you may be familiar with. My ultimate
goal is to model farmers responses to different forms of N regulation.

I know very sophisticated models exist to get at the issue of excess N but
what I need is a one line formula with one or two parameters that will
give me a rough idea of how much is too much. I would appreciate any
suggestions, or comments on the approach I am taking. Thanks.

-- Drew Laughland, Penn State

School of Agriculture & Forestry 7/12/93
14:16
SUBJECT: Introduction
G'day!

Introduction:
------------------
Name: Glyn Rimmington
Title: Senior Lecturer - Agricultural Systems
Affiliation: The University of Melbourne
Faculty of Agriculture and Forestry
Department of Agricultural Science
Crop Improvement Centre
Address: Parkville, Victoria 3052, AUSTRALIA
Phone: +61 3 344 5021
Fax: +61 3 347 7359
Email: gmr@agriculture.unimelb.edu.au

Background:
Basic training in science (BSc) with majors in botany -- mainly
ecophysiology, ecology, biometeorology -- and modelling -- computer
science, biometry, simulation, maths, statistics -- at the University
of Queensland, followed by PhD studies on effects of fluoride (mainly
from aluminium smelters) air pollution on plant growth and
development at UofQ, followed by a post doc in Agricultural
Engineering at The University of Melbourne, working on databases and
models of the wheat crop, then appointment to present position within
the Department of Agriculture in Agricultural Systems. Have been
lucky enough to have been able to visit fellow modellers at UCDavis,
Kansas SU, Michigan SU, Uof Hawaii, USDA-ARS Systems Lab Beltsville,
UFlorida Gainesville, Rothamsted and PRI Hurley UK, Taiwan and
PRChina as well as groups down under.

Modelling philosophy:
Most similar to Graeme Hammer (Toowoomba Qld colleague of Holger
Meinke), Tom Sinclair (USDA Florida) and Richard Vanderlip (Kansas
SU) [They may disagree ;-) ]. Influenced most by Charles-Edwards
(retired CSIRO Qld formerly GCRI Littlehampton), Thornley (UK),
Loomis (UC Davis), Connor (UofMelbourne) and various people from the
Wageningen group. Most different to Ritchie (Michigan SU) and Baker
(USDA Starkville Mississippi). The world would be boring if some of
us didn't agree on things. ;-)

Areas of Interest:
Modelling in general, as applied to biological systems -- ranging
from light relations, crop yield forecasting, through to simulating
effects of introducing new technology to the Ugandan dairy industry.
Computer Applications in Agriculture -- networks, simulation, GIS,
expert systems, spreadsheet modelling, DBMS, graphics, data
acquisition and control, multimedia, computer-aided learning, image
processing, remote sensing, animation and virtual reality.
Favourite area is L-Systems for visualisation of growth and
development and investigation of light relations.

Current Projects:
LandcareNET - a modem-based electronic network for rural users
testing hypotheses about problems with
adoption of networks
CEREAL GAME - a decision support system for cereal growers with
emphasis on
interface design and development
methodology
CAL by stealth ;-)
Virtual Sunflower - an L-system-based model of the sunflower plant

> I know very sophisticated models exist to get at the issue of excess N but
> what I need is a one line formula with one or two parameters that will
> give me a rough idea of how much is too much. I would appreciate any
> suggestions, or comments on the approach I am taking. Thanks.
>
> -- Drew Laughland, Penn State
>

I would like to suggest the use of extension recommendations for
N fertilizer application for your area of interest. This will
make the process modelers wince but it could fill your needs.
These guidelines are empirically based and usually derived from
observation of fertility trials with popular crop varieties over
a period of years. This would give you a simple equation which
you could reference and throw into your spatial model to screen
for 'trouble' spots. Let me know what you think.

Jerome Pier jp@unl.edu

> If you have been following this thread you'll remember that nitrogen
> uptake can be handled with either/both active and passive uptake
> processes, in my conception. I'd like to address two issues in this query:
>
> 1) Is there a preference within the plant uptake process, for NH4 or NO3?
>

My understanding is that corn takes up mostly NO3 but some
varieties take up NH4 in significant quantaties. The varieties
which take up NH4 are usually high yielding varieties. NH4 uptake
is greatest early in the growing season. Also, NH4 uptake appears
to increase following short periods of N starvation (Lee et al.
1992. Nitrogen assimilation and the control of ammonium and
nitrate absorption by maize roots. J. Exp. Botany 43:1385-1396.).
Nevertheless, NO3 appears to be taken up substantially more than
NH4. Anyone else care to comment?

> 2) What percentage of total daily uptake of N would be from active versus
> passive uptake?
>
Good question! I think it depends on the growth stage of the
plant, the soil water content, N availablity, etc. I am looking
over a pair of papers which examined diffusion and mass flow of
nitrate to plant roots (corn) in the green house and field. The
bottom line of these papers appears to be:

"The relative importance of diffusion to mass flow of
NO3-N is large when the ration of transpiration rate to the
constant of proportionality relating flux into the plant to
concentration of NO3-N in the soils solution is less than 0.2 but
is negligible when the ratio is greater than unity."

R.E. Phillips, et al. 1976. Diffusion and mass flow of
nitrate-nitrogen to plant roots. Agron. J. 68:63-66.

> The obvious limitation to the preference for nh4 or no3 is the adsorption
> of nh4 to soil colloids and organic matter. Not to mention that the nh4
> is not readily available in large quantities relative to no3. Could you
> soil fertility guys staighten me out on this?
>
>
> In reference to this passive/active uptake percentage, we use an estimate
> somewhere around 90% passive to 10% active. Any refinement on this would
> be appreciated.
>
> later,
>
> Ken
>
Jerome Pier jp@unl.edu

Introduction
**************
My name is Alan Munford and I work as lecturer in statistics and
OR at the University of Exeter, in South West England.

I am interested in livestock modelling, and I am the author of a
commercial livestock modelling and least cost feed formulation
package called Ultramix. Ultramix is a general purpose (species
independent) system, and consists of 4 integrated modules:
feed database manager, spreadsheet-like modeller,
linear programming optimiser, and report writer.
It is used by many universities and colleges in the
UK, by the government advisory service, and by many of the larger
animal feed companies.

I would be interested to make contact with others intested in the
general area of feed fromulation - all species (pigs, poultry, ruminants).

--
Alan Munford JANET: agm@uk.ac.exeter.msor
MSOR Dept, Exeter University BITNET: agm%uk.ac.exeter.msor@ukacrl
Laver Building, North Park Road, Tel: +44 392 264470 (home 215680)
EXETER, UK. EX4 4QE Fax: +44 392 264460

> >How about trying a resistance term for diffusion through the soil layer.
> >Tom Hodges
> >> however, the problem we face in RZWQM is that we do not simulate diffusion
> >> to the root surface. Therefore, all nitrogen in the a given layer is
> >> available to the root even though it may not be near to the root surface.
>
> I appreciate the brevity of your suggestion, but I, for one, am not so
> familiar with the models such that I can follow the full significance of a
> cryptic suggestion. I can take a good guess, but I might miss the
> significance, and I may not be as imaginative next time. Please coddle us
> a little!

> Cheers,
> R. H. (Dick) Richardson

I would like to thank Dick for reminding us that many (>180!)
people are reading this list and have an interest in the
discussion but are not completely familiar with the subject. Mind
if I take a stab?

I beleive the debate on modeling N uptake is whether N uptake by
plant roots can be modelled as simply the absorption of NO3-N and
NH4-N from the soil solution via the transpiration stream or if
there should be some consideration of N concentration gradients
which might develop due to root N uptake thereby causing the
diffusion of N (as opposed to mass flow/convection) to be a
significant process. In the late 70's and early 80's, there was a
great deal of effort on deriving an equation to describe plant
nutrient uptake based on the convective-dispersion equation. This
work culminated in the Barber-Cushman model (S.A. Barber. 1984.
"Soil Nutrient Bioavailability: a Mechanistic Approach"
Wiley-Interscience, N.Y., N.Y.) which is a numerical solution of
the convective-dispersion equation for non-steady state
conditions.

The above comment posted by Tom Hodges suggests that maybe there
should be some consideration of diffusion processes in the N
uptake model being discussed. Now exactly how one would model
that is a question the plant modelers on the list will have to
address but diffusion all hinges on the buffer or nutrient supplying
power of the soil, the tortuosity, and the diffusion coefficient
of the molecule in water.

Well, I'm not sure if I have helped or made matters worse since I
know just enough about this subject to be dangerous. Let me know.

Jerome Pier jp@unl.edu

In response to Ed Rayburn, YES, there are animal production modelers out there.
I'm one of them. Beef cattle production is my focus.

Rick Bourdon
Colorado State University

On Tue, 7 Dec 1993, jerome pier wrote:

> > If you have been following this thread you'll remember that nitrogen
> > uptake can be handled with either/both active and passive uptake
> > processes, in my conception. I'd like to address two issues in this query:
> >
> > 1) Is there a preference within the plant uptake process, for NH4 or NO3?
> >
>
> My understanding is that corn takes up mostly NO3 but some
> varieties take up NH4 in significant quantaties. The varieties
> which take up NH4 are usually high yielding varieties. NH4 uptake
> is greatest early in the growing season. Also, NH4 uptake appears
> to increase following short periods of N starvation (Lee et al.
> 1992. Nitrogen assimilation and the control of ammonium and
> nitrate absorption by maize roots. J. Exp. Botany 43:1385-1396.).
> Nevertheless, NO3 appears to be taken up substantially more than
> NH4. Anyone else care to comment?

Yes. One can find in nature plant species/varieties differing in
their perference for NH4 and NO3. It is my experience that during the fall,
winter, and spring, Oregon soil NH4/NO3 ratios are high (much >1). Ryegrass
grown under these conditions uses NH4 as a major N source. This is partly
due to higher levels of soil-NH4 levels but also because NH4 uptake at low
temperatures is greater for NH4 than for NO3 in ryegrass. Therefore, in
ryegrass ammonium uptake should be considered as well as NO3, or even more
so. By the time soil NO3 levels rise above NH4 levels, around May, the
ryegrass has already taken up most (80-90%) of its total N. I have also
found that ryegrass overall grows faster with higher NH4/NO3 ratios
whereas, Kentucky bluegrass grows faster with higher NO3/NH4 ratios.

It is often stated in the literature that "Most of the N of higher plants
is taken up as nitrate which is reduced to NH4 and simultaneously utilized
in the synthesis of amino acids and amides."

In modeling N uptake etc., should more consideration be given NH4 utilization?
What is our depth of knowledge of root ammonium assimilation of
NON-LEGUMES?? How does our knowledge of nitrate physiology compare to the
total knowledge of ammonium physiology in non-legumes?

> > 2) What percentage of total daily uptake of N would be from active versus
> > passive uptake?
> >
> Good question! I think it depends on the growth stage of the
> plant, the soil water content, N availablity, etc. I am looking
> over a pair of papers which examined diffusion and mass flow of
> nitrate to plant roots (corn) in the green house and field. The
> bottom line of these papers appears to be:
>
> "The relative importance of diffusion to mass flow of
> NO3-N is large when the ration of transpiration rate to the
> constant of proportionality relating flux into the plant to
> concentration of NO3-N in the soils solution is less than 0.2 but
> is negligible when the ratio is greater than unity."
>
> R.E. Phillips, et al. 1976. Diffusion and mass flow of
> nitrate-nitrogen to plant roots. Agron. J. 68:63-66.
>
> > The obvious limitation to the preference for nh4 or no3 is the adsorption
> > of nh4 to soil colloids and organic matter. Not to mention that the nh4
> > is not readily available in large quantities relative to no3. Could you
> > soil fertility guys staighten me out on this?
> >
> >
> > In reference to this passive/active uptake percentage, we use an estimate
> > somewhere around 90% passive to 10% active. Any refinement on this would
> > be appreciated.
> >
> > later,
> >
> > Ken
> >
> Jerome Pier jp@unl.edu

Dr. Stephen M. Griffith
Research Plant Physiologist
National Forage Seed Production Res. Ctn.
USDA-ARS
3450 SW Campus Way
Corvallis, OR 97331
Ph: 503-750-8742
Fax: 503-750-8750
Internet: griffits@ucs.orst.edu

First off I am not a modeler. My general research interest is grass
physiology/biochemistry. I am interested in cool-season (temperate) grass
N nutrition of both forage and turf types.

I am particularly interested in N fertilizer use efficiency, regulation
of plant growth and root physiology. I work mainly with the following grasses;
tall fescue, perennial ryegrass, Italian ryegrass, and Kentucky bluegrass.
I have been screening grass species and varieties for NH4 and NO3 use. I have
also been examining NO3 and NH4 use under field and hydroponic conditions.

These grasses (above) are grown for seed in the Pacific Northwest. Oregon
is the world's largest producer of temperate grass seed.

Name: Wesley Rosenthal
Affiliation: Blackland Research Center/Texas Agric. Expt. Stn.
Areas of Interest: Crop simulation modeling--sorghum,cotton &
Hydrologic modeling and applications
Email address: rosentha@brcsu0.tamu.edu

Recent areas of interest have dealt with hydrologic modeling applications on a river basin scale. Currently working on linking crop models with GIS to evaluate spatial production. This work is currently being conducted with the Dept. of Primary
Industries in Queensland Australia.

----------
X-Sun-Data-Type: text
X-Sun-Data-Description: text
X-Sun-Data-Name: text
X-Sun-Content-Lines: 32

Name: Daniel Wallach
Affiliation: INRA (Institut National de la Recherche Agronomique)
Biometry Department. Tououlse, France.
Email address: wallach@toulouse.inra.fr
Areas of interest: Statistics of crop models (parameter estimation,
evaluation of predictive quality). Statistics of optimal
decisions in agriculture.

I haven't yet seen another statistician introduce himself since I've
been following this group. I think a) that reflects the true level of
cooperation between modelers and statisticians b) that's a pity for
both.

My philosophy in a nutshell - A model can be expressed as
E[Y]=f(X,theta) where Y is what is being calculated (yield, or runoff,
or a whole list of things), X are the driving variables (again a whole
list of them), theta are the parameters, and f is some function
(usually embodied in a few pages of computer code). There is no
randomness here because what is being calculated is the expectation of
the result over some population E[Y].

The model with the randomness of the real world added is
Y=f(x,theta)+epsilon, with epsilon being a random error. The modeler
works on f(X,theta), and the statistician on epsilon. The way epsilon
is treated determines how the parameters theta should be estimated
from data, and how the predictive accuracy of the model can be
estimated.

I hope that hasn't just reinforced your convictions that statisticians
should at all costs be avoided.

----------
X-Sun-Data-Type: default
X-Sun-Data-Description: default
X-Sun-Data-Name: agmod.intro
X-Sun-Content-Lines: 31

Name: Daniel Wallach
Affiliation: INRA (Institut National de la Recherche Agronomique)
Biometry Department. Tououlse, France.
Email address: wallach@toulouse.inra.fr
Areas of interest: Statistics of crop models (parameter estimation,
evaluation of predictive quality). Statistics of optimal
decisions in agriculture.

I haven't yet seen another statistician introduce himself since I've
been following this group. I think a) that reflects the true level of
cooperation between modelers and statisticians b) that's a pity for
both.

My philosophy in a nutshell - A model can be expressed as
E[Y]=f(X,theta) where Y is what is being calculated (yield, or runoff,
or a whole list of things), X are the driving variables (again a whole
list of them), theta are the parameters, and f is some function
(usually embodied in a few pages of computer code). There is no
randomness here because what is being calculated is the expectation of
the result over some population E[Y].

The model with the randomness of the real world added is
Y=f(x,theta)+epsilon, with epsilon being a random error. The modeler
works on f(X,theta), and the statistician on epsilon. The way epsilon
is treated determines how the parameters theta should be estimated
from data, and how the predictive accuracy of the model can be
estimated.

I hope that hasn't just reinforced your convictions that statisticians
should at all costs be avoided.

I would like to thank Daniel Wallach for joining the list and
look forward to his contributions on how to account for
variability in models. On that note, I was wondering if, Daniel,
you would post a brief explanation on what Monte Carlo Methods
are all about. I have heard of this technique being used to add
an 'epsilon' component to the predictive equation; am I correct?

Thanks in Advance,

Jerome Pier

I'd like to second your suggestion, Jerome. It would be interesting to
discuss kinds and approaches to modeling. Not only Monte Carlo modeling
would be nice to fit into the discussion, but chaotic modeling as well. I
don't know if anyone would like to compare the features of complex models
with traditionally more "tractable" approaches or not. I'd find it
interesting, since my modeling days were in genetics and ecology rather
than physiology. But, to me, there are even more directions for using
mathematics and statistics in biology than the "epsilon - function"
dichotomy. Today we have a new kid on the block -- chaos. It's appeal to
me is that some features of both the stochastic and deterministic
approaches are found in chaos -- deterministic but unpredictable outcome.
Is it, however, even more untractable?
Dick

>I would like to thank Daniel Wallach for joining the list and
>look forward to his contributions on how to account for
>variability in models. On that note, I was wondering if, Daniel,
>you would post a brief explanation on what Monte Carlo Methods
>are all about. I have heard of this technique being used to add
>an 'epsilon' component to the predictive equation; am I correct?
>
>Thanks in Advance,
>
>Jerome Pier
R. H. (Dick) Richardson
Zoology Dept.
Univ. of Texas
Austin, TX 78712

email: dickr@utxvm.cc.utexas.edu
office: 512-471-4128
home: 512-476-5131
FAX: 512-471-9651

Received: by ccmail from gaia.ucs.orst.edu
>From root@crcnis1.unl.edu
X-Envelope-From: root@crcnis1.unl.edu
Received: from crcnis1.unl.edu by gaia.ucs.orst.edu with SMTP id AA26389
(5.67a/IDA-1.5); Wed, 8 Dec 1993 15:02:10 -0800
Received: by crcnis1.unl.edu id AA14252
(5.65c/IDA-1.4.4); Wed, 8 Dec 1993 17:00:13 -0600
Date: Wed, 8 Dec 1993 17:00:13 -0600
Message-Id: <199312082256.AA14145@crcnis1.unl.edu>
Originator: agmodels-l@unl.edu
Errors-To: jp@unlinfo.unl.edu
Reply-To: <agmodels-l@unl.edu>
Sender: agmodels-l@unl.edu
Version: 5.5 -- Copyright (c) 1991/92, Anastasios Kotsikonas
From: dickr@utxvm.cc.utexas.edu (R. H. (Dick) Richardson)
To: Multiple recipients of list <agmodels-l@unl.edu>
Subject: Re: Monte Carlo Methods for Modeling

I'd like to second your suggestion, Jerome. It would be interesting to
discuss kinds and approaches to modeling. Not only Monte Carlo modeling
would be nice to fit into the discussion, but chaotic modeling as well. I
don't know if anyone would like to compare the features of complex models
with traditionally more "tractable" approaches or not. I'd find it
interesting, since my modeling days were in genetics and ecology rather
than physiology. But, to me, there are even more directions for using
mathematics and statistics in biology than the "epsilon - function"
dichotomy. Today we have a new kid on the block -- chaos. It's appeal to
me is that some features of both the stochastic and deterministic
approaches are found in chaos -- deterministic but unpredictable outcome.
Is it, however, even more untractable?
Dick

Definitely more intractable. Deterministic in the sense that a given run will
produce a give time series output, but intractable in that chaotic systems are
extremely sensitive to initial conditions. To the extent, that we are unable to
exactly define initial conditions (not even close, frequently), chaotic systems
are problematic from a predictive point of view. At best we can learn something
from bifurcations or attractors that lie on the path to chaos or in chaotic
regions of "phase-space". While others may disagree and contrary to some
popularly expressed views, my own opinion is that the majority of systems of
interest to most biological modelers do not reside in chaotic regions, but are
rather most appropriately analyzed deterministically or stochastically. Still,
the concept of chaos is appealing and clearly is applicable to some biological
systems.

John Bolte
Bioresource Engineering
Oregon State University
boltej@ccmail.orst.edu

>I would like to thank Daniel Wallach for joining the list and
>look forward to his contributions on how to account for
>variability in models. On that note, I was wondering if, Daniel,
>you would post a brief explanation on what Monte Carlo Methods
>are all about. I have heard of this technique being used to add
>an 'epsilon' component to the predictive equation; am I correct?
>
>Thanks in Advance,
>
>Jerome Pier
R. H. (Dick) Richardson
Zoology Dept.
Univ. of Texas
Austin, TX 78712

email: dickr@utxvm.cc.utexas.edu
office: 512-471-4128
home: 512-476-5131
FAX: 512-471-9651

I think that things are getting pretty confused here. First of all,

>It would be interesting to
>discuss kinds and approaches to modeling. Not only Monte Carlo modeling
>would be nice to fit into the discussion, but chaotic modeling as well.

These are apples and oranges. Monte Carlo methods let one deal with
parameter uncertainties and offer an alternative to analytic sensitivity
analysis, and they can be applied to any kind of model (cf. Silvert, W.
1979. Practical curve fitting. Limnol. Oceanogr. 24:767-773.).

Chaos modelling is quite different, in that chaos is a real phonemenon
and one should model it if it exists in a system, and if not, don't.

<[chaos is]
>Definitely more intractable. Deterministic in the sense that a given run will
>produce a give time series output, but intractable in that chaotic systems are
>extremely sensitive to initial conditions. To the extent, that we are unable to
>exactly define initial conditions (not even close, frequently), chaotic systems
>are problematic from a predictive point of view. At best we can learn something
>from bifurcations or attractors that lie on the path to chaos or in chaotic
>regions of "phase-space". While others may disagree and contrary to some
>popularly expressed views, my own opinion is that the majority of systems of
>interest to most biological modelers do not reside in chaotic regions, but are
>rather most appropriately analyzed deterministically or stochastically. Still,
>the concept of chaos is appealing and clearly is applicable to some biological
>systems.

Chaos is not necessarily intractable. Chaotic systems are often quite
well-behaved and predictable, but not EXACTLY predictable. Systems
which exhibit high-order fixed point behaviour, like 7-year locusts, may
well be representable by chaotic models.

Dear Ag-models list,

Has anyone out there in ag-model land attempted to incorporate
users in the model building process? I am interested in learning
about participatory approaches to science and technology
development and would be interested in hearing of other people's
experiences along these lines. Correct me if I am wrong, but it
seems that the dominant approach in developing models is for
subject area "experts" to build models based upon assumptions that
they and their peers are comfortable with. Then they turn the
model over to software experts who put some "user-friendliness"
onto it. Then some users (farmers, policy makers, etc.) get to
push the buttons, relatively blind to assumptions and limitations of
the equations within the model. Then the modeler gets to
complain that the users are "misusing" the models. Perhaps limits
could be put on certain parameters within the model to prevent
some misuse and to maintain the model builder's control over the
model as well as maintaining the users' blindness and dependence.
Rarely, I think, do complex models provide confidence intervals
for their predictions in novel situations. There may not be a
reliable basis for constructing such confidence intervals. Such
confidence intervals might give users a tool for judging whether or
not model predictions are meaningful for their particular
applications.

An alternative to the top-down, expert driven approach that I've
caricatured above, might be to help users build their own models
based on their own assumptions. This way, the users are not
dependent upon the assumptions of the "expert" modelers. Larry
Smarr, the Director of the National Center for Super Computing
at the University of Illinois, gave a speech about a year ago where
he envisioned different stake holders in social disputes coming
together and using modeling to work out their problems. In this
vision, street gangs, politicians, scientists, planners, and regular
citizens, could address their problems by coming together to either
build models of what is, or what could be. I wonder if a similar
vision could be developed for agriculture. Rather than developing
models, perhaps modeling tools and environments could be
developed so that farmers, ranchers, land-managers, etc., could
build and test models that they think best fit their assumptions and
situations. It could be a good learning experience for all. You
may say that I'm a dreamer, but I'm not the only one. I've been
reading Thomas Jefferson lately.

I realize that these ideas may be controversial. But when has
"progress" ever made by sticking only to the tried and true?

Best regards,

Greg McIsaac
Senior Research Specialist (Soil erosion and water quality)
Agricultural Engineering
University of Illinois
Urbana, Illinois

Received: by ccmail from gaia.ucs.orst.edu
>From root@crcnis1.unl.edu
X-Envelope-From: root@crcnis1.unl.edu
Received: from crcnis1.unl.edu by gaia.ucs.orst.edu with SMTP id AA24529
(5.67a/IDA-1.5); Fri, 10 Dec 1993 13:21:59 -0800
Received: by crcnis1.unl.edu id AA11294
(5.65c/IDA-1.4.4); Fri, 10 Dec 1993 15:19:51 -0600
Date: Fri, 10 Dec 1993 15:19:51 -0600
Message-Id: <MAILQUEUE-101.931210110156.480@age2.age.uiuc.edu>
Originator: agmodels-l@unl.edu
Errors-To: jp@unlinfo.unl.edu
Reply-To: <agmodels-l@unl.edu>
Sender: agmodels-l@unl.edu
Version: 5.5 -- Copyright (c) 1991/92, Anastasios Kotsikonas
From: "Greg McIsaac" <GFM@age2.age.uiuc.edu>
To: Multiple recipients of list <agmodels-l@unl.edu>
Subject: model builders and users

>An alternative to the top-down, expert driven approach that I've
>caricatured above, might be to help users build their own models
>based on their own assumptions. This way, the users are not
>dependent upon the assumptions of the "expert" modelers. Larry
>Smarr, the Director of the National Center for Super Computing
>at the University of Illinois, gave a speech about a year ago where
>he envisioned different stake holders in social disputes coming
>together and using modeling to work out their problems. In this
>vision, street gangs, politicians, scientists, planners, and regular
>citizens, could address their problems by coming together to either
>build models of what is, or what could be. I wonder if a similar
>vision could be developed for agriculture. Rather than developing
>models, perhaps modeling tools and environments could be
>developed so that farmers, ranchers, land-managers, etc., could
>build and test models that they think best fit their assumptions and
>situations. It could be a good learning experience for all. You
>may say that I'm a dreamer, but I'm not the only one. I've been
>reading Thomas Jefferson lately.
>
>I realize that these ideas may be controversial. But when has
>"progress" ever made by sticking only to the tried and true?
>
>best regards,
>
>Greg McIsaac
>Senior Research Specialist (Soil erosion and water quality)
>Agricultural Engineering
>University of Illinois
>Urbana, Illinois

Greg - this is a great idea, but very difficult to implement with
current technology and modeling mindsets. Because the current model
building process is so laborious and assumption-intensive, the task
of interactively creating models with sufficient basis and resolution
to be useful is a non-trivial at best. Hope may be in sight in the
area of object-oriented model construction, where component
preconstructed models with well-defined, standardized interfaces which
can be interactively "mixed and matched" to reflect different problem
assumptions and viewpoints are at least technically feasible at this
point. Beyond the technical challenge is a potentially more severe
"mentality of model construction" problem, getting modelers to buy
into object and communication standards which will force a rethinking
of our fundamentel approaches to model building.

- John Bolte
Oregon State University
Bioresource Engineering Dept
(boltej@ccmail.orst.edu)

Dear Soils-l and Agmodels-l Subscribers,

I am passing on this information from Dr. Glendon Gee who is in
charge of the Soil Physics Newsletter. Dr. Gee has moved into the
Information age by making the Newsletter accessable thru a
Listserv list. Let's hope that the other divisions of the
Society follow suit. Below is the information which describes the
Newsletter and how to subscribe to the listserv. NOTE: this is
NOT a Soil Physics discussion group. For those who would just
like to receive the Newsletter without subscribing, I have
offered to post the Newsletters on both Soils-l and Agmodels-l as
long as no one objects.

Jerome Pier List Owner
jp@unl.edu

------------------------------Cut----Here------------------------

Welcome to S1-NEWS, the Pacific Northwest Laboratory's
Email Discussion list for
Division S1 of the Soil Science Society of America

Overview

The S1-NEWS group has been set up to provide electronic communication
for the soil physics division (S-1) of the Soil Science Society of
America. This communication link is organized to cover topics of
mutual interest and to act as a supplement to the Agronomy Journal and
other official organs of the SSSA to communicate things of interest to
the S-1 division.

This is an initial attempt to provide an electronic network service to
members of the society interested in soil physics topics. An action
item at the recent S-1 business meeting was for Glendon Gee to
initiate a newsletter that would provide current news to the S-1
members. It was felt that INTERNET could be the vehicle for providing
that service. E-mail addresses were requested at the business meeting
from those who attended.

Since the S-1 business meeting, Bill Jury has taken the initiative to
secure e-mail addresses and circulate them over INTERNET. You should
have received his several messages sent in November. This is the next
step in the process of setting up an electronic newsletter and a rapid
communication link.

In the interim, the S-1 planning committee has been communicating
about activities for the next annual meetings and items that should be
brought up to the SSSA executive board meeting.

Any response on the past annual meetings, notices of upcoming meetings
of interest to S-1 members, or other subjects of general interest will
be welcome. Please send information to the LISTSERV address described
below.

A very simple process for a free "subscription" to the LISTSERV group,
S1-NEWS, is described below.



Messages mailed to S1-NEWS@VOGLNP.PNL.GOV are distributed to all list
subscribers. Please do not send messages longer than about 500 lines;
break long messages up into multiple shorter messages.


Selected LISTSERV Commands

To join S1-NEWS, send the following e-mail message to
LISTSERV@VOGLNP.PNL.GOV:

SUBSCRIBE S1-NEWS First_Name Last_Name

Example: SUBSCRIBE S1-NEWS Joan Smith

To sign off S1-NEWS, send the following e-mail message to
LISTSERV@VOGLNP.PNL.GOV:

UNSUBSCRIBE S1-NEWS

To stop S1-NEWS mail when you go on vacation, send the following
e-mail message to LISTSERV@VOGLNP.PNL.GOV:

SET S1-NEWS MAIL POSTPONE

To resume S1-NEWS mail delivery, send the following e-mail message to
LISTSERV@VOGLNP.PNL.GOV:

SET S1-NEWS MAIL NOACK

To receive groups of messages instead of each as it is sent to S1-NEWS,
send the following e-mail message to LISTSERV@VOGLNP.PNL.GOV:

SET S1-NEWS MAIL DIGEST

To send a message to the S1-NEWS, send your e-mail message to

S1-NEWS@voglnp.pnl.gov

To obtain a list of S1-NEWS participants, send the following e-mail
message to LISTSERV@VOGLNP.PNL.GOV:

REVIEW S1-NEWS

To get help on LISTSERV commands, send the following e-mail message to
LISTSERV@VOGLNP.PNL.GOV:

HELP

We have attempted to include end-users in the model development cycle for
a couple of different efforts. Not always has it been possible, however,
to determine who the end-user of the model will be. In a very research
oriented model, we try to capture the science and are not always concerned
if someone will actually use the model. However, in the development
of systems that will support management decisions (Decision Support Systems)
the end-user is very important. In this case, the user should be included
as part of the development team. We are currently doing this at our
location.

+---------------------------------------++---------------------------------+
| Dr. Jon D. Hanson || Comm: (303)490-8323 |
| USDA, Agricultural Research Service || Fax: (303)490-8310 |
| Great Plains Systems Research || jon@gpsrv1.gpsr.colostate.edu |
| 301 S. Howes, P.O. Box E || FTS2000: a03jonhanson |
| Fort Collins, Colorado 80522 || |
+---------------------------------------++---------------------------------+

Yet, there is at least one group, a combination of several federal
agencies, that is attempting to build a user-directed system that
will build a new model according to the uers specifications. I agree
that this is a difficult task. I look forward to seeing how well this
group fairs.

+---------------------------------------++---------------------------------+
| Dr. Jon D. Hanson || Comm: (303)490-8323 |
| USDA, Agricultural Research Service || Fax: (303)490-8310 |
| Great Plains Systems Research || jon@gpsrv1.gpsr.colostate.edu |
| 301 S. Howes, P.O. Box E || FTS2000: a03jonhanson |
| Fort Collins, Colorado 80522 || |
+---------------------------------------++---------------------------------+

Dr. Anthony Starfield has done some work building wildlife management
models in close conjunction with the resource managers of big game parks.
The process sought to explicitly incorporate the experience and "rules
of thumb" into the rules of the model. Mangers were later asked to
observe and evaluate the model's performance and their own expectations
in a given situation. This approach is outlined in a book on modeling
for wildlife conservation which he did with Bleloch.

Jonathan Haskett
jhaskett@asrr.arsusda.gov

On Fri, 10 Dec 1993, JONATHAN HASKETT wrote:

> Dr. Anthony Starfield has done some work building wildlife management
> models in close conjunction with the resource managers of big game parks.
> The process sought to explicitly incorporate the experience and "rules
> of thumb" into the rules of the model. Mangers were later asked to
> observe and evaluate the model's performance and their own expectations
> in a given situation. This approach is outlined in a book on modeling
> for wildlife conservation which he did with Bleloch.
>
> Jonathan Haskett
> jhaskett@asrr.arsusda.gov
>

I am a graduate student at the university of Minnesota studying plant
breeding and conservation biology. I recently took Tony Starfield's class
on modeling. I think that the "rule-of thumb" approach is a very powerful
one. The types of models that he teaches students to build and use are
rather simple, but yet they can allow good insight into a problem. His
main point is that given a speadsheet and a few modeling concepts just
about anyone should be able to develop and use a simple model. Several
people in our Agronomy Department are interested in models that can
incorporate a broad range of data into a general easy to understand tool.
Dr. Starfield is currently working on frame-based models, and I think
that they may be a tool to allow us to do that. Is anyone out there
familiar with that concept and have you had any success?

Phil Schwab
Department of Agronomy and Plant Genetics
University of Minnesota
schwa015@maroon.tc.umn.edu

Thanks to Jon Hanson, John Bolte, Johnathan Haskett and Philip
Schwab for their thoughtful replies to my posting on model builders
and users. I will pursue the leads that you have given me.

Philip, could you explain what "frame-based models" are? I plan to
look up Starfield's book, but if you know of any particularly good
examples of these user rule of thumb models in wildlife biology or
agronomy, I would be interested in learning more about them.

After sending my posting, I thought of the modeling package called
STELLA, which an ecological modeler on this campus, Bruce Hannon,
uses in his research and in teaching an ecological modeling class for
people with little or no computer experience. It runs on MacIntosh,
and is visually oriented in that the programmer draws lines between
boxes to represent model information flow. I think there may be
a recently released IBM version of this model too. Has anyone out
there had experience with it?

Best Regards,
Greg McIsaac

> Date sent: Sun, 12 Dec 1993 10:33:50 -0600
> Send reply to: <agmodels-l@unl.edu>
> From: Philip M Schwab-1 <schwa015@maroon.tc.umn.edu>
> To: Multiple recipients of list <agmodels-l@unl.edu>
> Subject: RE: model builders and users

>
>
> On Fri, 10 Dec 1993, JONATHAN HASKETT wrote:
>
> > Dr. Anthony Starfield has done some work building wildlife management
> > models in close conjunction with the resource managers of big game parks.
> > The process sought to explicitly incorporate the experience and "rules
> > of thumb" into the rules of the model. Mangers were later asked to
> > observe and evaluate the model's performance and their own expectations
> > in a given situation. This approach is outlined in a book on modeling
> > for wildlife conservation which he did with Bleloch.
> >
> > Jonathan Haskett
> > jhaskett@asrr.arsusda.gov
> >
>
> I am a graduate student at the university of Minnesota studying plant
> breeding and conservation biology. I recently took Tony Starfield's class
> on modeling. I think that the "rule-of thumb" approach is a very powerful
> one. The types of models that he teaches students to build and use are
> rather simple, but yet they can allow good insight into a problem. His
> main point is that given a speadsheet and a few modeling concepts just
> about anyone should be able to develop and use a simple model. Several
> people in our Agronomy Department are interested in models that can
> incorporate a broad range of data into a general easy to understand tool.
> Dr. Starfield is currently working on frame-based models, and I think
> that they may be a tool to allow us to do that. Is anyone out there
> familiar with that concept and have you had any success?
>
> Phil Schwab
> Department of Agronomy and Plant Genetics
> University of Minnesota
> schwa015@maroon.tc.umn.edu
>
>

Received: by ccmail from gaia.ucs.orst.edu
>From root@crcnis1.unl.edu
X-Envelope-From: root@crcnis1.unl.edu
Received: from crcnis1.unl.edu by gaia.ucs.orst.edu with SMTP id AA17298
(5.67a/IDA-1.5); Sun, 12 Dec 1993 08:54:25 -0800
Received: by crcnis1.unl.edu id AA26369
(5.65c/IDA-1.4.4); Sun, 12 Dec 1993 10:52:59 -0600
Date: Sun, 12 Dec 1993 10:52:59 -0600
Message-Id: <MAILQUEUE-101.931212105719.320@age2.age.uiuc.edu>
Originator: agmodels-l@unl.edu
Errors-To: jp@unlinfo.unl.edu
Reply-To: <agmodels-l@unl.edu>
Sender: agmodels-l@unl.edu
Version: 5.5 -- Copyright (c) 1991/92, Anastasios Kotsikonas
From: "Greg McIsaac" <GFM@age2.age.uiuc.edu>
To: Multiple recipients of list <agmodels-l@unl.edu>
Subject: RE: model builders and users

Thanks to Jon Hanson, John Bolte, Johnathan Haskett and Philip
Schwab for their thoughtful replies to my posting on model builders
and users. I will pursue the leads that you have given me.

Philip, could you explain what "frame-based models" are? I plan to
look up Starfield's book, but if you know of any particularly good
examples of these user rule of thumb models in wildlife biology or
agronomy, I would be interested in learning more about them.

After sending my posting, I thought of the modeling package called
STELLA, which an ecological modeler on this campus, Bruce Hannon,
uses in his research and in teaching an ecological modeling class for
people with little or no computer experience. It runs on MacIntosh,
and is visually oriented in that the programmer draws lines between
boxes to represent model information flow. I think there may be
a recently released IBM version of this model too. Has anyone out
there had experience with it?

Best Regards,
Greg McIsaac

Greg -

We don't use STELLA for a number of reasons, some of which are:

1) it has limited representational capabilities. Good for
continuous models, but when you need to do anything more than
relatively simple representations it becomes unworkable pretty
quickly.

2) runs slow. for larger, computationally intensive models, it isn't
great. Particularly true for models used in optimization contexts
were the ability to "clone" models and do repeated optimization runs
is critical. STELLA has no support for this sort of thing.

3) Resulting models are non-portable. Last I heard, STELLA runs only
on Macs, and obviously uses a proprietary language. The have been
talking about a Windows-based version for years, but I have yet to see
it.

4) Partly because of 2) and 3), STELLA has limited capabilities for
communicating with the rest of the world. We prefer more "open"
solutions.

Still, for simple models, particularly for instructional purposes,
STELLA is pretty nice. Another (maybe nicer) alternative under
Windows is a package called VisSim which has some very nice features
as well. We have written a simpler package ourselves (EcoSim) that
we use for developing and demonstrating ecological models in a
graduate simulation class. It is written for Windows and has
capabilities for rapid model construction and graphical analysis.

For our "production" models, our approach has focused in many ways in
the opposite direction than STELLA. We have developed of over the
past several years a heavily object-based simulation framework
which is based on the "mix-and-match" concepts previously espoused.
All simulation logical and control is handled by a "simulation
environment" which is model independent, clonable, and handles
continous, discrete and knowledge-based simulation components as well
as a variety of inter-object communication protocols. All "actors" in
a given simulation subclass from a single, high-level "simulation
object" class which provides an abundance of services for collecting
data, setting timesteps, plotting, viewing, etc. Individual
simulation objects (and for that matter, complete simulations) can be
cloned as necessary. Works out really nicely for mixing
representation strategies (we mix a lot of expert-system-like "agents"
in our simulations to provide intelligent monitoring and control
services). All written in portable C++ with the exception of the
user-interface code, which is Microsoft Windows-based and necessarily
operating environment dependent.

I am encouraged to hear people begin to think along the lines you
mentioned in your original message. My own opinion is that we need a
major rethinking of how we do model building if we are going to have
any chance at all of applying these models to a broad range of
practical situations. We have been notoriously bad at this in the
past. Maybe there's hope in the future...

- John Bolte
Bioresource Engineering
Oregon State University
boltej@ccmail.orst.edu

Greg McIsaac writes:
>Has anyone out there in ag-model land attempted to incorporate
>users in the model building process?

The answer of course is yes (and no). Let me describe a couple of things
we are doing at the Agricultural Production Systems Research Unit here in
Toowoomba in southeast Queensland.

Participatory on-farm research
We are using our crop/cropping systems/erosion models in work with farmers
on commercial plots. We link in with farmers' current or planned
experimental program. So, farmers fix both the content and the design
of the on-farm experiments. Typically, these are fertilizer response
experiments at the moment, but we are also looking at crop sequencing
problems, opportunity cropping decisions etc. The farmers' experiments
are typically single large plots covering 2 or 3 rates. The challenge
for our researchers is to improve interpretation and generalisation of
farmers' results within these constraints. The researchers contribute
improved monitoring of the treatments. We attempt to model experimental
outcomes at a single point (site, year). If successful, we have some
basis for generalising the response to other sites and other years to get
a better understanding of the response surface. This process is embedded
in a cultural process aimed at improved communication between farmers and
researchers. Crop models allow us to work with farmers in this way, using
their rules for the conduct of experiments. The farmers experiments alone
would be dismissed by fellow researchers as anecdotal; their solution
would be to impose formal statistical designs on the farmers' experimental
program. We are trying to avoid doing this by using a combination of
monitoring plus models to get maximum benefit/insight/ownership/generality
from farmers' own experiments. Even though, in this case, models are still
the tools of the professional researchers and are built by them, the way
these tools are developing is beginning to be influenced by farmers.

Ethnographic decision tree models
In another, though related activity, we are attempting to understand the
rules farmers use to make decisions about the use of different technological
components. We expect this will provide some basis for the design of
decision support of some kind (although probably not Decision Support
Systems). The approach is a variant of Gladwin's "Ethnographic Decision
Tree Modelling". In this case, we are attempting to model what the farmer
does in her own language and to capture what she believes to be important.
Already, we are detecting significant discrepancies between farmers' rules
and those of professional research.

Both these are beginning studies. For those of you who want references see:

Cox, P.G., A.L. Garside and R.L. McCown (1993). Participatory design of an
agronomic on-farm research program in Central Queensland. Proceedings 7th
Australian Agronomy Conference, Adelaide, pp. 325-327.

Cox, P.G., A.L. Garside and R.L. McCown (1993). Market research for decision
support for dryland crop production in Central Queensland: a preliminary
report. Proceedings 7th Australian Agronomy Conference, Adelaide,
pp. 332-335.

Our general approach fits increasingly (although not yet very well) with the
ideas of Participatory Rural Appraisal. For a discussion of this, see the
article by Lamb, Designs on Life, New Scientist 1897 (30 October 1993)
pp. 37-40. We have some way to go before farmers' models and those of
professional research converge. Indeed, such convergence may not be
desirable since the two kinds of models are being used to do different
things. We think it is important to understand the two kinds of model,
and use the appropriate one. But this is probably the basis for a future
contribution to this discussion.

-------------------------------------------------------------
Peter Cox _--_|\
Agricultural Production Systems Research Unit / T\
PO Box 102, 203 Tor Street \_.--._/
Toowoomba o
Queensland 4350 Australia Way to go!

Ph: (076) 314298
Fax: +61 (076) 332678
Email: peterco@apsrusg.sth.dpi.qld.gov.au
============================================================

A few comments in response to Greg McIsaac's post:

>An alternative to the top-down, expert driven approach that I've
>caricatured above, might be to help users build their own models
>based on their own assumptions.

George Leavesley (george@snow.cr.usgs.gov) and coworkers are working
on the Modular Modeling System (MMS) with a GIS interface.
Basically, they've integrated compatible modules of hydrologic
algorithms so that the user can put together a model specific to
their needs.

>I wonder if a similar
>vision could be developed for agriculture. Rather than developing
>models, perhaps modeling tools and environments could be
>developed so that farmers, ranchers, land-managers, etc., could
>build and test models that they think best fit their assumptions and
>situations.

A goal of MSEA (Management Systems Evaluation Areas) has been to
regionalize information on ag management impact on the environment and
to feed that infomration into a spatial decision support system
(SDSS). I've heard that they may have lost some funding, so I have
no idea how much progress they'll make towards this end.

There's quite a bit of literature on SDSS (especially in GIS circles).

Maribeth Milner
milnerm@phibred.com

Our research has interested in both the MMS and MSEA projects mentioned
in a previous message. MMS is no proceeding as a joint project with
USGS, ARS, and Forest Service. It is progressing nicely. The objective
of this program is to develop a computer program that generates
simulation models for a specific user and according to that user's
specifications. I am most interested in seeing how well this approach
works and if the resulting models are valid.

Second, the MSEA project was originally scheduled for two years of
funding. We are currently in the second year and seeking funding
for a third year. I am very hopeful that the objectives of
regionalizing simulation results can be obtained. My opinion is
that funding has been adequate. The chances of receiving a third year
funding are very good.

+---------------------------------------++---------------------------------+
| Dr. Jon D. Hanson || Comm: (303)490-8323 |
| USDA, Agricultural Research Service || Fax: (303)490-8310 |
| Great Plains Systems Research || jon@gpsrv1.gpsr.colostate.edu |
| 301 S. Howes, P.O. Box E || FTS2000: a03jonhanson |
| Fort Collins, Colorado 80522 || |
+---------------------------------------++---------------------------------+

24TH ANNUAL WORKSHOP ON CROP SIMULATION

FIRST CALL FOR PAPERS

The 24th Annual Workshop on Crop Simulation will be held on
March 15-18, 1994 in Raleigh, North Carolina. This workshop is
organized by the Biological Systems Simulation Group and co-
sponsored by North Carolina State University (Gail Wilkerson,
see address below), the USDA Crops Research Lab (Sally Schneider,
phone 919-693-5151 ext 256, e-mail SSCHNEIDER@asrr.arsusda.gov),
and Division A3 (Agroclimatology and Agronomic Modeling) of the
American Society of Agronomy.

The workshop will consist of invited and volunteered papers as
well as a software demonstration and poster section. An open
discussion (free-for-all) is scheduled for one evening. Invited
papers will address the role of modeling in sustainable agriculture
from the viewpoint of a plant modeler, a plant pest modeler, a
soils/water/chemical fate and transport modeler, the Soil
Conservation Service, an economist, and a farmer.

Volunteered papers, posters, and software demonstrations are
requested on topics dealing with the development of simulation
models or components as well as their applications in research,
agricultural education, systems analysis of cropping options and
environmental quality studies. Papers on a broad range of topics
related to the simulation of biological systems are welcome.
Please submit a one-page abstract by February 16, 1994 to:

Gail G. Wilkerson
Crop Science Department
North Carolina State University
Raleigh, NC 27695-7620
Phone: (919) 515-5816 Fax: (919) 515-7959
E-mail: ggwilker@csemail.cropsci.ncsu.edu

If possible, submit your abstract on floppy disk (WordPerfect,
Word, or ASCII format), as well as in a hard copy. We will attempt
to standardize appearance of the abstracts included in the
Proceedings. Abstracts should fit on one page, single-spaced, with
one-inch margins, and type no smaller than 10 point. Include
authors' affiliations and addresses at the top of the abstract.
Please state in a cover letter whether you wish to present the
paper orally (15 minute limit), in poster form, or as a software
demonstration. We may be unable to honor all format requests or
schedule all papers that are submitted. All abstracts which are
submitted by mail by February 16 will be included in the
Proceedings of the Workshop.

The Workshop will be held at the Brownestone Hotel, 1707
Hillsborough Street, Raleigh, North Carolina (NC 1-800-331-7919,
Outside NC 1-800-237-0772, or 919-828-0811). A block of rooms has
been reserved in the name of "Crop Simulation Workshop" at the
special rate of $48.00 single/$54.00 double + 12% tax. You should
make your hotel reservations by February 23, 1994 to obtain the
Workshop rate. The hotel provides a free shuttle service to and
from the Raleigh-Durham International Airport.

A registration fee of approximately $50.00 will cover the cost
of the catered barbeque dinner, welcome reception, meeting room
fees, proceedings, poster rentals, and other ancillary expenses.

PLEASE POST AND/OR CIRCULATE AMONG INTERESTED COLLEAGUES

Global Change and Terrestrial Ecosystems:
The First GCTE Science Conference

Monday, 23 May - Friday, 27 May 1994
Woods Hole, Massachusetts, USA

This conference will present a state-of-the-science assessment of
research into terrestrial ecosystem interactions with global
change. It will present and review new results in the areas of
global change impacts on terrestrial ecosystems, and the feedback
of change in terrestrial ecosystems to the atmosphere and to the
physical climate change. The meeting will include the early
achievements of the Global Change and Terrestrial Ecosystems (GCTE)
Core Project of the International Geosphere-Biosphere Programme
(IGBP). Conference themes include:

Ecosystem Physiology
Ecosystem Structure and Composition
Agriculture, Forestry and Soils
Ecological Complexity

In addition to these four major themes, which form the four Foci of
GCTE, the conference will address a number of cross-cutting themes
which involve more than one Focus and which involve GCTE in
collaborative work with other IGBP Core Project/Framework
Activities and with other groups.

The conference will feature a mixture of overview talks on major
themes (30 minutes each), invited papers (30 minutes each), and
contributed papers (20 minutes each). In addition to this oral
presentations, contributed papers will also be presented in special
poster sessions. There will also be working group sessions and a
special plenary session at the end of the conference to evaluate
progress within GCTE and suggest changes to the GCTE research
programme and its implementation.

Conference Organizing Committee

Dr. Brian Walker (Chairman), CSIRO, Australia
Prof. Hal Mooney, Stanford University, USA
Prof. Hank Shugart, University of Virginia, USA
Dr. Bernard Tinker, University of Oxford, UK
Prof. Osvaldo Sala, University of Buenos Aires, Argentina
Dr. Jerry Melillo, Marine Biological Laboratory, USA
Dr. Will Steffen, GCTE Core Project Office, Australia

Paper Submission Form

To submit a paper for presentation at this conference, please
return this form (or a copy), along with a 1/2 to 1 page abstract,
to:
Dr. Brian Walker
Chairman, GCTE
GCTE Core Project Office
CSIRO Division of Wildlife & Ecology
PO Box 84
Lyneham ACT 2602
AUSTRALIA

Tel: 61-6-242-1742; Fax: 61-6-241-2362
E-mail: BHW@CBR.DWE.CSIRO.AU

Abstracts must be received before 15 JANUARY 1994 to be considered
for presentation at the conference. Authors will be notified
acceptance by 28 February 1994

Author(s):__________________________________________________
______________________________________________________________
______________________________________________________________
______________________________________________________________

Contact Name:_____________________________________________________

Address:__________________________________________________________
__________________________________________________________________
_________________________________________________________________
_______________________________________________________________
_______________________________________________________________

Tel:_______________________Fax:________________________________
E-mail:_________________________________________________________

Attached is an abstract of a paper I/we would like to present at
the conference. I/we would prefer the paper presentation to be
(please tick one)

o Oral o Poster o No preference

Paper
Title:________________________________________________________
_________________________________________________________________
_________________________________________________________________

THIS IS A SAMPLE ABSTRACT:
THE TITLE SHOULD BE IN BOLD, UPPER CASE LETTERS

A.B.Author, C.D.Colleague, and E.D.Collaborator

Name and Location of Institution(s)

A book of abstracts for all invited contributed and poster papers
to the First GCTE Science Conference will be assembled and copies
made available to all registered attendees at the beginning of the
meeting. To facilitate preparation of this book, the organizing
committee would appreciate the cooperation of all those preparing
papers in adhering to the format presented here.

The entire abstract including title, author(s) and text should be
confined to one A$ page at most and contained within a imaginary
box of 16.5 cm wide by 23 cm high (6.5" by 9") centered on the
page. Please use a Palentino 12 font in printing the abstract is
possible. Otherwise, use the closest type font available.

The title of the paper should be centered at the top and printed in
bold, UPPER CASE letters. Leave 2 blank lines and then enter the
author(s) names. Author affiliations should be given (in brackets
and lower case) on the line following the name. Use superscript
numbers to identify authors with particular institutions. Leave 2
blank lines and then begin the body of the abstract. In typing the
body of the abstract, please ensure that it is single-spaced, not
right-justified and that the start of each paragraph is not
intented. Each paragraph should be separated by 1 blank line.

Tentative schedule

Monday 23 May
1400-1900 Registration
1900-2030 Dinner
2030- Miser

Tuesday 24 May

0830-0900 Welcome and GCTE Overview
0900-1045 Session I (Ecosystem physiology)
1045-1100 Break
1100-1215 Session I (continued)
1215-1430 Lunch/Poster Session
1430-1545 Session II (Ecosystem structure/composition)
1545-1600 Break
1600-1730 Session II (continued)

Wednesday 25 May

0830-1030 Session III (Cross-cutting theme)
1030-1045 Break
1045-1245 Session IV (Cross-cutting theme)
1245-1430 Lunch/Poster Session
1430-1800 Parallel working group sessions

Thursday 26 May

0830-1030 Session V (Agriculture, Forestry, Soils)
1030-1045 Break
1045-1230 Session V (continued)
1230-1430 Lunch/Poster Session
1430-1545 Session (Ecological Complexity)
1545-1600 Break
1600-1700 Session VI (continued)
1700-1800 Plenary Discussion (Evaluation of GCTE)
1900-2130 Conference Dinner

Friday 27 May

0830-1045 Session VII (Cross-cutting theme)
1045-1100 Break
1100-1200 Concluding Session
1200-1300 Lunch
1330 Conference adjourns

Registration and Accommodation

The registration fee is $US 50. All registered participants will
receive a copy of abstracts of all papers to be presented at the
conference and a copy of the conference proceedings.

Accommodation/meals are available at the Marine Biological
Laboratory (MBL) or in local motels. Costs are ($US):

Shared twin room MBL (includes meals) 325
Single room MBL (includes meals) 390
Motel accommodation (no meals 300-450
(approximately)
Motel accommodation + MBL meals 550-700
(approximately)
MBL meals only 250

Refunds for cancellations will not be issued after 1 May 1994.

Note: Some limited support will be available for participants from
developing countries and from Eastern Europe.

General Information

Location: Woods Hole is located on the southern side of Cape Cod
peninsula, Massachusetts, USA.

Climate and Clothing: The weather in late May is cool (13-21 C, 55-
70 F). Causal atmosphere and dress.

Places of Interest: MBL/WHO1 Library, Woods Hole Oceanographic
Institution, National Marine Fisheries Aquarium, U.S. Coast Guard
Station, Martha's Vineyard.

Local Transport: Woods Hole is about two hours south of Logan
International Airport, Boston, Information on local transport will
be provided to registered participants in the third circular, April
1994.

Conference Registration Form

Intending participants are asked to register by April 15 1994.
Also, please note the cancellation policy for accommodation on the
preceding page. Note that MBL housing is available on a first-
come/first-serve basis (can not be guaranteed after 15 April 1994).

Registration - by April 15, 1994 to:

Lou Ann King, Manager of Conferences
Marine Biological Laboratory
Woods Hole, MA 02543
USA

Name:____________________________________________________________
Male/Female (please circle) (Required to assign roommates)

Postal Address:____________________________________________________
_________________________________________________________________
_________________________________________________________________

Tel:___________________________Fax:_______________________________

E-mail:_________________________________________________________

I will require accommodation/meals in Woods Hole for the following
dates:
ARR:_________________________DEP:___________________________

I will share a double room at MBL (includes all meals) with
_________________________________________________________
(if no name is specified, a roommate will be assigned).

I enclose herewith a cheque in the amount of $US
(room/meals/registration)_______________made payable to MARINE
BIOLOGICAL LABORATORY

Signature:______________________Date:_______________________

Send to: Lou Ann King, Manager of Conferences
Marine Biological Laboratory
Woods Hole, MA 02543, USA
Tel: 1-508-548-3705; ext.212;
Fax: 1-508-540-7187
E-mail: L.KING@HOH.MBL.EDU 

Another livestock modeler heard from.

I am an Agricultural Economist with an interest in modeling range livestock
management issues. At present I have ongoing projects in range cattle
nutrition management, optimal culling straegies, and optimal preg testing
strategies. The nutrition work has been done in Stella while the culling
and preg testing models are dynamic programming models sumarized by
statistical development of decision trees. Papers describing these models
are available over the internet by anonymous ftp to ag.srnr.arizona.edu in
the pub/irm directory. Postscript and Adobe Acrobat .pdf files are
available. We have also set up a discussion list for integrated resource
management for livestock operations. Anyone interested in subscribing can
send me e-mail at rgum@ag.srnr.arizona.edu and I will add you to the list.

While not completly ag related a team in which I participated has just
completed a Stella model of the water management options for dealing with a
severe drought in the Colorado River basin. Included in the modeling
effort are economic, institutional, and hydrologic systems.

Russell Gum 602-621-6265
rgum@ag.srnr.Arizona.Edu
Department of Agricultural and Resource Economics
University of Arizona
Tucson AZ 85721

In response to your request to introduce myself, I am Don Wauchope, Research Ch
emist with the USDA Ag. Res. Service. I do research on the behavior and fate o
f pesticides in the environment. I was involved a little in the development of
the CREAMS model and am currently on the periphery of the testing and tuning of
the RZWQM model. I can best be described as a field and laboratory researcher
in the chemodynamics of pesticides, but it is clear to me that models are the
only way to integrate and extrapolate such measurements.

Participation in such an electronic discussion is a new experience for me and I
ask that you all be patient while I learn how.

I'm Bill Bland, Asst. Prof. and Agricultural Climatologist in the
Department of Soil Science, Univ. Wisconsin-Madison. I operate the
Wisconsin Agricultural Weather Observation Network, now 19 stations;
recent research includes inexpensive and rugged anemometers,
evaporation from cranberry and potato, and spatial structure of solar
radiation.
I am about to start a project involving biophysical
modeling of grazing dairy cattle. Anyone out there doing this, or
know of someone who is? Are there Stella-like software systems for
PC Windows?

>Are there Stella-like software systems for PC Windows?
I haven't seen anything on the PC end of things which approximates
STELLA, although a number of dedicated modelling packages exist. In
terms of ease of use and user interface STELLA remains the best. If
I remember correctly, the folks at High Performance systems, who
make STELLA are working on a Windows version so that "Stella-like"
won't be necessary.

Jonathan Haskett

unsuscribe

I am working at IRRI, Philippines, as theme coordinator in the SARP (Systems
Analysis for Rice Production) network, coordinating modelling activities in 15
National Agricultural Research Centres in various south-east Asian countries.

My interests are in climate change effects on regional rice production,
genotype x envt interactions, tiller formation, and sink/size relationships.
Previously, I worked on developing the IBSNAT cassava model with Tony Hunt,
Guelph, Canada.

I look forward to interactions with other modelers with similar interests.

ROBIN MATTHEWS
International Rice Research Institute, P O Box 933, 1099 Manila, Philippines.
email: R.MATTHEWS@CGNET.COM

Sorry abou this test message. I have been having problems posting to
the group.

Graham Lewis

Bill Bland,
I am interested in developing a modified-penman or other ET module
for potato that can use wind and dew point or RH as well as simulated
leaf-area-index (LAI). Would your data be suitable for that?
Tom
Tom Hodges Cropping Systems Modeler ___ ___
USDA-ARS / \_/ \
Rt. 2, Box 2953-A Telephone: 509-786-2226 | |
Prosser, WA 99350 \______/^\/
USA potato tuber
thodges@beta.tricity.wsu.edu thodges@asrr.arsusda.gov
=====================================================================
...photosynthesis makes the world go around... Mr. Potato Head

On Thu, 16 Dec 1993, William L. Bland wrote:

> I'm Bill Bland, Asst. Prof. and Agricultural Climatologist in the
> Department of Soil Science, Univ. Wisconsin-Madison. I operate the
> Wisconsin Agricultural Weather Observation Network, now 19 stations;
> recent research includes inexpensive and rugged anemometers,
> evaporation from cranberry and potato, and spatial structure of solar
> radiation.
> I am about to start a project involving biophysical
> modeling of grazing dairy cattle. Anyone out there doing this, or
> know of someone who is? Are there Stella-like software systems for
> PC Windows?
>

Hi!
This is just test.
Regards, Feliks Landa

Hi,
I'm Adrienne Roelofs. I'm working on the ORYZA rice production model
and hopefully soon on a constructed wetlands model which will be
my thesis. I am a student of John Bolte. My background is in botany
and genetics and using those two I have found areas of interest in
genetic algorithms and expert systems. If there are any wetlands
modelers out there, please jump in (the discussion not a wetlands).
Thanks, Adrienne
============
Adrienne Roelofs
Bioresource Engineering
Oregon State University, Corvallis, OR 97331-3906
Telephone: 503-737-3218 FAX: 737-2082

Hi everybody !

I just heard of agmodels-l and immediately subscribed, with enthusiasm.
Many thanks to those who initiated it.

My name is Marc TCHAMITCHIAN and I am working for the French INRA, National
Institute for Agricultural Research. My lab is "Bioclimatology", located in
Avignon (south east of France, near Marseille).
Bioclimatology is a department of INRA dealing with life-environment inter-
actions, life mainly meaning plants !

In Avignon, we are dealing with three main topics:
- Remote sensing (sensors, signal processing, extracting biological
information)
- Agrometeorology (crop models, scale change like in coupling remote
sensing data with crop models, evaluation of regional potentialities
with these crop models)
- Protected Cultivation (Greenhouse physics understanding and modelization,
greenhouse crop ecophysiology, greenhouse environmental control and
optimization)

I am working for the Protected Cultivation Group (Greenhouse Environment
Modeling and Management) in the area of climate control using mathematical
theories like optimal control. My interest in agricultural models is twofold:
- I worked on a greenhouse tomato crop photosynthesis model during my PhD
and still work on it a little
- to use optimal control, I desperately seek models of the greenhouse and
the greenhouse crop that are dynamic, simple, validated, (shall I
dreamingly perfect ?).

One key issue in using models for control lays in the validity domain. This
point may seem trivial but is not.
Most models have limits (no one will ever think of calculating a photo-
synthesis for a negative radiation value or a negative CO2 value !). But most
models do not include these limits in their formulation. Limits are just
common knowledge to scientist and modellers, written down in the model
presentation paper.
When using a control optimization routine, limits MUST be stated either as
hard bounds, or as bounding functions. Hard bounds are not easy to deal
with (constrained optimization is always more difficult and tricky than
unconstrained optimization). The problem with bounding functions is that
I would prefer them to tell something representative of the process to
simulate and bound. But little work is available on the fringes of available
models.

To you all

Marc
------------------------------------------------------------------------
Tchamitchian Marc Tel: (+33) 90 31 60 83
Bioclimatology Secr: (+33) 90 31 60 94
INRA BP 91 Fax: (+33) 90 89 98 10
F-84143 Montfavet Cx, France E-mail: arm@avignon.inra.fr
------------------------------------------------------------------------

I've been off internet for some days due to renovation activities within
the building. Maybe my reply comes a bit late. Anyway, I'd like to answer
one of the questions of Bill Bland concerning a PC WINDOWS STELLA-like
program.

Personally, I do not have any experience with STELLA or another such
program, but I know that my colleagues at the University of Amsterdam use
STELLA and a PC tool called VENSIM.

It was developed by VENTANA Systems, Inc.

Information can be obtained from the following address:

VENSIM Product Center
Ventana Systems, Inc.
149 Waverley Street
Belmont MA 02178
USA

Phone: 1 617 489 5249
Fax: 1 617 489 5316

Regards,

Mark Snethlage
m.snethlage@frw.ruu.nl

I am a post-doc in the Agricultural Engineering Department at the Ohio
State University.
I am working as part of the Management Systems Evaluation Area (MSEA) modeling
team.
My interests are in Crop and Environmental Physics modeling,
Decision Support System design.
I look forward to interacting with other modelers.

Sincerely, Feliks Landa
+--------------------------------------+--------------------------------------+
|Feliks Landa | Phone: (614)292-8893 |
|The Ohio State University, | |
|Department of Agricultural Engineering| Fax: (614)292-9448 |
|590 Woody Hayes Drive | |
|Columbus,OH 43210-1057 |Email:flanda@magnus.acs.ohio-state.edu|
+--------------------------------------+--------------------------------------+

I am a veterinarian with the Centers for Epidemiology and Animal Health in
Ft. Collins CO. We are part of the USDA Animal and Plant Health
Inspeciton Service under Veterinary Services. I am not currently doing
any modeling but in the past have worked on have worked on a series of
linear programming models to evaluate economics of animal health and
health management programs at the farm/ranch level. In the past our group
has been involved with simulation models of respiratory disease in dairy
calves as well as Johnes disease on dairy operations. Another portion of
our center is working with a simulation model to evaluate control
stategies for pseudorabies in pigs.

I'm looking forward to discussions of modeling techniques in general and
specifically projects dealing with the modeling of animal and/or plant
health.

Dave Dargatz
ddargatz@aphis.ag.gov
(303) 490-7800 voice
(303) 490-7899 fax
USDA:APHIS:VS
Centers for Epidemiology and Animal Health
------------------------------ Cut here ------------------------------