Dina  •  Annual report  2000
 

Introduction
The year 2000 was the first one in a new working period with a reduced budget  and focus on the following core activities:
 

The fifth Dina priority research area, Information Technology and Agricultural Engineering, has mainly developed outside the Dina framework, and it is undecided in which way Dina should support this area in the future.

As a rule, actual project activities take place outside the field of Dina's own budget and responsibility. However, Dina still shares the responsibility for a few earlier projects with external funding; besides there is some activity - involving, for instance, the IT coordinators - to coordinate applications for the funding of new projects.

Below follows the detailed report, ending with the plans for 2001.
 

The board
Two board meetings took place: on January 21, 2000 at KVL, Copenhagen, and on Dec. 12 2000 at DTU, Lyngby. - Important items on the agenda of the January meeting included the completion of the status report “Dina 1991-1999”, and the general plans and budget for the period 2000-2002. A new scheme of funding from the member institutions, comprising a slightly larger contribution from KVL, was adopted; so were the proposed working plan and the budget for 2000. - At the December meeting, the board approved a provisory issue of the present report, that is, activities in 2000 and the plans for 2001; general plans for the Annual Meeting 2001 were discussed and laid out; and the budget for 2001 was adopted.
 

The management
The management still consists of  Mogens Flensted-Jensen (head) and Iver Thysen (deputy head); Academic secretary: Poul Einer Hansen; Secretary: Jeanette Kublitz (until June 30), Carina Jensen (from July 1); Webmaster: Sven Esbjerg (until Sept. 30), Thomas Iversen (from Oct. 1). Plans of moving the webmaster to Foulum are under consideration.

The beginning of 2000 was affected by the transition to a new working period, following the decision of the member institutions on the continuation of Dina, provisionally until the end of 2002. The status report Dina 1991-1999 was completed and came out in March; a thousand copies were printed of which about 900 have been distributed; the report was received favourably. In May, the first issue of a semiannual newsletter, DinaNyt, appeared (circulation: 600); following a decision at the board meeting in December, the newsletter will in future be edited in English as DinaNews; a second issue came out in January 2001.

The Dina website is updated and developed continuously, according to available resources. In connection with the report Dina 1991-1999 (which is also accessible on the Internet), two essential improvements were made: The directory of “Dina people” was updated and made more user-friendly; and the publications list was rewritten in a fairly complete form.
 

The Network Committee
The Dina Network Committee consists of representatives from Dina member institutions and of the management. The Network Committee meets 2-4 times per year; the IT coordinators and the research school experts are invited to the meetings. In 2000 the Network Committee met twice: on January 11 via VideoLink, to prepare the board meeting ten days later; and on August 29 at Tune Landboskole, to plan the Dina Annual Meeting for 2001.

The Network Committee is a co-ordinating body: it assists the Dina management in preparing material, data and recommendations for the board; it carries out the general planning of annual meetings and workshops; and it coordinates large project applications, typically by appointing an ad hoc committee.

The Network Committee is of major significance to Dina’s work. Without formalities, the committee maintains the solidarity and the cooperation that was established in the period where Dina was developed on the basis of shared project grants.
 

The Annual Meeting
For the first time, Dina organised its annual meeting in cooperation with DSIJ, Danish Society for Informatics in Agriculture. DSIJ has held its own annual meeting three times, while Dina has organised a thematic day or an annual meeting over a number of years.  The scope of the cooperation was to create an all-embracing event which could appeal to virtually everybody with an interest in informatics in agriculture, be it in research, development or extension service - and thus an event which might attract a fairly large number of participants.

The annual meeting took place on March 16-17 at the Fuglsø Center on Mors. The meeting lasted two half days, each one covering an exciting theme of current interest. The first day was about Portals on the Internet, with three subheadings: “On the portal concept: how to cooperate and to run a portal”, “Trade oriented portals: some examples”, and “We build the portal of agriculture”.

The theme of the second day was Multimedia. The programme fell into two sessions: “State of the Art”, illustrating how far 3D visualisation has come; and “Examples of practical applications of multimedia”, one example being construction advising in agriculture.

Of the registered 58 participants, a few contented themselves with just one of the themes, but the majority took part in both, and they had the opportunity, in the evening of the first day, to make contact with other participants - an obvious success, as it is every year.
 

The Dina Research School
Head:  Anders Ringgaard Kristensen.  Experts attached to the school: Per Christian Hansen, Peter Sestoft and Henrik Stryhn.

During 2000, the research school organised the following events:

Workshop on Biosequence analysis, Koldkærgård Landboskole, April 6-7. Biosequence analysis means data processing and statistical analysis of data for DNA/RNA and protein sequences. Heredity mapping has been carried out on numerous organisms in recent years, and rapidly too. However, some questions arise: What use can be made of the data that has been collected? How does one proceed in practice? How do the computer programs that lie behind the mapping work? The workshop aimed at answering these questions through sessions on
- searching the existing sequence databases,
- building phylogenetic trees (genealogical trees) for organisms else unknown,
- organising alignment algorithms for DNA/RNA and for protein sequences in the computer,
- identifying the parts of the genome that decide phenotypical traits such as the gene map, linking analysis, and quantitative trait loci,
- overviewing the future perspectives for biological research in the ever-growing data sets, and in the development of new methods in data processing and statistical analysis.

The workshop had 41 participants which is the largest number to date. Two were Nordic PhD students, two were IT coordinators from the Dina priority area Bioinformatics and Quantitative Genetics, and four were contact persons from the Nordic network coordinated by the research school. The professional content of the workshop was organised by Peter Sestoft and Henrik Stryhn. Further information, including the programme (in Danish), can be found at the web address   http://www.dina.dk/phd/w/w6/w6.htm.

Nordic summer school: Differential Equations and Dynamic Systems in Agriculture, Tune Landboskole, August 20-31.  The scope was to give a general introduction to mathematical modelling of dynamical systems described in terms of ordinary differential equations, with focus on applications in agriculture. Computer based methods to solve differential equations were illustrated, using either specialised software or suitable modules of more general software. Following the summer school, participants were able to formulate simple dynamical models and to assess the accuracy and the quality of a numerical solution.

The main themes at the summer school:
- mathematical modelling of dynamical systems by ordinary differential equations,
- implementing a model on the computer,
- using the resulting system for simulation and parameter identification,
- evaluating the result with respect to accuracy and quality.

The summer school had 24 participants, among these 11 Danish PhD students or other young researchers. Per Christian Hansen, in cooperation with docent Per Grove Thomsen, DTU, was responsible for the professional content. The international guest lecturer was Prof. Ken Jackson assisted by Petter Wiberg, both from University of Toronto. Further information, including the program and evaluation by participants, can be found at the web address  http://www.dina.dk/phd/s/s3/s3.htm.

Workshop on Digital image processing, Tune Landboskole, Dec. 7-8. Digital image processing is an interdisciplinary activity aimed at the automatic extraction of information from images. Digital image analysis is in particular applied with the purpose of automatising routine visual inspection or classification; this is relevant not only in industry but also in research, where image processing allows for the analysis of large data sets in image form. In other situations, image processing can be applied to objective measurements of objects that were hitherto only judged visually and were classified in a more or less subjective way. Examples of applications: counting and measuring cells in microscope pictures; automatic visual sorting of seeds or plants; mapping of area application through satellite photos; measuring the volume of pulp-wood loaded on a truck; navigation systems for robot drivers. The workshop gave an introduction to image processing, with focus on general image-oriented techniques and on an algorithmic approach based on data processing and technology. Still, some lectures used other approaches: one based on statistical modelling, another - traditionally applied in remote sensing - based on "local points".

The workshop had 25 participants, including 5 PhD students from Sweden. IT coordinator Morten Larsen was responsible for the professional content in cooperation with the research school experts. Further information, including the program, can be found at the web address  http://www.dina.dk/phd/w/w7/w7.htm.
 

Priority research areas

Spatial Statistics in Agriculture
IT Coordinators:  Rasmus Waagepetersen and  Morten Larsen.
The 2000 contribution focused on the workshop “Spatial Statistics and Image Analysis in Precision Agriculture” which took place on Nov. 16 at Aalborg University, with 21 participants from AAU, DJF, DMU (The National Environmental Research Institute), KVL, Risø and DAAC.

Data on soil properties, weed spread, yields, and crop status are a prerequisite for precision agriculture; the 9 lectures at the workshop dealt with various techniques for collecting data as well as with their statistical treatment. Image data was the focal point in several lectures; both remote sensing (air/satellite images) and photos from tractor-mounted cameras were treated. In addition, presentations were given regarding 1) yield measurements in the form of yield per meter data and 2) geoelectric measurements of soil properties.

The workshop was most successful, with an animated discussion. For more information on the website, including the program, list of participants and abstracts, consult the website of the priority area: http://www.dina.kvl.dk/~ml/dina/itcoord/  under ‘Previous events’.
 

Information Technology and Agricultural Engineering
IT Coordinator:  Anders P. Ravn.
Within the field of agricultural techniques 2000 was characterized by much activity, although most of it took place outside Dina. At Aalborg University, the new Center for Agricultural Technique has started up and has, among other things, announced a specialization in the civil engineering study programme, see http://cjt.auc.dk. Simultaneously, KVL is in the process of renewing the agricultural technique area by forming a Center for Agrotechnology where a study programme is planned in cooperation with DTU. Finally, DJF has also strengthened the input by appointing a new deputy head, Svend Christensen, who will coordinate the area. The parties mentioned above coordinate education and other activities through the Dina inspired network DANET (Danish Agricultural Network in Engineering and Technology), see http://www.agrsci.dk/jbt/danetny.
 

Internet-based Advisory Systems in Agriculture
IT Coordinators: Jens Peter Hansen and Ellen Juel Christensen.
In connection with the Dina Annual Meeting on March 16-17 (see above), a workshop was organised that went into more detail on the two themes of the meeting, Portals on the Internet and Multimedia.
 

Model-based Decision Support Systems in Agriculture
IT Coordinators: Erik Jørgensen and Allan Leck Jensen.
A workshop on "Stochastic modelling in plant production - presenting the problems, and methods to solve them" took place on Oct. 31. The workshop focused on practical experience with the development of stochastic models in plant production. Two general presentations on stochastic modelling and four on specific applications were given. Furthermore, Prof. Stein W. Wallace, Trondheim gave a one-hour lecture on decision making in uncertainty.

The workshop demonstrated that most applications have the following problems in common: laying down the model structure; degree of complexity; discretization of continuous variables; representation of time; providing data for parameterization. It was briefly discussed whether typical problems in plant productions would allow for representation in Bayesian networks of influence diagrams, since they are often characterised by dynamics, interaction and unevenly distributed continuous variables.

The workshop was attended by 31 people, most of them from DJF, LR, KVL and AAU. Several people have expressed their satisfaction with the arrangement. Most of the presentations, including that of Prof. Wallace,  can be found at the webpage of the workshop, www.planteinfo.dk/alj/dina/program_WS2000.html.

Among the other activities in 2000, two can be mentioned: coordinating project applications; and intensifying the cooperation between AU, KVL and the Research Center Foulum on developing new methods in decision support. This cooperation can be seen as combining the expertise on hierarchic multi-level Markov processes; on influence diagrams with limited recall; and on data processing. Both activities are expected to continue in 2001.
 

Bioinformatics and Quantitative Genetics
IT Coordinators: Mogens Sandø Lund and Gunter Backes.
A workshop with the title “Gene mapping and marker assisted selection - Methods and analyses for plants and domestic animals” took place on Oct. 6. The organisers found that the day went very well, not least because the 30 participants came from so many different places: from the universities (Copenhagen, Aarhus, Aalborg, South Denmark), from KVL, Risø, The National Bureau for Cattle, NLH (Norway) and DJF. The program was versatile, with applied as well as theoretical contributions on mapping of single genes for properties in cattle, plants and drosophila; furthermore there were contributions on 1) statistical methods of multivariate analyses, 2) taking incomplete marker information in complex genealogies into account, 3) taking into account that the number of loci may be unknown.
 

Projects and activity reports

This section gives an account of what has happened to a number of specific projects and Dina members during the year. The order of the reports is random and we have not attempted to unify their form, nor their size.

Application of weather prognoses in agriculture
Project worker: Nina Detlefsen, Dina Foulum.
In cooperation with the Danish Meteorological Institute, the project aims at investigating
   1) the predictive value of decision support based on weather prognoses,
   2) possibilities for time-varying uncertainty in weather prognoses.
Concerning the first objective, comparisons were made between decision support derived from prognoses and decision support derived from subsequent observations. With regard to the second objective, the variation in ensemble prognoses (50 simultaneous prognoses calculated with varying initial data) is compared with the accuracy of the very prognosis. Publications: Contributions given at 1) Conference of Danish Center for Plant Protection, March 2000, and 2) Dina Workshop, Oct. 2000.

Combination of remote sensing data for invention and health surveillance of a forest
Project workers: Mats Rudemo (contractually liable) and Morten Larsen.
This three-year project is funded by the Danish research councils within the program "ESA Consequence Research" and is carried out in cooperation with FSL. The purpose is to develop tools of analysis for combining panchromatic high solution air photographs and multispectral, less highly soluted satellite photographs, aiming at invention and monitoring of a mixed forest. Two FSL sample plots in Jutland have been selected for the project. In the  autumn of 2000 an attempt was made to achieve both colour-infrared air photographs and multispectral data from the Ikonos satellite, to allow for working with real as well as with simulated satellite images.

Activity report from Jesper Møller
Ole Fredslund Christensen's PhD project on geostatistics and Markov chain Monte Carlo methods is progressing well. He has written two articles together with Rasmus Waagepetersen and Jesper Møller: one deals with the analysis of spatial data by generalised linear mixed models and Langevin type Markov chain Monte Carlo, where models and methods are, e.i., applied in weed censuses; the other article investigates convergence properties of the algorithms, in particular geometrical ergodicity. In the spring semester of 2000, O.F.C. spent time at Lancaster University; his cooperation with Peter Diggle and Paulo Riberio on model based geostatistics has resulted in an article on the analysis of spatial data using the transformed Gaussian model, where the methods were used to analyse rainy weather. The scientific findings of O.F.C. have been presented at several national and international scientific meetings: at a conference in Ambleside, Sept. 2000, his and Paulo Riberio's poster was nominated best poster. Ole Fredslund Christensen's PhD project is expected to be concluded in the autumn of 2001.

Frede Aakmann Tøgersen's PhD project, on spatial statistics in the agricultural sciences, is also proceeding well. Apart from his daily routine at Foulum, F.Aa.T. visited the University of Washington in the spring of 2000, to consult Peter Guttorp, among others. During his stay he also, together with Martin Bøgsted Hansen and Jesper Møller, completed an article on a Bayesian analysis of deformable template models for time-space processes. In addition, F.Aa.T. and Rasmus Waagepetersen have started to write a paper on spatial deconvolution of yield metre data, with application in precision agriculture. Besides this, F.Aa.T. intends to work with spatial process models (kriging), where the covariate information has not been measured in the same places as the response variable. This too aims at an application in precision agriculture. The scientific results of F.Aa.T. have been presented at national and international scientific meetings. Frede Aakman Tøgersen too expects to complete his PhD project in the autumn of  2001.

Activity report from Søren Hansen
Continued development of the Daisy model.  Through 2000, the Daisy code has been provided with additional "default values" so that it is no longer necessary to specify a number of model parameters and initialisation variables; this has led to a more user friendly model. The model has also been extended with new logging possibilities, including checkpoints where a simulation can be stopped and possibly reinitialized at a later stage. - As a new option, one can now disconnect the nitrogen part and simulate only the water dynamics; this could be valuable in water balance studies. The management language has been expanded, to among other things allow for simulating precision agriculture in the sense that the model uses information on mineral nitrogen in the root zone to calculate the need of fertilizer, and at the same time it takes legislative claims into account if organic fertilizer is applied. Models for cultivation in ridges and for subsoil irrigation have also been developed.

The plant growth models have been supplemented with simple models for simulating 1) forest and other permanent vegetation, and 2) vegetation for which only limited information is at hand; both are particularly important for area distributed simulations. Simulations with these simple models have a larger uncertainty than simulations with the original growth models. A parameterisation for coniferous forest has been established. Furthermore the mineralisation model has been adapted to take the effect of soil treatment into account. Work has been done 1) to integrate a new state of the art model for SVAT (Soil-Vegetation-Atmosphere Transfer) with an experimental GUI front end (Per Abrahamsen); and 2) to complete the user interface (Per Abrahamsen and Henrik Svendsen).
 

Activity report from Steffen Lauritzen (Jette Cloos)
During 2000, Dennis Nilsson has been working with decision problems that can be represented by Limited Memory Influence Diagrams (LIMIDs). He has written two articles together with Steffen Lauritzen, one for the journal Management Science, the other for the conference Uncertainty in Artificial Intelligence, UAI. As of November 2000, D.N. was working on three articles, with co-authors  a) Michael Höhle,  b) Anders R. Kristensen and Erik Jørgensen,  c) Anders L. Madsen. D.N. has given lectures at  1) Stanford, UAI (in July),  2) Foulum, the workshop "Decision problems in plant breeding" (in October),  3) Aalborg, Dept. of Decision Support Systems (in October). D.N. has cooperated with Michael Höhle on the implementation of the LIMID algorithm.

S.L. organised a workshop in Skagen (Klitgården) in week 37; the title was: "DINA Research Kitchen on Probabilistic Expert Systems for Genetic Analysis". It was a follow-up of a workshop in Italy earlier in 2000. The participants were: Robert Cowell,  Dept. of Actuarial Science and Statistics, City Univ. London;  Philip Dawid, Dept. of Statistical Science, Univ. College London; Thore Egeland, Center for Epidemiology, Rikshospitalet, Oslo; Julia Mortera, Dept. of Economics, Univ. of Rome; Vincenzo Pascali; Nuala Sheehan, Dept. of Mathematical Sciences, Loughborough Univ.; Paola Vicard, Dept. of Economics, Univ. of Rome; and Steffen Lauritzen, Aalborg University.
 

Activity report from Anders Ringgaard Kristensen
Nils Toft, PhD project.  A mathematics-economics candidate from Odense University, Nils Toft received a KVL stipend "for special areas" reserved for Dina as part of the co-financing of the network from KVL. The stipend expired on Nov. 30, 2000, and Nils Toft handed over his thesis in mid December. The main theme in the project has been simultaneous optimisation of decisions on multiple time scales. N.T. has written a survey article on the theme and has also developed an application concerning optimal decisions in disease control in slaughter pig production, simultaneously on a short and a long view. Since the strategy of supply (to the slaughterhouse) is an integrated part of the problem, it is also optimized in the application. At the end of 2000, an article on the application was completed and supplementing analyses for a chapter in the PhD thesis were in progress. In his project, N.T. has also been working with parameter estimation on livestock level, of which he gave a presentation at a conference in Spain in December; and together with Thomas Nejsum Madsen he has written an article on the analysis of eating patterns in growing pigs. As of Dec. 1, he has been employed at Dept. of Animal Science and Animal Health, KVL as an assistant professor in epidemiology.

Michael Höhle, PhD project.  A candidate in computer science from Aalborg University, Michael Höhle started on an ordinary KVL stipend on Feb. 1, 2000. After consultations with the reporter (A.R.K.) he has chosen to be attached to the Dina Research School. Prof. Finn Verner Jensen, Aalborg University, is formally assigned as a joint supervisor. The subject of M.H.'s project is: Temporal influence diagrams in pig production. The first months of the stipend were mainly used to carry out PhD courses at KVL and DTU, but there has also been time for research considerations, as reflected in a contribution by M.H. at a conference in Spain in September. At the end of 2000, M.H. together with Dennis Nilsson, was completing an article on the asymptotic properties of the LIMID algorithm, Limited Memory Influence Diagram.

A new project, in cooperation between Anders Ringgaard Kristensen and Per Abrahamsen. For a five month period,  Per Abrahamsen is working at the Dept. of Animal Science and Animal Health. His effort will be closely linked to a cooperation between KVL and DJF, Foulum on the integration of Limited Memory Influence Diagrammer and multi-level hierarchic Markov processes. Besides Per Abrahamsen, A.R.K. and Michael Höhle from KVL and Erik Jørgensen and Dennis Nilsson from Foulum are also involved in the project. Specifically, Per Abrahamsen will start by developing a language of specification and with implementing the algorithm for "Multi-level hierarchic Markov Processes", described in an article by Kristensen & Jørgensen (2000) in Annals of Operations Research. The integration with LIMIDs is described in an article (to appear) by Erik Jørgensen, Dennis Nilsson and A.R.K.; another article (in the journal Management Science, in print) by Steffen Lauritzen and Dennis Nilsson gives an account of the proper LIMID algorithm.

Thomas Nejsum Madsen, PhD project.  An agricultural economist trained at KVL, Thomas Nejsum Madsen has held a stipend funded by the Danish Research Academy and the National Committee for Pig Production/Danish Slaughterhouses. He has chosen to be attached to the Dina Research School during his PhD studies. The stipend expired in the spring of 2000, and at the end of the year, T.N.M.'s thesis was about to be completed. The project has focused on developing surveillance methods for use in pig production. Directly inspired by Iver Thysen's work with Kalman filtering methods in milk production in the first years of Dina, T.N.M. has developed these methods to monitor 1) growth of slaughter pigs, and 2) water consumption of piglets, as an indicator of their health status. The project has shown that the water consumption of a section of piglets follows a very regular pattern, and deviations from this pattern quite often indicate disease or other irregularities in the production. For the surveillancs, where Dynamic Linear Models are used, T.N.M. has developed a warning system based on methods from quality control. As a direct consequence of the PhD project, the National Committee for Pig Production has had a commercial software system developed, marketed under the name "FarmWatch (TM)". Pilot experiments with the system have proved most promising and it has been welcomed in the business. It is marketed in more than 100 countries across the world and has been translated into several languages. The thesis will be handed in in early 2001.
 

The effect of macropores on the transportation processes in the root zone
Charlotte Tofteng, PhD project.
This project, abandoned in 2000, investigated transport processes in the root zone for a structured agricultural soil containing biologically conditioned macropores. A theoretical part aimed at constructing a physics-based computer model to quantify the conditions for leaching through macropores; and an experimental part investigated, in a well-controlled laboratory experiment, the physical conditions for generating the macropore flow.

When C.T. called off her PhD stipend, the main elements of the computer model were already functionable, and interesting results had been achieved in the lab experiment. Two quite different macropore flow forms had been discovered: a slow film flow along the macropore wall, and a fast, pulsating stream. While C.T. has agreed to complete an article on her observations, this has been delayed by pregnancy and birth. The experimental layout was utilised by stud.agro. Birgitte Gjettermann to investigate the significance of the two flow forms for the sorption of phosphor in artificial macropores consisting of ceramic tubes coated with iron oxide.

Within the PIFT programme, the Danish Research Councils have contributed 1.96 million DKR which has covered a substantial part of the expenses of the above-mentioned projects.
 

Balance sheet 2000 and budget 2001

A balance sheet for Dina for 2000 and the approved budget for 2001 (not including PIFT funding, of above) can be found in an appendix. Largely the 2000 budget has held. However, there is a minor surplus which reflects activities that were not realised. The surplus has been carried over to the 2001 budget. Over the year, a few adjustments between the nodes were made.
 

Plans for year 2001

Annual Meeting
It has been decided that the annual meeting cooperation between Dina and DSIJ, The Danish Society for Informatics in Agriculture, should in 2001 also include the new agricultural technology network, DANET. By April 1, the meeting has already taken place, viz. on March 1-2 in Rebild Bakker; its theme was:  Information and Technology in Agriculture.

Another important event in relation to Dina is The 3rd International EFITA Conference, taking place on June 18-21 in Montpellier. Several Dina members are expected to participate.
 

Dina Research School
This year's summer school will take place in Ås, Norway on June 18-29. Title: Geographical Information and Spatial Analysis in Agriculture. The detailed planning of the school will be carried out by the Nordic cooperation partners of the research school. Besides this, the research school will as usual organize two PhD workshops, one on April 26-27 with the theme Optimization. The other on December 6-7 with the theme Tools for visualisation. Locations  have not yet been selected.
 

Priority research areas

Spatial Statistics in Agriculture
A workshop is planned for Nov.1, 2001; its theme has not yet been laid down.

Information Technology and Agricultural Engineering
The most important task in 2001 has already been carried out: to organise and to take charge of the annual meeting, March 1-2  (see http://cjt.auc.dk/Aarsm2001.html), arranged in a collaboration between Dina, DSIJ and DANET. As a consequence, the regular annual workshop for the priority area has been cancelled. It is hoped in Dina that the annual meeting has contributed to furthering the development in IT/agricultural technique. It is not clear to what extent this development will involve Dina, or if DANET should rather take charge of the entire area. A joint annual meeting indicates that a policy of cooperation has been drawn up.

Internet-based Advisory Systems in Agriculture
A workshop has been planned for May 10. Theme: How may the Internet supplement or possibly entirely replace traditional advisory service in agriculture and silviculture? Subheadings will, among others, include the coordination of central and local information; the use of video as a means of communication; and the support of the exchange of experience. The workshop is closely connected to current projects at the Danish Agricultural Advisory Center, Skejby and at the Danish Forest and Landscape Research Institute, Hørsholm.

Model-based Decision Support Systems in Agriculture
In connection with the Dina Annual Meeting, March 1-2, 2001 (see above), a presentation of research in the area was given, with focus on the connection between automatic data collection in stables, methods for data filtering and monitoring, and decision support with continuous updating of knowledge based on the observations just referred to. The discussion at the annual meeting will be continued at a workshop on Nov. 7, 2001, with focus on controlling the risk of false alarms from the many automatic monitoring/surveillance devices. Finally, at the EFITA Conference in Montpellier, a contribution will be given which summarizes experience with the use of Bayesian methods in relation to decision support in animal husbandry.

Bioinformatics and Quantitative Genetics
It is the intention to organise a workshop on biodiversity and phylogeny. Further information will be made available on the Internet as soon as it is available.
 

Publications
 

Dissertations  and  monographs

Møller, J., 2000. Aspects of Spatial Statistics, Stochastic Geometry and Markov Chain Monte Carlo. DSc Thesis accepted by the Faculty of Engineering and Science, Aalborg University.

Toft, N, 2000.  Elements of Decision Support Systems in Pig Production.  PhD thesis. Department of Animal Sciences and Animal Health, Royal Veterinary and Agricultural University, Copenhagen.  90 pp.

Thysen, I.; Kristensen, A.R. (eds.), 2000. Special Issue: The First European Conference for Information Technology in Agriculture. Computers & Electronics in Agriculture: 25.  307 pp.
 

Refereed  scientific  papers

Abrahamsen, P.; Hansen, S., 2000. Daisy: An open soil-crop-atmosphere system model. Environmental Modelling and Software: 15, p. 113-330.

Baddeley, A.; Møller, J.; Waagepetersen, R., 2000. Non- and semi-parametric estimation of interaction in inhomogeneous point patterns. Statistica Neerlandica: 54, p. 329-350.

Jensen, A.L.; Boll, P.S.; Thysen, I.; Pathak, B.K., 2000. PlanteInfo® - A web-based system for personalised decision support in crop management. Computers and Electronics in Agriculture: 25, p. 271-294.

Jørgensen, E., 2000. Calibration of a Monte Carlo simulation model of disease spread in slaughter pig units. Computers and Electronics in Agriculture: 25, p. 245-259.

Kendall, W.S.; Møller, J., 2000. Perfect simulation using dominating processes on ordered spaces, with application to locally stable point processes. Advances in Applied Probability: 32, p. 844-865.

Kicherer, S.; Backes, G.; Walther, U.; Jahoor, A., 2000. Localising QTLs for leaf rust resistance and agronomic traits in barley (Hordeum Vulgare L.). Theoretical and Applied Genetics: 100, p. 881-888.

Kristensen, A. R.; Jørgensen, E., 2000. Multi-level hierarchic Markov processes as a framework for herd management support. Annals of Operations Research: 94, p. 69-89.

Lund, J.; Rudemo, M., 2000. Models for point processes observed with noise. To appear in: Biometrika, vol 87 (2000).

Madsen, T. N.; Ruby V., 2000. An application for early detection of growth rate changes in the slaughter pig production unit. Computers and Electronics in Agriculture: 25, p. 261-270.

Nilsson, D.; Jensen, F.V., 2000. Probabilites of future decisions. In: Bouchon-Meunier, B.; Yager, R.R.;  Zadeh, L.A. (eds). Springer Lecture Notes in Artificial Intelligence: Information, Uncertainty, Fusion, p. 161-71.

Nilsson, D.; Lauritzen, S., 2000. Evaluating influence diagrams using LIMIDs.  UAI (Uncertainty in Artifical Intelligence), p. 436-45.

Østergaard, S.; Sørensen, J. T.; Kristensen, A. R., 2000. A stochastic model simulation of the feeding-health-production complex in a dairy herd. Journal of Dairy Sciences: 83, p. 721-733.
 

Papers  in  proceedings

Backes, G.; Jensen, J.; Jahoor, A., 2000. Localising QTLs for malting quality traits in barley. 8th International Barley Genetics Symposium II, p. 217-219.

Detlefsen, N.K.; Steffensen, M.; Thysen, I., 2000. Prognoser for godt sprøjtevejr - et indledende studium. 17. Danske Planteværnskonference. DJF-rapport nr. 24.

Grando, S.; Backes, G.; Ceccarelli, S.; Sabbagh, A.; Jahoor, A;, Baum, M., 2000. Quantitative Trait Loci (QTL) analysis for agronomic traits in recombinant inbred lines of the cross Arta x H. spontaneum 41-1. 8th International Barley Genetics Symposium III: 61-63.

Höhle, M.; Jørgensen, E.; Nilsson, D., 2000. Modelling with LIMIDs - Examplified by disease treatment in slaughter pigs. In: Proceedings from International Symposium on Pig Herd Management Modelisation,  Lleida, Spain Sept. 18-19, 2000, p. 17-26.

Jahoor, A.; Backes, G., Jensen, J.; Baum, M.; Walther, U., 2000. Are quantitative genes different from major race-specific resistance genes? 8th International Barley Genetics Symposium, Adelaide, Austrailie I, p. 53-55.

Jensen, A. L.; Lassen, P.; Röhrig, M.; Thysen, I., 2000. Transfer of knowledge and technology between countries mediated by the Internet. In: Proceedings of the second Asian conference for Information Technology in Agriculture, Suwon, Korea, June 15-17 2000, p. 158-167.

Jørgensen, E., 2000. Monte Carlo simulation models: Sampling from the joint distribution of “State of Nature”-parameters.  In: Van der Fels-Klerx, I.; Mourits, M. (eds). Proceedings of the Symposium on “Economic modelling of Animal Health and Farm Management”, November 23-24, 2000, Wageningen, The Netherlands. Farm Management Group, Dept. of Social Sciences, Wageningen University, p. 73-84.

Jørgensen, E.; Søllested, T. A.; Toft, N., 2000. Evaluation of SIR epidemic models in slaughter pig production units using Monte Carlo simulation. In: Proceedings of the International Symposium on Pig Herd Management Modelisation,  Lleida, Spain, Sept. 18-20, 2000, p. 27-36.

Kristensen, A. R., 2000. A software system for multi-level hierarchic Markov processes. In: Proceedings of the International Symposium on Pig Herd Management Modelisation. Lleida, Spain, Sept. 18-20, 2000.

Madsen, T. N., 2000. A model for detection of changes in the drinking behaviour of young pigs. In: Proceedings of the International Symposium on Pig Herd Management Modelisation. Lleida, Spain, Sept. 18-20, 2000.

Nilsson, D.; Jørgensen, E., 2000. Representing and evaluating POMDPs with limited memory. In: Workshop: Beyond Markov Decision Processes: Representations and Algorithms, Stanford University, USA, June 2000.

Röhrig, M.; Jensen, A. L.; Felkl, G.; Kjeldsen Bjørn, G., 2000. HortInfo - a web-based information system for horticultural production. In: Birkner, U., Amon, H., Ohmayer, Reiner, L. (eds). Proceedings of the 21st GIL annual conference, Freising-Weihenstephan, Germany, Sept. 6-8 2000, p. 174-179.

Schiemann, A.; Backes, G., 2000. The use of molecular markers in practical plant breeding. 8th International Barley Genetics Symposium III, p. 42-44.

Søllested, T. A.; Kristensen, A. R., 2000. A multi-process dynamic linear model for oestrus detection in group housed sows by modelling eating behaviour. In: Proceedings of the International Symposium on Pig Herd Management Modelisation, Lleida, Spain, Sept. 18-20, 2000.

Thysen, I.; Kristensen, A. R., 2000. Eds. COMPAC, Special Issue. The First European Conference for Information in Agriculture. Computers and Electronics in Agriculture 25 (3).

Thysen, I.; Jensen, A.L.; Röhrig, M., 2000. Pl@nteinfo - Status og fremtid. 17. Danske Planteværnskonference 2000. DFJ-rapport nr. 24, 101-111.

Toft, N.; Jørgensen, E.; Dethlefsen, C., 2000. Estimation of farm specific parameters in a longitudinal model for litter size with variance components and random dropout. In: Proceedings of the International Symposium on Pig Herd Management Modelisation, Lleida, Spain, Sept. 18-20, 2000, p. 127-136 .

Toft, N.; Kristensen, A.R., 2000. Simultaneous optimization of decisions on multiple time scales in slaughter pig production. In: Workshop: Beyond MDPs: Representations and Algorithms, Stanford University, Stanford, CA, June 30.

Toft, N.; Kristensen, A.R.; Jørgensen, E., 2000. A framework for decision support related to an infectious disease in slaughter pig production. In: Van der Fels-Klerx, I.; Mourits, M. (eds.). Proceedings of the Symposium on “Economic Modelling of Animal Health and Farm Management”, Nov. 23-24, 2000, Wageningen, The Netherlands. Farm Management Group, Dept. of Social Sciences, Wageningen University, p. 109-117.
 

Other  publications

Christensen, O.F.; Diggle. P.J.; Ribeiro Jr., P.J., 2000. Analysing positive-valued spatial data: the transformed Gaussian model. In: Allard, D.; Monestiez, P. (eds). GeoENV III - Geostatistics for environmental applications, Quantitive Geology and Geostatistics, Kluwer Series (to appear).

Christensen, O.F.; Møller, J.; Waagepetersen, R.P., 2000. Analysis of spatial data using generalized linear mixed models and Langevin-type Markov chain Monte Carlo. Research report R-00-09, Aalborg University.

Christensen, O.F.; Møller, J.; Waagepetersen. R.P., 2000. Geometric ergodicity of Metropolis-Hastings algorithms for conditional simulation in generalised linear mixed models. Research report R-00-10, Aalborg University.

Döge, G.; Mecke, K.; Møller, J.; Stoyan, D.; Waagepetersen, R., 2000. Grand canonical simulations of hard-disk systems by simulated tempering. Research Report R-00-2003, Department of Mathematical Sciences, Aalborg University.  Submitted.

Flensted-Jensen, M.; Hansen, P.E.; Thysen, I.  (eds.), 2000. Dina 1991 - 1999.  Dina Publication Series no. 3. ISBN: 87-89817-05-2.  53 pp.

Flensted-Jensen, M.; Hansen, P.E.; Thysen, I., 2000. Dina 1991 – 1999 :  Publications. Dina Notat no. 85.  6+31 pp.

Hansen, M. B.; Møller, J.; Thøgersen, F., 2000. Bayesian Contour Detection in a Times Series of Ultrasound Images Through Dynamic Deformable Template Models. Technical Report 2000-04, Biometry Research Unit, Danish Institute of Agricultural Sciences,  Research Centre Foulum.

Jørgensen, E., 2000. Elements of Bayesian network specification in an animal health eco-nomy research project.  Internal Report 2000-02, Biometry Research Unit. Danish Institute of Agricultural Sciences.  24 pp.

Jørgensen, E., 2000. Decision Support Systems in Animal Production. A Bayesian Future? Abstract submitted to the symposium of “Decision support systems in animal production”. The 3rd Conference of the European Federation for Information Technology in Agriculture, Food and the Environment. Montpellier, France, June 18-20, 2001.

Kristensen, A. R., 2000. Dinas Forskerskole: Workshop 6-7 april 2000. Dina Notat no. 87.

Lauritzen, S.; Nilsson, D., 2000. LIMIDs of decision problems.  Submitted to: Management Science.

Møller, J., 2000. A review on perfect simulation in stochastic geometry.  Research Report R-00-2016, Department of Mathematical Sciences, Aalborg University.  Submitted.

Nilsson, D.; Goldberger, J., 2000. An efficient algorithm for sequentially finding the N-Best list. Submitted to: IJCAI 2000.

Appendix
Dina balance sheet 2000 and budget 2001

Dina balance sheet 2000 (UNIT: 1000 DKR)

Activity

Budget type *)

Budget

KVL IMF

KVL IHH

DJF

AAU

Skejby

DTU

FSL

Risø

Others (SVS)

Deviation from budget

Network

Head	A	25	25									0
Deputy head	A	10			10							0
Annual meeting	B	15					10					5
10 IT Corrdinators, basic allowance 5	A	50	5		15	5	5		5	5		10
IT Coordinators, Special allowance (10 per workshop per coordinator)	A	100	10		30	10	10		10	15		15
IT workshops	B	50			20	10				5		15
Secretarial (KVL) and Webmaster	A/B	100	55		13							32
General costs	B	20	18		18		4					-20
Total, network		370	113	0	106	25	29	0	15	25	0	57

Research School

Head	A	200		200								0
3 Eksperts basic allowance 30	A	90	30					30			30	0
Experts, special allowance 25 per summer school, 5 per workshop	A	35	5					25			5	0
2 workshops	B	50			52							-2
1 summer shool (funded by NOVA- NorFA)	B	0		0								0
General costs	B	5		5					11			-11
PhD scholarships	B	0		0								0
Total, Research School		380	35	205	52	0	0	55	11	0	35	-13
Total expences		750	148	205	158	25	29	55	26	25	35	44
Membership contribution		750	95	205	200	50	50	50	50	50	0	0
Deviation - to be settled			53	0	-42	-25	-21	5	-24	-25	35	-44

*) A: Fixed amount.  Each institution disposes of the money as long as the work is done.
*) B: Estimated cost.  The final amount should be accounted for.

*) A: Fixed amount. Each institution disposes of the money as long as the work is done.
    B: Estimated cost. The final amount should be accounted for.
 
 

Author:webmaster. Updated: April 2001.