Table of Contents
Integrated Modeling Approaches and System Dynamics in
Education related to Sustainable Resource Management,
Forestry, and Land use Management
Experiences and Perspectives at the School of Forestry and Resource
Management, Technische Universitat Miinchen, Germany
Peter Biber and Hans Dieter Kasperidus
Dr. Peter Biber
Chair of Forest Yield Science
Technische Universitat Miinchen
Am Hochanger 13
85354 Freising, Germany
Tel.: 0049 8161 714708
Fax.: 0049 8161 714721
Email: peter.biber@lrz.tu-muenchen.de
Dipl.-Ing. Hans Dieter Kasperidus
UFZ Centre for Environmental Research Leipzig-Halle
Department Urban Regions
Permoserstr. 15
04318 Leipzig, Germany
Tel.: 0049 341 235 2601
Fax.: 0049 341 235 2534
Email: hans.kasperidus@ufz.de
Abstract
The School of Forest Science and Resource Management at Technische Universitat
Minchen, Germany offers two different streams of education. The traditional stream is
education in forest science (BSc, MSc, Diploma), which traditionally qualifies forest
and related enterprise managers and scientists. The other stream is the 2001-founded
international master course in Sustainable Resource Management (SRM), which
qualifies managers or consultants who have to deal with interdisciplinary problems in
resource management worldwide.
With the SRM program, the organisers put more emphasis in education on teaching
systems thinking and practice system dynamics skills to increase the number of qualified
academics in this field. Thus, resource management and forestry education included
and improved system dynamics (SD) and modeling issues in their curricula. The goal is
to improve the student’s systems thinking skills to enable them applying resource
management and foresty in their home countries highly based on systems principles.
This paper describes the structure of the courses and demonstrates, which aspects of
SD, modelling and models are currently used for which purpose in these fields of
education. In our experience up to now, students appreciate these issues and are
interested in more SD and modelling related courses. We also observe typical
misunderstandings among the students concerning the meaning of stocks and flows and
the role of simulation software.
Our main conclusions and future perspectives are, that by providing high-quality
lectures in SD-related issues we should try to raise the student's demand for getting
more subjects presented in an SD context, which in turn should stepwise increase the
supply. As especially the SRM students are a very international group, we hope to be
multiplicators for systems thinking worldwide.
Keywords: Education, Systems Thinking, System Dynamics, Integrated Modelling,
Resource Management, Forestry, Sustainability.
1 Introduction
Despite the widely accepted systemic quality of many issues in natural resource
management and land use planning, the widespread application of systems thinking,
integrated systems modelling, and system dynamics in these fields’ research and
education activities is nowadays not well developed, and particularly in German
speaking countries far beyond its inherent possibilities. Moreover, the awareness of the
value of theory and application of system dynamics in forest science, forest
management practice, and in forestry related regional planning issues is by far not
adequately recognised. One reason seems to be the fact that systems thinking and multi-
disciplinary model building is mostly absent in the regular universities’ education
programs. Astonishing is that this matter of fact also applies to other fields of science
like engineering sciences (Nadtke 2003) or social sciences (Repenning 2003). Even in
forest science and forestry, with its long tradition in observing the forest as complex
ecosystem and in developing a wide range of modelling approaches (Mies 1998),
integrated modelling at the landscape level and above is at the beginning and is seen as
the challenge of the near future (Chertov et al. 1999). Integrated modelling implies a
great potential in combination with the central paradigm of the concept of sustainability
in the use of forest resources that has been coined by German foresters like Hans Carl v.
Carlowitz (1713), J. Ch. Hundeshagen (1828), and Georg Ludwig Hartig (1820), and
Heinrich Cotta, who 1811 established worldwide the first school of forestry in Tharandt,
Saxony. Sustainability is still the self-evident red thread of today’s professional forestry
education. Furthermore, forest management in general has a long tradition in long-term
thinking and interdisciplinary approaches in both science and practical management that
are of particular value in all related fields of resource management.
Based on these considerations the idea for a new integrative and international Master of
Science Programme was born and developed at the School of Forest Science and
Resource Management at the Technische Universitat Miinchen in Freising, Germany.
By pooling expertise from the School of Forest Science, at place for the classical
forestry education, completed with expertise from teachers of other faculties of the
University and from external experts, the School managed to compile an
interdisciplinary spectrum of courses, which, in the end, have one common goal:
training of skills required for achieving sustainability in real-world resource systems. As
a result, the International Master of Science (MSc) Programme in ‘Sustainable Resource
Management’ (SRM) was implemented at the Technische Universitat Miinchen in 2001.
Furthermore, the organisers took the opportunity to offer special courses that deal in
particular with systems thinking, systems analysis and system dynamics with the goal to
put these kinds of knowledge and skills at the centre of the education. In the following
chapters we will discuss the structure of the SRM program and the traditional education
at the school in relation to system thinking and system dynamics. We will give a short
overview which kind of system dynamics models are currently in use and which are
available for various educational purposes in future. Finally, we will share our first
years’ experiences with the students’ attitudes regarding System Dynamics and express
our plans we have for the future.
2 The courses
2.1 Sustainable Resource Management (SRM)
The general objective of the international MSc program is to prepare students, who
already possess an eligible academic degree, for professional work in the various fields
of resource management that addresses all kinds of natural resources like water, air, soil,
plants, and wildlife as well as land use management and human resources.
Master's thesis
3.Sem.
2 months internship abroad
Extended courses: 2 outof 8:
Wildlife & Protected Area Management Sustainable Agriculture
Landscape Management Forest Ecosystems 2.Sem.
Resource & Environmental Economics Material and Waste Management
Watershed & Soil Management Renewable Resources
Basic courses
Management of Human-Nature-Technology-Systems
System Analysis and Information Management 1Sem.
Inventory and Planning
Management Concepts
Communication Techniques
Fig. 1. Structure of the MSc course Sustainable Resource Management (SRM) at the
School of Forest Science and Resource Management, Technische Universitat Miinchen.
This requires an interdisciplinary and integrated teaching programme that provides the
systemic understanding and skilful application of important concepts, tools, and
techniques in resource management including important ‘soft’ skills for professionals
like rhetoric, conflict management and moderation. These special management skills
will increase the capabilities of the Graduates to solve natural and human resource
problems in their own way and appropriate to their own cultural and natural
environment. The goal is that graduates learn to design and implement management
regimes for sustaining high-productivity use to meet real human needs in that way that
it is done sufficiently and reliably, efficiently and appropriately, justly and enjoyably.
Moreover, the education in this program takes the students beyond the boundaries of the
traditional disciplines such as forestry and agriculture.
The MSc program is designed for 3 semesters (Figure 1). If required, the students can
participate in an optional preparatory course which provides orientation and add-on
language skills. The first semester requires the attendance in a series of five basic
courses dealing with concepts and techniques of resource management. These courses
include a) Management of Human-Nature-Technology-Systems, b) System Analysis
and Information Management, c) Inventory and Planning, d) Management Concepts,
and e) Communication Techniques. During the second semester the students have to
chose two elective fields out of eight fields of specialisation which are a) Wildlife and
Protected Area Management, b) Landscape Management, c) Resource and
Environmental Economics, d) Watershed and Soil Management, e) Sustainable
Agriculture, f) Management and Protection of Forest Ecosystems, g) Material and
Waste Management, and h) Renewable Resources. The third semester is dedicated to
prepare the master thesis. The programme is accompanied by field trips and case
studies, and it includes an internship in another than the student’s home country.
Starting with about 20 students from 6 countries in the first year, 76 new students from
more than 40 countries in the second year, the MSc-Programme reached its full capacity
with about 80 international beginning students in the third year, which seems to
confirm the attractiveness of the concept and the lectures.
In three lectures of the MSc-Program the students directly get in touch with SD and
integrated modelling: a) “Introduction to Sustainable Resource Management”, b)
“Systems Analysis and Model Building”, and c) “Forest Growth for Resource
Managers”.
2.1.1 Lecture “Introduction to Sustainable Resource
Management”
The lecture “Introduction to Sustainable Resource Management”, as main part of the
basic course “Management of Human-Nature-Technology-Systems”, is designed to be
the “red thread” through the whole master course in order to provide basic and
theoretical knowledge as well as coherence. Therefore, the aim of the lecture is to
define the most important perceptions of sustainable resource management, to show
typical properties of biological, social, and technical systems, their combinations and
interactions. The lecture ends up discussing examples for actual interdisciplinary
problems of resource management and the corresponding efforts to find sustainable
solutions.
Each topic is represented by the university’s expert scientists and by external experts.
One key issue of this lecture is to communicate the value of systems thinking, system
dynamics and systems analysis as appropriate tools for a comprehensive
interdisciplinary understanding and handling of complex systems that will result in
better management strategies for real-world-systems. We intend to provide the basis to
qualify our students to be managers or consultants able to solve interdisciplinary
problems in the context of sustainable use of natural, human and technological systems.
The goal is to improve the student’s systems thinking skills to enable them applying
resource management in their home countries based on systems principles. This shall
contribute to a wise management of resources for the benefit of future generations as
well as the present one and that they produce high yields sustainably. We strive for
deepening their understanding of system theory and system thinking, the relationships
between structure and long-term behaviour of resources, teach the principles of
relationships between stocks and flows, the concepts of feedback, non-linearity and
delay. In addition, we introduce important tools and methods, such as causal-loop and
stock-and-flow diagrams, and emphasise the role and importance of simulation models
in the SD context.
2.1.2 Lecture “Systems Analysis and Model Building”
This lecture is designed to amplify and to consolidate the basic knowledge from the
systems dynamics part of the lecture “Introduction to Sustainable Resource
Management”. The lecture defines the most important perceptions of systems analysis
and model building. The value of models in practical decision support and in the
process of scientific cognition is discussed as well as examples of different model types
for different purposes.
A very important part of this lecture is to delineate the path of abstraction leading from
a real system to a mathematical model and its implementation as a simulation tool.
STELLA turned out to be our choice as teaching aid for this task. The university’s
computer laboratory provides computers with a STELLA campus installation that offers
students unlimited access to this tool. We give lectures to guide the students how to
operate the software. By this way, STELLA is intensively used to explain the most
important generic system structures, their typical dynamics and their occurrence in
actual biological, social, technical and psychological systems. The exercises with the
software shall enable and encourage the students to experiment with self-constructed
models. As a part of the examination, students get an assignment to analyse a story
containing a dynamic problem and find approaches for solutions using STELLA.
2.1.3 Lecture “Forest Growth for Resource Managers”
As mentioned above, advanced SRM students have to choose two fields of
specialisation. One of these fields is called “Management and Protection of Forest
Ecosystems”. Here we communicate the basic static and dynamic properties of forest
ecosystems, and the most important sustainable forest management concepts on
landscape level or above to future decision-makers.
The lecture “forest growth” offered by the chair of forest yield science is a part of this
elective field. It teaches driving forces and biological laws of forest growth, how to steer
forest growth, and how to transform it into specific forest products.
2.2 Bachelor, Master, Diploma in Forest Science
The classical and traditional qualification offered by the school is the education in
Forest Science on different levels: Bachelor, Master, and Diploma. Traditionally, its
main focus is to educate forest managers on enterprise or larger level (> ~5.000 ha) and
forest scientists of all specialisations. In the past decades, the study programmes have
been revised several times to adapt to changing needs; the last revision has been done in
2000. Nevertheless, this education has always necessarily been very comprehensive, as
forest managers have to act all along in the biological, technical and human dimension
at the same time. Here we identify the greatest potential benefit of systems thinking and
SD in education. This recently attracts very different kinds of students. Contrarily to the
situation ten and more years ago, most of the today students don’t want to work in
classical forestry from the beginning of their studies but chose to study forest science
anyway, because they feel to be qualified for a large variety of professions from applied
ecology to pedagogics.
Degree Master Ssem.
Diploma-thesis Master-thesis
Extensionblock
Optional I++
subjects
Bachelor Ssem:
Extensionblock Bochelorunesis
es Opticnal
subjects
Extensionblock |
Extension programmes
Foundation asems
Foundation courses
Fig. 2. Structure of the Forestry Education at the School of Forest Science and Resource
Management, Technische Universitat Miinchen. Degree: Diplom-Forstingenieur (Univ.)
in Forest Science, Bachelor of Science (BSc) in Forest Science, Master of Science
(MSc) in Forest and Wood Science.
Depending of the degree aimed at, the study programme in forestry is designed for 9
semesters (fig. 2). The first four semesters consist of the foundation courses, which deal
with a) general basics of natural and social sciences, b) natural-scientific, and c) social-
scientific basics of forest science, and d) basic aspects of forest production and
technology in forestry. From the fifth to the sixth semester, we operate obligatory
extension programmes, where the knowledge from the foundation courses b) to d) is
deepened in order to reach a professional level. Parts of the sixth semester are already
open for students to choose extension blocks from a large range of possibilities, such as
International Forestry, Nature Protection, Land Use Planning, Dynamics and
Management of Population, Forest Enterprise Management, Geographical Information
Systems, Methods of Wood Science and many more. Depending of the degree one or
three extension blocks have to be chosen and a thesis has to be prepared. There is also
room for attending optional subjects. Regularly the education is finished within 9
semesters (Master, Diploma) or 6 semesters (bachelor) respectively (Figure 2).
Before the last revision of the study programme in 2000, the numbers of beginning
students oscillated around 50 per year. Since 2000, the beginning student numbers
steadily increased from 48 in 2000 to 100 in 2003. Issues related to modeling and SD
are directly dealt with in the lecture “Forest yield Science”.
Lecture “Forest Yield Science”
In the framework of the classical forestry education, the Chair of Forest Yield Science
offers a lecture called “Forest Yield Science”. Its purposes and contents are similar to
the lecture forest growth offered in the resource management framework, but the
knowledge communicated here is much more detailed as it is aimed at qualifying people
who directly have to manage a forest. The lecture consists of two parts, where the first
belongs to the foundation courses and the second to the extension programmes.
It focuses on understanding which driving forces and biological patterns produce the
typical dynamics of single trees and forest stands. Furthermore, we instruct, how, in
quantitative terms, these dynamics can be influenced by silvicultural management. As
forest growth models play an important role in this context, students learn to know,
which main types of such models are existing, which purposes they are serving, how
they work, and how they can be constructed. In a practical part of the lecture, students
perform management scenario simulations for different forest stands with the forest
growth simulator SILVA (Pretzsch et al. 2001), where they learn to handle such a
model in practice, and at the same time deepen their understanding of forest dynamics.
3 Available models
3.1 Models currently in use
3.1.1 Simple model games for teaching important system
archetypes
In the practical part of the lecture Systems Analysis and Model-Building, students learn
to know some of the very important system archetypes by practice modelling in the
STELLA environment. We start with exponential growth and asymptotic decay,
proceed to their interactions, which leads to s-shaped growth and overshoot and collapse
patterns. Furthermore we introduce system structures producing oscillation and main
chains. When introducing a system archetype, we always try to connect it to a very
striking example, e.g. exponential growth is developed by example of a bacteria
population starting to grow with —for the time being- unlimited resources. When the
typical process-generating pattern has been made clear, we show, again by examples,
that basically the same structures generate the same process in very different real world
systems, from biology to economics to psychology. E.g. for many students it seems to
be a new perplexing insight, that constant relative growth rates for economy —as
postulated by most politicians- are absolutely impossible on the long run, as this would
mean unlimited exponential growth.
When these subjects have been understood by the students, they are already able to
solve somewhat larger problems on their own. As a part of the examination they have to
do an assignment, where a resource management problem is presented as a story; in the
story, there is always a client, who has the problem, and the student act as the
consultant. The story has to be analysed carefully in order to identify the important
structures of the system. Then, it must be step-by-step transformed into a model using
STELLA. This model must be used to answer some questions related to dynamic
problems. Finally, the students have to write a short report for the client, where they
have to express their own view of the problem and show possible solutions.
recitation in mm
conga KD
o—_o—_
Ground Water nus
Regeneraion round Wgler
Pool eo
Draling Rate
1: Suppy per Head 2:Pepuation Total Harvest tons
: 0,
EF sasaos
x eof A
® 04s vA
z
5 ‘S000 <i
os a IE =
im 5
c .
2 $8000,
Ei 10000
50 28:00 3750 50.00
Pace 1 Years 18:27 Mon, 8 Apr 2003
United
Fig. 3. Constructing a model for population dynamics in the fictive country
Nabakoumaland as a task of examination in Systems Analysis. In the scenario shown,
unwise use of ground water resources causes first exponential population growth, then a
drastic breakdown due to famine.
As an example, the 2001 story introduced king Nabakouma III, absolute ruler of a
fictitious small and poor country in Africa. Nabakouma III wants to enhance prosperity
and welfare for his country and people by utilising a newly discovered underground
water reservoir for large-scale irrigation of their fields. World bank announced financial
support, but demands a professional advisory report. With an appropriate model (Figure
3), one can show, that only a very thought-out water management regime together with
birth control policies can improve life conditions and prevent a catastrophic overshoot
and collapse pattern in the population dynamics.
3.1.2 Normal Forest Model NORMFOR
In the Lecture ‘Forest Growth’ for the SRM students learn about the so-called ‘normal
forest model’, and we developed the STELLA-based model NORMFOR to demonstrate
the idea of the ‘normal forest’, its typical dynamics and the possibilities to move a real
forest towards the ideal state of ‘normality’.
Area Age Class ‘Transition
co
Felling Area 2 ae
o—{lIKs
@|
Felling Area 3 ee.
6 @
vm
‘Area Age Class 3 reaDeficte 4
Feling Areas ation & greater 4
5B (i A
‘Area Age Class 4
Fling Area greater 4
ie
‘Area Age Class greater 4
Oo
Vol age Class greater 4 pera ———
Fig. 4. Main structure of the Normal Forest Model NORMFOR. Stocks represent the
areas occupied by respective forest age classes, flows represent the sequential transition
of forest areas from age class to age class and the sudden transition to age class 1, when
an area is clearcut and replanted. Note that most converters are ‘ghosts’; their
calculation is not shown in this diagram.
The ‘normal forest model’ is a mental model, every professional forester carries in
his/her mind. In its basic form it had been invented and already formulated
mathematically by Hundeshagen (1826) and Heyer (1841). It still forms the backbone of
sustainable forest management worldwide. A ‘normal forest’ is a forest in a larger area,
which is in an ideal state of perfection under sustainable management in the strictest
sense (Assmann 1961, Ostmaston 1968, Speidel 1972). This ideal state is a dynamic
equilibrium where the forest delivers continuously the same yield in terms of quality
and amount. Annual increment, growing stock, annual harvest, yield and expense are
constant. Forest scientists developed many indices, which quantify the deviation of
‘normality’ for an actual forest. In addition, a whole set of indicators is available -and
obligatorily used by foresters- which are used to derive a felling budget which moves a
forest from its present state nearer towards normality. In practice, this is an iterative
10
process, which is often disturbed by unforeseeable events. Thus, true normality is never
reached, but the normal forest model guarantees, that this exemplary state of perfection
never gets out of scope.
In the model, a normal forest is defined as a set of forest stands, where each stand age is
represented on the same area. Each year, the oldest stand is always cut and immediately
replanted. In this idealised forest, Except for age, everything else, like tree species, site
conditions, etc., is the same and constant over time.
In NORMFOR, each area covered by a forest at a special age (class) is represented by a
stock. These stocks are sequentially connected by flows, which force the transition of
areas in higher age classes as time passes by. Each age class has a certain wood volume
per unit area, so, transition of an area from one age class to another is connected with an
accumulation of wood on this area. When wood on a certain area is harvested, this area
is considered as to be replanted and will be shifted to the youngest age class (Figure 4).
Timber harvest is defined by an annual felling budget, which expresses the timber
volume to be cut. Cutting starts at the oldest age class and will —sequentially- affect
younger age classes, when this is necessary to reach the felling budget.
For the sake of simplicity, the parameters of NORMFOR are set on values, which
approximately reflect the dynamics of tropical or subtropical short rotation plantation
tree species, such as certain eucalypts. The model includes four age classes each
covering a span of five years and one age class, which contains everything older. The
whole model forest area amounts to about 100 ha.
Felling Volume Total: 1-2-%-4-
4000-
4: 3000:
a, a
1 2000: t
0.00 12.50 25.00 37.50 50.00
Page 2 Years 14:00 Don, 1. Apr 2004]
Untitled
Fig. 5. Scenario runs with the Normal Forest Model NORMFOR. Development of felled
timber volume in a virtual forest enterprise controled by different felling budget
indicators. Although adjusted for aiming at the same equilibrium point each indicator
produces a different level of steady state.
Before running the model, the user must define the start frequency distribution of the
age classes and a constant annual felling budget. After some time, the forest will reach
normality, but its key parameters, annual volume increment and total standing wood
volume will stabilise on very different levels. As an alternative to setting a constant
felling budget, the model user can also choose one of five common forestry felling
11
budget indicators (cf. Speidel 1972) to be applied iteratively at each time step. These
indicators use different methods to calculate a felling budget based on information
about the forest, e.g. total standing wood volume, current and past wood increment,
desired time period to achieve normality. It can be demonstrated, that the different
indicators don’t perform equally well depending of the start situation (Figure 5). While
each of the indicators finally leads to a normal forest, the desired normal standing and
regularly felled timber volume and increment level are often missed significantly.
While, classically, the normal forest model is being taught as a set of mathematical
equations, its main consequences can be made clear more quickly via simulating with
NORMFOR. The strength of the model is, that it is a valuable teaching aid to put across
the core concept of sustainable forest management worldwide. Understanding this
concept is absolutely essential for resource managers, when they are concerned with
forestry issues.
3.1.3 Forest Growth Simulator SILVA
The forest growth simulator SILVA (Pretzsch et al. 2001) is a tool, which allows to
simulate forest development for mixed and pure, even- and uneven-aged forest stands
under a large variety of silvicultural concepts. When the first version of the model was
constructed end of the 1980s, the main purpose was to support forest management
decisions from the point of view of multi-criterial sustainability, but we increasingly
recognized SILVA as an ideal tool for teaching forest dynamics and its feedback
structures with management issues.
Site conditions
Stand structure Tree growth
Disturbances
(Thinnings,
Windthrow,
Snowbreak)
Fig. 6. Coarse cause effect-structure of the forest growth simulator SILVA.
In the simulator a forest stand is a set of single trees whose coexistence is modelled as a
spatial-temporal dynamic system. Compared to traditional tools, already the coarse
cause-effect-structure of SILVA (Figure 6) is based on an enlarged dynamic
understanding of forests. Thus, the increment of a tree results from the site conditions,
the present state of the tree, expressed through crown dimensions, and the spatial
structure of the stand, expressed by the three-dimensional space utilisation of the single
trees. This stand structure results in a specific spatial growth constellation for each
single tree, which more or less reduces its growth. Changes of this structure, due to
12
thinnings, growth or mortality, consequently change growth constellations and therefore
have a significant impact on the further development of the trees, the stand and its
structure. Figure 7 illustrates the sequence of steps in a simulation run with SILVA.
Forest stands are dynamic systems, which are steered by strong feedback effects.
Exactly these feedback mechanisms are utilised by foresters when they remove trees in
order to steer the development of the remaining stand and to generate revenues. As
SILVA explicitly models these feedbacks, the simulator has proven to be a very
effective tool for simulating the dynamic response of forests to a large variety of
management scenarios.
Starting values | > module reflecting site/
site conditions growth relationships
data stand
missing structure
2 generator
<<!
3D-competition module
thinning module
mortality module
allocation module |
regeneration module
rowth and yield
y g! y! a
t=0..n
ecology
vy ia
t=0...n
e socio-economy output
v >
t=0...n
Fig. 7. Forest growth simulator SILVA: flow-chart (right) and forest stand development
simulated over 100 years (left).
13
SILVA is no so-called mechanistic model, which explains tree growth starting at the
level of physiology. Single-tree growth in SILVA is described as the change of
dendrometrical variables as diameter in breast height, total height, crown diameter and
length. The driving variables are highly aggregated; such as there is e.g. a single index
for soil water supply. Competition among trees is described by geometrical methods.
The growth equations driving the model are fitted empirically based on a unique data
set from long-term research plots partly covering a time span of more than 100 years.
Nevertheless, biological plausibility never came out of view when formulating the
model equations. The input data needed for using the model is constrained to
information, which is accessible for forest managers from their standard site
classification or regular inventory. The output covers a large range of information
covering timber production as well as economic and ecologic issues (Biber and
Weyerhaeuser 1998, Pommerening et al. 2000). Thus, the simulator provides
information and generates results, which can be used to assess sustainability from a
multi-criteria point of view. The simulator is not limited to small forest areas. Data from
large-scale sample forest inventories can be used to perform scenario calculations on
large regional level (e.g. Biber et al. 2001). Due to the broad range of information
provided by SILVA and due to its biological plausibility, it has played a key role in the
interdisciplinary sector study “German Forest Sector under Global Change” (Lindner
and Cramer 2002, Pretzsch and Dursky 2002)
Students come in touch with SILVA mainly in the courses related to forest science,
where they intensively work with the model. But we also present SILVA as an example
for a management-related dynamic model in the lectures ‘systems analysis’ and
‘introduction to sustainable resource management’.
3.2 Models considered for future application
The models shown in the following chapter are readily available and can in our opinion
contribute to a great deal to transporting problem-specific and SD related knowledge. In
the framework of UNESCO’s ‘Man and Biosphere MAB’ program and _ national
research activities in the 1970s and 1980s some of the integrated projects incorporated
model-building activities, which aimed in the construction of large integrated, socio-
economic, ecological models. From these activities and additionally from international
trainings courses, mostly organised from the German Foundation for International
Development (formerly DSE and nowadays INWENT), a remarkable set system
dynamic models have been developed. Most of these models are described in research
reports, technical reports, scientific papers and other media, at least a complete
documentation of the source code is available (e.g. Grossmann et al. 1993). These
models are of particular value for the system dynamics knowledge and model base and
in particular for education from various points of view. Firstly, they represent important
examples of applications of system dynamics models of real-world systems at regional
or community level; secondly they principally allow a review of the predictive
capabilities of the models as time progresses and the modelled real systems evolved;
thirdly they embody the state-of-the-art in system dynamics modelling of that time that
seems to be of equal importance nowadays. Fourthly, they are valuable teaching and
learning models to get acquainted with more complex models than the small ones
usually supplied with system dynamics tutorial material. Mostly, it is necessary to
transform the models into up-to-date simulation environments like STELLA, which is a
good training for beginners and advanced modellers.
14
From this pool the following models were selected for introduction in this paper:
Proto - National Model with Extended Forest Sector (Grossmann 1994)
The model depicts the interaction between a national economy and the forest sector
(Figure 8).
Fig. 8. Stock-and-flow diagram of the forest sector part of the model Proto. This model
is completely documented (Grossmann et al. 1993).
It also includes a renewable and a non-renewable resource sector, a sector with four
land use categories, a food sector (as energy equivalent), an economy sector, and a
pollution sector. Although, it had been applied and modified by other groups,
nowadays, it would need considerable adaptations to new insights in e.g. economical
development. Nevertheless, it is well suited for training purposes and as a starting point
for modelling similar problems.
SECENG - Forest Damage by Secondary Pollutants (Grossmann 1991)
The model deals with the problem of forest die-back caused by secondary pollutants and
it explores the most probable reasons for this phenomenon. It reproduces the impacts of
secondary pollutants in synergistic combination with SO2 and NOx studied in the area
of the ‘Rosalia Lehrforst’, a forest demonstration centre of the BOKU - University of
Natural Resources and Applied Life Sciences, Vienna, Austria. The model contains the
stock and flow structures for pollutants (NOx, SOx, VOC), biomass of the forest stand,
and biomass of the tree foliage (Figure 9). Damage is modelled as total impact by
pollutants affecting foliage biomass. The model was linked to risk maps for forest
damage produces with geographic information systems (GIS) to compare model results
with field data. This approach has been applied in other projects (cf. Grossmann and
Schaller 1986).
15
Fig. 9. Stock-and-flow diagram of SECENG. This model is completely documented
(Grossmann et al. 1993) but not transferred to up-to-date modelling environments yet.
Agroforestry Village Community in G ebang, J ava, Indonesia (Grossmann 1998)
The purpose of the model is to simulate the agroforestry scheme of a village community
in the long run with respect to food production, income from forestry, and development
of the village population. Therefore, the model allows assessing the sustainability of
different types of agroforestry systems and of population policies. Figure 10 shows the
Meta-structure of the model and the interactions that have been established between the
following sectors: forestry, agricultural crops, commercial fruit and wood production,
population, income, and food demand and supply. Figure 11 shows the results of a
scenario run with the agroforestry model.
4 Experiences
At the end of each semester the students give their opinion about the quality and
relevance of the lectures. The students’ response shows that the lectures and courses
teaching systems thinking and systems analysis belonged to the ones that are most
appreciated. In general, we experience our students to be very open-minded for SD and
modelling issues. Many are very interested and try to keep on track with SD also
beyond the normal lectures and consider its methods as important tools in their coming
professional career. Some resource management students already started their master’s
thesis addressing problems that require SD application. Not surprisingly, students
generally like STELLA as a valuable and useful tool, which allows to construct
executable computer simulation models and test ideas without being too familiar to the
inherent mathematics. Also, from the teacher’s point of view this tools allows us to
interest far more persons for modelling and SD than we could reach with a pure
difference-equation-based teaching concept.
16
Model Meta-Structure
Income 7 Population S77
Income fe Population
4 4
Forest Fruits
Agricultural
Crops
Fig. 10. The sector structure of the Agroforestry Village Community model prepared
with the software STELLA.
However, the apparent easiness of handling models with STELLA often conceals two
common misunderstandings among students. Although repeatedly instructed and
highlighted in the lessons, when revising the exams, which include similar questions as
proposed by Booth Sweeney and Sterman (2000) and Sterman (2002) as ‘Bathtub
Dynamics Tasks’, we find out, that there is also a considerable lack in understanding the
basic concept of stocks and flows in a considerable portion of the students. The results
correspond well with the experiences published by Booth Sweeney and Sterman (2000).
In winter semester 2003/2004 only 34 out of 68 examinees gave fully correct answers to
the bathtub task presented in figure 12 (see also figure 13). Regarding ‘Bathtub
Dynamics’ we plan to apply the same tasks and procedures as proposed by Booth
Sweeney and Sterman to test our future students. As the test group is very
heterogeneous, we hope to contribute a very interesting sample to the Bathtub Research
(cf. Kapmeier 2004)
Another misunderstanding often heard when discussing with students, is the opinion,
that the ability to operate STELLA already makes a professional modeller and problem-
solver, while we try put across, that such software is nothing less but nothing more than
a very helpful tool. Despite such misunderstandings of some students, we experience a
rising demand for SD issues from highly motivated students, who consider to do their
17
master thesis using SD methods. As personal resources are limited, we cannot satisfy
the accelerating demand at the moment by only the small number of modellers, which is
currently teaching them. Due to the high degree of interdisciplinary in forestry, we see a
great potential to enlarge the group that applies SD and modeling methods in their
special field of teaching.
4: Popul total 2 Income total 3: Agric erop 4: Fruit
1 850-5
2 a
3 400.00
4 200.00
1
2
3
1
2
3
4
1985.00 1993.00 2001.00 2008.00 2017.00 2026.00
IPage 1 Years
Village Community Performance
Fig. 11. Performance of the agroforestry village community with high birth rate policy.
Flows (Units/hour)
le a a a
ab A pe
oume ff
flap LE NP el
OA UA eee
EE
io
of
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
hours
Fig 12: ‘Bathtub Task’ used in an exam in winter semester 2003/2004: Consider a stock
with an inflow and an outflow. Calculate and draw the values of the stock for each time
step. The initial value of the stock is 100 units.
18
Stock (Units) ‘Stock (Units)
012.3 45 6 7 @ 9 1011 12 13 14 15 1617 18 19 20
O12 es 4S BT ee WN RBM 6 1 1 OD
hours fee
Stock (Units) Stock (Units)
D1 2a 4 FOTO OR D4 1:0 17 8 19
Page eTAS HHH UIST eo
Fig. 13: Typical student’s responses to the bathtub task given in Fig. 12. Top left is the
correct solution, the other solutions are errorneous.
5 Conclusions
The concept to put a strong emphasis on systems thinking and system dynamics into the
education of the international MSc program for Sustainable Resource Management
seems to be a successful strategy as other international MSc programs at the University,
which also started in 2001, could not gain the same popularity and student enrolments
as SRM. From our point of view, SD and integrated modeling are very valuable tools
for the students and also for the later professionals. We see many promising possibilities
and also the need for extending its application in teaching at our school, in forestry as
well as in resource management. Clearly, progress can be only made step by step. We
are neither competent to teach all subjects -we can only show obvious and possible
interdisciplinary links- nor would our teaching time budget allow for that.
We are rather counting on a market process: our teaching must simply be very good in
order to raise student’s demand for getting more in contact with SD and related
modeling in other subjects. This should, on the long run, have effect on the supply side.
A promising step has just been done: the bachelor programme in forestry is currently
being revised and a whole teaching module has been dedicated to Systems Analysis and
Modelling, partly due to the successful implementation of systems thinking and system
dynamics in the SRM programme. We also hope, that, on the long run, our efforts
contribute to increase the international community of highly qualified systems thinkers
and system dynamicists. We see ourselves at the beginning of a promising process and
would appreciate to get advice and support from the System Dynamics Community.
19
6 Literature
Assmann E. 1961. Waldertragskunde. BLV Verlagsgesellschaft Miinchen Bonn Wien.
490 p.
Biber P, Moshammer R, Pretzsch H. 2001. Prognose des Fichten (Picea abies)-
Starkholzaufkommens auf der Basis von Inventurdaten. Kongressbericht zur 60.
Jahrestagung des Deutschen Forstvereins, Dresden, 20.9.-23.9.2001: 275-284.
Biber P, Weyerhaeuser H. 1998. Numerical Methods for characterizing Structure and
Diversity applied to a natural Tropical Forest and an even aged Teak stand. In
Mies, E. 1998 (Ed.): Natural and Socioeconomic Analysis and Modelling of
Forest and Agroforestry Systems in Southeast Asia. Publication by the German
Foundation for international Development, Zschortau, Germany: 83-104.
Booth Sweeney L, Sterman JD. 2000. Bathtub dynamics: initial results of a systems
thinking inventory. System Dynamics Review 16(4): 249-294.
Carlowitz HC. von. 1713. Silvicultura oeconomica oder haufwirtschatfliche Nachricht
und naturgemaBe Anweisung zur wilden Baumzucht. J. Fr. Braun, Leipzig
(Reprint, TU Bergakademie Freiberg, 2000, p 414.
Chertov OG, Komarov AS, Karev GP. 1999. Modern Approaches to Forest Ecosystem
Modelling. European Forest Institute Research Report 8. Brill, Leiden, p 116.
Deutsches Nationalkomitee MAB. 1987. Ecologic-Socioeconomic System Analysis and
Simulation: A Guide for Application of System Analysis to the Conservation,
Utilization and Development of Tropical and Subtropical Land Resources in
China.
Grossmann WD, Schaller J. 1986. Geographical Maps on Forest Die-Off, Driven by
Dynamic Models. Ecological Modelling 31: 341-353.
Grossmann WD. 1991. Ergebnisse der Anwendung von dynamischen Karten
("Zeitkarten") im Lehrforst Rosalia. (Results of the application of dynamic maps
in the demonstration forest Rosalia). In: Centralblatt fiir das gesamte
Forstwesen. 108. Jahrgang, p. 215-235.
Grossmann WD, Mayerhofer E, Kasperidus HD, Eberhardt S. 1993. Dynamische
Modelle zur 6konomisch-dkologischen Zukunftsgestaltung - Dokumentation
von fiinf komplexen Modellen. Umweltforschungszentrum Leipzig-Halle
GmbH, Leipzig, 871 Seiten, unpublished.
Grossmann, W.D., 1994: Socio-economic ecological models: criteria for evaluation of
state-of-the-art models shown on four case studies. Ecological Modelling 75/76:
21-36.
Grossmann WD. 1998 Strategies for tropical forestry, agriculture and rural areas to cope
with the information society, globalization and sustainability. In: Mies, E. (Ed.)
1998: Natural and Socioeconomic Analysis and Modelling of Forest and
Agroforestry Systems in Southeast Asia. German Foundation for International
Development, Food and Agriculture Development Centre: 105-183.
Hartig GL. 1795. Anweisung zur Taxation der Forste, oder zur Bestimmung des
Holzertrages der Walder. Giessen.
Hartig GL. 1820. Lehrbuch der Forster und die es werden wollen. 6. Aufl.
J.G.Cotta’sche Buchhandlung, Stuttgart und Tiibingen.
Heyer C. 1841. Die Waldertrags-Regelung. Verlag von B.C. Ferber, Giefen.
Hundeshagen JC. 1828. Encyclopadie der Forstwissenschaften , 2. Aufl. H. Laupp,
Tiibingen.
20
Kapmeier F. 2004: Bathtub Dynamics at Higher Management EducationInstitutions in
Stuttgart. Proceedings of the2004 International System Dynamics Conference.
System Dynamics Society: Albany, NY.
Lindner M, Cramer W. 2002. German Forest Sector under Global Change: An
Interdisciplinary Impact Assessment. Forstwissenschaftliches Centralblatt, 121,
Supplement 1: 3-17.
Mies E. (Ed.) 1998. Natural and Socioeconomic Analysis and Modelling of Forest and
Agroforestry Systems in Southeast Asia. German Foundation for International
Development, Food and Agriculture Development Centre, p 183.
Natke HG. 2003. Introduction to Multi-Disciplinary Model-Building. WIT Press,
Southampton, p 379.
Ostmaston FC. 1968. The management of forests. George Allen and Unwin Ltd.
London.
Pommerening A, Biber P, Stoyan D, Pretzsch H. 2000. Neue Methoden zur Analyse
und Charakterisierung von Bestandesstrukturen-New Methods for the Analysis
and Characterization of Forest Stand Structures, Forstwissenschaftliches
Centralblatt, 119: 62-78
Pretzsch H, Biber P, Dursky J. 2002. The single tree-based stand Simulator SILVA:
construction, application and evaluation. Forest Ecology and Management 162:
3-21.
Pretzsch H, Dursky J. 2002. Growth reaction of Norway Spruce (Picea abies (L.) Karst.)
and European Beech (Fagus silvatica L.) to Possible Climatic Changes in
Germany. A Sensitivity Study. Forstwissenschaftliches Centralblatt, 121,
Supplement 1: 145-154.
Repenning NP. 2003. Selling system dynamics to (other) social scientists. System
Dynamics Review 19(4): 303-327.
Speidel G. 1972. Planung im Forstbetrieb. Verlag Paul Parey. Hamburg und Berlin. 267
p-
Sterman JD. 2002. All models are wrong: reflections on becoming a systems scientist.
System Dynamics Review 18 : 501-531.
Back to the Top
