WIS.2

General System Description

WIS.2 logo

Brief Overview

WIS.2 is a decision support system for monitoring and ensuring the sustainable, goal-oriented and efficient management of forest ecosystems, especially with regard to the:

• integral handling of significant spatial- and temporal scales in forest ecosystems (i.e., multiscalarity, sustainability);
• valorisation of intrinsic multiple ways of using the forest (i.e., multifunctionality, adaptability);
• implementation of a liberal and pragmatic view of silviculture as applied in Switzerland (i.e., flexibility, close-to-nature).

WIS.2, based on MS Access and ArcGIS ArcView, is composed of different applications, each handling a main aspect of the management of forest ecosystems. As a prototype, WIS.2 is currently being tested in three forest enterprises and promoted by a forest engineering firm. The tool is used at all levels of education for forest management in Switzerland.

WIS.2 was developed during 2001-2005 within the framework of a PhD thesis at the ETH in Zurich (Rosset, 2005 ), and has since then been successively developed in an iterative manner using inputs gained from practical experience. Plans are currently being made to refine the tool in a research and development project and launch a commercial product called SMART-Forest.

Scope of the System

Scope of WIS.2

WIS.2 focuses on providing guidance in sustainable forestry management. The system gives strong organisational and decision support in implementing and monitoring silvicultural practices. The system provides specific information for the different phases of the silvicultural process, allowing planners to simulate the mid- to long-term consequences of various action scenarios, and supporting users in solving complex silvicultural problems at multiple scales (i.e., from the whole enterprise level down to the single stand level; from short-term to very long-term). The scope of WIS.2 covers:

• defining different variations of silvicultural strategy and implementation concepts, in order to assess the engineering strategy and recognize the mid to long-term consequences;
• formalising the manifold demands of society, forest owners and stakeholders for implementation in forest management;
• pointing out the priority-of-action by connecting long-term strategic planning with mid-term harvesting planning;
• systematically documenting and reporting the decisions taken with regard to the managerial plan;
• facilitating regular control of the development of forest areas with regard to the realisation of the enterprise strategy;
• giving guidance on how to react in a flexible way to changes such as market fluctuations and unforeseen events like storms;
• enabling the user to handle the given complexity - step by step.

System Origin

As multi-purpose, near-to-nature silviculture seeks to steer forest dynamics towards socio-economic and conservation goals through minimal anthropogenic intervention, the necessity for a powerful forest management decision support system that implements and monitors such silvicultural practices is getting stronger. In order to face this challenge, WIS.1 was initially designed in the 1990s by Good and Pistor (1992) as an information system for silvicultural planning at stand level. Under the name of WIS.2 the system subsequently evolved from an information system to a DSS by integrating knowledge and supporting the entire decision process of the management of forest ecosystems. Over the past few years, further adaptations have been subsequently made, such as e.g. during a 10-day course on silvicultural planning held at the ETHZ in a small forest enterprise (150ha) owned by the Canton of Zurich.

WIS.2 has real-life application scenarios on the level of forest management, education, as well as in research and development.

Forest management

In forest management the tool has been extensively applied in the:

• State forest of Teufen in Canton Zurich (i.e., the 10 day course);
• teaching and research forest of the ETH in Zurich;
• forest of the city of Winterthur[1];
• forest of the city of Rheinfelden[2];
• forest of the village of Bullet in Canton Vaud (diploma thesis);

Plans are currently being made to apply the tool in a forest in Canton Freiburg, as well as in a forest in Thailand (diploma thesis).

Education

WIS.2 is being applied at all levels of the education in forest management in Switzerland in order to illustrate and demonstrate forest planning.

The main motivation behind introducing WIS.2 in teaching is to:

• enhance awareness of the consequences of very slow tree growth by visualising long-term trends resulting from decisions taken (i.e., the necessity to ensure certain continuity);
• concentrate on concept building to derive urgency of intervention to be undertaken at stand level (i.e., how to steer forest dynamics in order to reach the production target as efficiently as possible) and subsequently visit each individual stand separately;
• elaborate on and compare different variations in forestry management.

Students who have used the prototype have been providing valuable feedback that is in an iterative manner to improve the IT-solution. Research and development: a further tool based on the same platform (WVK) is currently being developed[3].

Research and Development

WIS.2 is not yet a commercial product, however, a proposal is currently being written in order to be able to refine the system within the framework of a research project and launch a commercial product through collaborative partnerships under the name of SMART-Forest.

Support for specific issues

WIS.2 takes a multi-level, top-down, situation-oriented approach and provides foresters with room for maneuvering when specifically determining the most appropriate silvicultural techniques to use. The recommended forest strategy is subsequently implemented within the framework of sustainable forest development.

Silviculture in Switzerland

The liberal and pragmatic view of silviculture in Switzerland implies that the suggested strategy should leave as much room as possible for maneuvering when implementing the planned intervention.

Formalisation of the manifold demands of society and forest owners in silvicultural profiles

In order to meet the demands of society and forest owners, the objectives with regard to the desired status of the stands (what to reach), the given conditions of the location (where to intervene) and the related spatial organization (how to coordinate the procedure) need to be clearly specified. In WIS.2, this specification is described in the form of silvicultural profiles that define the desired status of the stands under the given local conditions. Silvicultural profiles are comprised of criteria (e.g., types of trees) and indicators that further specify the criteria (e.g., oak, fir, maple). The grading of the criteria is given at three levels: very important/ very appropriate, important/ appropriate, less important/ less appropriate.

Capabilities to support the decision-making process

WIS.2 supports silvicultural decision-making phases by:

• providing valuable information for the different phases of the process, and thus allowing planners to simulate mid- to long-term consequences in varying action scenarios;
• defining different variations of silvicultural strategy and implementation concepts;
• assessing the feasibility of action scenarios and revealing the long- and mid-term consequences;
• pointing out the priority of action by connecting long-term strategic planning with mid-term harvesting planning;
• giving guidance on how to react in a flexible way to changes such as market fluctuations and unforeseen events.

Data and Data Models

Typical spatial extent of application

The typical scale of use for the application is the forest enterprise that is comprised of forest stands. The average size of stands range from one to several hectares, and the size of enterprises range between a few hundred to several thousand hectares.

Forest data input

The following forest data are required as input to the system:

• Stand information (i.e., regular or irregular stand structure, development stage, tree species composition, cover density, hdom, Ddom, m3 per ha, level of coverage, location, stand goal, quality, sensitivity, origin, natural rejuvenation, biodiversity, measures to take, urgency);
• Phytosociological composition of the stands;
• Description of the relief such as exposition and slope;
• Information on the road network (i.e., accessibility);
• Tree species equilibrium;
• Necessity for rejuvenation of the stand;
• Categorisation of forest societies;
• Projection of demographic structure;
• GIS Layers (i.e., raster and vector files).

Models

Forest Models

Forest models that are taken into account with WIS.2 are:

• Forest growth models that are implemented to determine production goals and concepts, rejuvenation politics for steering the demographic development, and urgency of action in the stands with regard to the current state and future milestones;
• Intervention concepts for e.g. tending and thinning (i.e., when to intervene during the stand life, dominant dhb, how to reach the production targets efficiently);
• Demographic models for e.g. comparison of current age structure and demographic equilibrium, harvested area per decade and over maturing degree of the harvested stands; development of the overall tree species composition (i.e., visualization of how slowly changes occur);
• Growth functions for e.g. planned intervention for the next 10 years (i.e., overview of what, where and when);
• Expert recommendations on the scope of action for e.g. proportion of softwood or other phytosociological units.

Decision Support

Definition of management interventions

WIS.2 takes a multi-level, top-down, situation-oriented approach that provides the forester with enough room for manoeuvring when determining which silvicultural techniques to use. WIS.2 enables structuring the process of decision-making by indicating the necessary information for the different stages of the process, allowing planners to simulate the mid- and long-term consequences of different actions. Managers are provided with valuable information on the consequences of different action scenarios.

Typical temporal scale of application

The approach to perform planning with WIS.2 spans from a vision in the very long term for a whole forest area (entrepreneurial strategy) to interventions undertaken in the short term at stand level. Silvicultural measures taken with WIS.2 at stand level are typically defined for a planning period of 10 years.

Types of decisions supported

As a silvicultural concept, WIS.2 offers a coherent solution at a strategic and tactical level to reach previously-defined targets. Because trees are long-living and slow-growing organisms, it is necessary to take long-term decisions at the strategic level for the whole forest area (or a part of it) and to define general guidelines. The imbrication of stands in a large and heterogeneous mosaic, for example, means that priorities must be defined and measures co-ordinated at the tactical level between 10 to 30 years).

Decision making process and models

The decision making process is an iterative process. First, the user analyses the necessary input to make a decision. This input consists of information and explanatory models. After making the decision, the user analyses its feasibility by comparing it to the present situation and to the targets defined in the previous decision steps. The impact of the decision is simulated using a decision model. The next interface consists of the outputs of the simulation. The decisions taken are recorded at the end, along with comments. In addition to the principle that it must be the user who takes the decision, the following interaction principles have been formulated:

• Information is useful to the user if he has easy access to the meta-information (precision, actuality, method of collecting the data);
• Decision variables must be introduced to the system by the users (no standard variables);
• Interpretation of the simulation outputs is left up to the user,
• Models are a simplification of reality. They are only useful if the user knows about the hypothesis and the validation area of the model;
• Decision models have many variables so that the user can understand the simulation outputs.

Output

Types of output

WIS.2 outputs are typically maps and graphs, such as;

• maps showing the assessment of forest profiles with regard to the “usefulness” of each stand;
• maps depicting the importance of certain mid- to long term intervention scenarios;
• graphs indicating the mid- to long term development trends of various compositions of tree species.

Spatial analysis capabilities

The link to GIS enables spatial analysis (e.g., topology overlays, multi layering of different maps, selection of objects based on selection criteria, aggregation of attributes, showing statistics by area, etc).

Abilities to address interdisciplinary, multi-scaled, and political issues

WIS.2 clearly has the ability to address interdisciplinary, multi-scaled- and political issues. The tool utilises semi-quantitative data describing the stands, as well as expert knowledge linked to the (phytosociological units in the) stands. WIS.2 combines wood production with other socio-economic and conservation goals, using a minimum of anthropogenic interventions and taking into account the heterogeneity of site conditions and stand structural types.

System

Architecture and major DSS requirements

Software architecture WIS.2 was developed and adapted on the basis of Microsoft® Access 2000/2007 and ESRI ArcGIS® 8.2/9.3. These two applications were selected because of their widespread usage, their full integration into the Windows system (i.e., COM-technology), the possibility of extending their capabilities by using Visual Basic, (a simple programming language), and the ease with which they can create user interfaces. Although Microsoft® Access is not an OO database, it is possible to implement the object-oriented model. Tables correspond to object classes, records to objects, modules (i.e., collection of program codes) to models (functions) and methods (procedures).

Modularity The system is built modularly with geographic, general and specific thematic data modules, along with interface modules such as the data analysis module or the regeneration concept module. Specific data corresponds to a particular forest enterprise. The main characteristic of an object-oriented (OO) approach is how it combines data and methods (program code) together into an object. The attribute values of an object are accessible and modifiable only through a method, which is assigned to the respective object. Objects are organised in a hierarchical class structure. The main strength of the OO approach is that it builds up a modular model, which is independent of implementation issues. Modularity is a very important characteristic related to DSS in general, since not all future applications are predictable (Power and Saarenmaa, 1994); Schönsleben, 2001). The UML (unified modelling language) was used to design and structure the object model. It has the advantage of being standardised and widespread. This approach helps organise and structure the elements of the decision support, i.e. the necessary information (structured data), knowledge (models) and methods (algorithms).

Scalability WIS.2 has strong scalability capabilities. Experience has shown that the decision model used at the strategic level is relatively simple (age based), but differentiated enough to assess the evolution of tree demographic structure in the very long term. At the tactical level, it is necessary to use dbh growth functions to determine the urgency degree of regeneration and to consider other factors. In general, however, the smaller the time and spatial scale, the more complex and intertwined with other systems the decision models become (e.g., productivity models, which consider more than just harvesting costs and market prices).

Usage

WIS.2 can be used at research-, industry-, as well as at Government levels.

User Interface

WIS.2 comprises an intuitive user interface that does not require any prerequisite knowledge of the system. The user interface is organised in iterative decision steps. First, the user analyses the necessary input to make a decision; this input consists of information and explanatory models. After making the decision, the user analyses its feasibility by comparing it to the present situation and to the targets defined in the previous decision steps. The impact of the decision is simulated using a decision model. The next interface consists of the outputs of the simulation. The decisions taken are recorded at the end, along with comments.