Using GIS-based multicriteria evaluation and path optimization for effective forest field inventory
Introduction
In Finland, the information needed in forest management planning in non-industrial private forests is traditionally collected by means of the standwise inventory method. In this method, the planning area is divided into homogenous subareas called forest stands, and the information related to the properties of the growing stock and soil is gathered separately on every forest stand through field inventory. The cycle of holding-level forest planning is 10–15 years, meaning that most of the non-industrial private forests will be subject to one field inventory during that period.
Recently, a lot of attention has been directed to increasing the cost effectiveness of holding-level forest planning (Hujala et al., 2007, Tikkanen et al., 2006). Because the field inventory is the most expensive phase of the planning process, new, improved methods for data collection are being sought. Currently in Finland, a new forest inventory system concerning privately owned forests is under development, in which the information about growing stock is obtained primarily through airborne laser scanning and digital aerial photography, growth models, real-time database updates carried out in connection with forest management operations, and precisely allocated field inventories.
In the current holding-level inventory and planning system, the forest planner visits all the forest stands in every forest holding, whereas in the new system the planner is only expected to visit the most important or difficult stands. These stands are the ones where it is not possible to acquire the needed information with acceptable accuracy level through methods based on remote sensing, or where some additional information is required, for example, for multiple-use forestry or roundwood trade. Implementation of the new system generates some problems: first, a definition problem concerning the actual necessity of inventorying each stand, and second, a decision problem concerning the appropriate way of selecting the stands for inventory when an area is subject to field inventory.
One possibility to address these problems is the use of multicriteria evaluation methods (MCE) together with Geographic Information System (GIS) techniques (Carver, 1991, Jankowski, 1995). In this approach, different objectives for field inventory can be presented as decision criteria and be weighted according to the relative importance they have to the final selection of stands. Deriving the appropriate weights and combining the evaluations according to the different criteria can be assisted by utilizing expert knowledge modeling. Especially in situations where objective information and applicable models based on empirical data are inadequate or non-existent, or the decision criteria are measured in different units, the methods of expert knowledge modeling have been found very useful (Kangas et al., 1993, Store and Kangas, 2001).
In addition to the variability in the stand-specific properties pertinent to inventory importance, the decision process is further complicated by the spatial distribution and accessibility of the stands. As fewer stands need to be inventoried in the field in the new system, the length and duration of a field inventory tour in proportion to the number of stands inventoried is expected to increase substantially compared to the traditional inventory practice. However, not only does the decision to include a stand into a field inventory tour depend on the individual importance and accessibility of the stand in the field, it is also affected by the stand’s location with respect to all other stands and their importance.
Considering the complicatedness of the decision process involving the spatial dimension, GIS-assisted planning based on path optimization can be expected to add value to field inventory planning in the new forest inventory system. In the context of GIS, path optimization is based on the notion of graph theory, where connectivity between discrete locations is modeled in terms of a connectivity graph structure consisting of zero-dimensional nodes and one-dimensional edges connecting pairs of nodes (Chou, 1997). If each edge is assigned a weight denoting the impedance of movement through the particular edge, an optimal path can be calculated between any locations within the connectivity graph using a graph search algorithm (Heywood, Cornelius, & Carver, 1998). The connectivity graph approach has been used mostly in applications related to transportation networks (Waters, 1999), but it can be extended to the cross-country context by representing the traversability of terrain by means of a quantitative traversability measure, discretizing space into a finite set of nodes, and connecting these discrete nodes by weighted edges (Miller & Shaw, 2001). Typically, this amounts to representing the terrain as a raster consisting of a tessellation of square cells of equal size and establishing the connectivity graph by connecting the centers of adjacent raster cells.
Although optimal terrain paths can be determined quite efficiently by utilizing the raster-based connectivity graph approach, in this study the path optimization problem is more complicated, involving a potentially large number of locations. Furthermore, it is influenced by the spatial distribution of important forest stands on the one hand, and the working hours available for a field inventory tour on the other. In the literature, this kind of optimization problem is usually referred to as the orienteering problem (OP), a variation of the widely known traveling salesperson problem (TSP) (Tsiligirides, 1984). In general terms, the OP seeks for the maximum amount of value to be collected subject to a budget limit for the tour (Fischetti, Salazar-González, & Toth, 2002). As concerns forest field inventory planning, this would signify designing an inventory tour encompassing a subset of stands generating the maximum attainable overall utility in terms of inventory importance within a prescribed amount of time.
This study presents a new method for effective standwise allocation of forest field inventory based on expert knowledge. The method combines different objectives set to the forest field inventory by means of multicriteria evaluation, and integrates the spatial dimension to the final selection of stands to be inventoried by applying a set of path optimization techniques. The method is demonstrated through a case study, in which a field inventory importance map and effective inventory tours are produced for the case study area.
Section snippets
Basic steps of the method
The forest field inventory allocation method proposed in this study consists of two main components. The first is the production of the importance map for forest field inventory, based on the objectives and requirements associated with the field inventory in the new forest stand inventory system. Second, the spatial dimension is incorporated into the allocation process by using path optimization methods to take the location of different stands into consideration. The path optimization procedure
Case study area and data
The field inventory allocation method proposed in this study was demonstrated and evaluated by applying it to a case study area in the municipality of Kuortane, located in western Finland (62°48′N and 23°30′E). The reason for selecting this particular location as the case study area was based on it being a testing site for many currently ongoing forest management planning development projects. By virtue of this, adequate data for carrying out the evaluation were available. The principal data
Conclusions and discussion
The forest field inventory allocation method developed in this study for holding-level forest planning seeks to provide an effective way for selecting the forest stands to be inventoried in the field. In the context of this approach, the different objectives and requirements concerning field inventory are described in the form of decision criteria. Determining the criteria is an extremely important part of the method, because the criteria selection has a crucial effect on the results. The
Acknowledgements
This research was sponsored by the Ministry of Agriculture and Forestry under the project titled “Continuous updating of uneven-aged forest resource information in private owned forests – the utilisation of the functional information of various enforcers in the macro area, 310346” and by the Finnish Forest Research Institute and the University of Oulu. We also wish to thank the Forestry Centre Etelä-Pohjanmaa for their cooperation and the data provided to be used in this project.
References (32)
- et al.
A fast and effective heuristic for the orienteering problem
European Journal of Operational Research
(1996) - et al.
Aggregating individual judgments and priorities with the analytic hierarchy process
European Journal of Operational Research
(1998) - et al.
An efficient four-phase heuristic for the generalized orienteering problem
Computers and Operations Research
(1991) A scaling method for priorities in hierarchical structures
Journal of Mathematical Psychology
(1977)- et al.
Integrating spatial multi-criteria evaluation and expert knowledge for GIS-based habitat suitability modelling
Landscape and Urban Planning
(2001) - et al.
Applying cognitive mapping approach to explore the objective-structure of forest owners in a Northern Finnish case area
Forest Policy and Economics
(2006) - Blum, A., Chawla, S., Karger, D., Lane, T., Meyerson, A., & Minkoff, M. (2003). Approximation algorithms for...
Getting to know ArcObjects
(2003)Integrating multi-criteria evaluation with geographical information systems
International Journal of Geographical Information Systems
(1991)Exploring spatial analysis in geographic information systems
(1997)
What is ArcGIS?
The generalized traveling salesman and orienteering problems
A multifaceted heuristic for the orienteering problem
Naval Research Logistics
An evaluation of lattice solutions to the problem of corridor location
Environment & Planning
An introduction to geographical information systems
Cited by (12)
Routing model applied to forest inventory vehicles planning
2020, Computers and Electronics in AgricultureCitation Excerpt :Studies on route optimization in forest inventory planning are scarce in the literature. Store and Antikainen (2010) propose the best selection of stands to be inventoried in a specific area of interest in a prescribed amount of time determining the best physical path in the field for carrying out the inventory using GIS-based multicriteria evaluation and path optimization. The most recent study of Etula and Antikainen (2014) describes a general solution procedure for the Areal Inventory Problem (AIP), and introduces an implementation of a heuristic algorithm that can be used to solve a real-world AIP within a reasonable time frame.
Analyzing uncertainties and estimating priorities of landscape sensitivity based on expert opinions
2017, Landscape and Urban PlanningCitation Excerpt :Their importance can be measured by weights. To produce these weights, expert knowledge modelling has been utilized especially in situations in which objective information and applicable models have not been available or sufficient, or the miscellaneous decision criteria coming from many sources and with incommensurable units of measure have been observed to be difficult to value (e.g. Kangas & Leskinen, 2005; Store & Antikainen, 2010; Store & Kangas, 2001). In expert knowledge modelling, the weights can be produced by an aid of pairwise priority comparisons of the criteria.
Open geo-spatial data for sustainable forest management: Lithuanian case
2020, Research for Rural DevelopmentHeuristic methods
2015, Advances in Geographic Information Science