Case Study #3 Oslo - Urban Tree Valuation Modelling

By: David N. Barton
WWW: Martin Karlsen & Anders L Madsen

City trees constitute perhaps the single most important green structure in addition to open green space in cities. City trees potentially provide a bundle of ecosystem services. Oslo Municipality has a strategy to replace every municipal tree felled within the built zone with at least one new one. Damage to trees on municipal land is subject to a fine and a compensation claim using an established liability assessment model for city trees.

VAT16-BBN

This liability or compensation value for city trees is calculated using the so-called VAT16 assessment model based on VAT03 developed by [Randrup, 2005] and adapted to a BBN by [Barton et al, 2015]. The VAT03 assessment model takes as a basis the replacement cost of a city tree, including purchase and planting costs. This base value is then adjusted for the tree.s structural health and for its qualities in a neighbourhood context, including adaptation and contribution to its local environmental. Environmental qualities include aesthetics, noise and pollution reduction, in other words several regulating ecosystem services. An overview of the VAT16 model structure is provided in Figure 1.

Map Interpretation

Each yellow circle on the map represents a tree based on LiDAR data. When the left mouse button is pressed on a yellow element, the value of the selected tree is computed and added to a layer on top of the trees. This layer has a circle for the selected element with color the defined by the value of the tree as computed using the model in Figure 1. The intensity of the color is determined by the expected tree compensation value. If the user wants to calculate the sum value of a number of trees, the area over which tree value is to be aggregated is selected by tracing the polygon using the mouse (while pressing shift).

The intensity of the chloropleth color is determined by the expected tree compensation value (normalized). An expected value of zero produce a transparent placemark whereas the value one (or above) produce a placemark with color green at full intensity.

Full compatibility of the maps can be viewed with IE11 or Firefox 33.

Model Usage

Below are some HUGIN widgets for interacting with the model shown on the right (click on the probability bar to instantiate a node or remove evidence). Notice that Dashboard contains a lot of variables that can be instantiated.

Associated with each tree is information on properties of the tree. The tree compensation value is computed by propagating this information in the Bayesian network shown in the figure above. The intensity of the green color is determined by expected tree compensation value. The more intense the color is, the higher the expected value of the tree is.

The Compute button computes the sum of the expected values of the selected trees on the map, i.e., the trees selected so far. The set of selected trees can be reset (and all evidence removed from the models) by pressing Reset.

The Save data button allows the user to download a file with the information entered using the widgets for a single tree.

---

---

Total expected tree compensation value of selected trees:

Selected tree:

References

Useful references for those interested in BBN include:

[Kjærulff, U. B. and Madsen, A. L. (2013)] Bayesian Networks and Influence Diagrams: A Guide to Construction and Analysis. Springer, Second Edition.

[Randrup, T.B. (2005)] Development of a Danish Model for Plant Appraisal. Journal of Arboriculture 31(3): May 2005.

[Barton, D.N. (2015)] Barton, D.N., Stange,E., Blumentrath, S., and Traaholt, N.V. (2015) Economic valuation of ecosystem services for policy. A pilot study on green infrastructure in Oslo. NINA Report 1114. 77 pp.

Contact information

For further details contact: David Barton (David(dot)Barton(at)nina(dot)no)

---