Intergroup Discussions
Stand Variability

Group - Seppo, Erik, Chris, Uva, David, Marianne, Juan

The Group discussed some of the issues relating to stand variability and what aspects should be considered within STORMS.

Variability is likely to be important because a number of the STORMS models are trying to assess the RISK of damage, and clues as to the level of risk, may be obtained from establishing relationships between homogeneity, patterns and processes within the forest stand. It is not yet clear as to whether or not variability is in itself significant in determining levels of risk or rather a reflection of processes within the stand and thus can be used as a warning of changed levels of risk.

Each parameter, within each model, is subject to variability, at the tree, stand and regional level, however, the sensitivity to variability is not always known.

Key: W=Wind, S=Snow, F=Fire, s=stand, t=tree

Elements of the discussion are summarised within the Figure above, which attempts to categorize some of the modelling efforts with respect to our knowledge of the inputs to areas of modelling within STORMS (but should not be taken as a definitive statement on the quality or effectiveness of the models); sensitivity to aspects of variability; levels of stand variability.

For example:

Ws, Fs - the modelling of wind damage and fire damage at a stand level is sensitive to variability within the stand but the knowledge of the interactions between sensitivities and variability are not well known.

WSs, WSt - the sensitivity of the modelling of damage due to combined wind and snow at a stand and tree levels, with respect to variation in measured variables, is relatively well understood. The sensitivity to variability of damage at the stand level are low compared to those at the tree level.

Requirements:

1. An assessment will be required of what measures of variability are most appropriate as inputs to the models of forest damage and at what scales and resolutions each parameter should be assessed. Options include: distance from stand edge, absolute tree height, canopy variability, species, crown diameter v stem diameter, tree character.

2. A reference level and spatial unit are required against which variability can be assessed. Such references will be dependent, in part, upon the method of measurement and some options are listed below:-

spatial units:

compartments or sub-compartments; soil type; some form of landform unit; interpreted patterns in forest canopy reflecting landform and forest management.

reference level:

absolute tree height; statistical measures at a global level (eg. differences in tree height with respect to minimum, mean or maximum canopy level or dominant tree height within a chosen spatial reference unit); distribution of the ratio of crown diameter to stem diameter within the spatial reference unit; local variation in statistical measures (at the resolution of the data, eg. cell by cell measure of variance in tree height).