AAIR Project: Sub Task 7.3 Validation of wind damage model against field data NRS

Participant : The Forestry Authority,

Northern Research Station,

Roslin, Midlothian,

Scotland, EH25 9SY

Contact: C P Quine, B A Gardiner, P D Bell.

Description:

Three windthrow monitoring areas will be used to validate the predictive model formed at task 6, and test the impact on alternative silvicultural strategies by presenting the results to managers of those forest areas.

Inputs:

The predictive model of wind damage from task 6 (MLURI) plus the necessary site, stand and actual damage data prepared by Forestry Authority from task 5.2 needed to run the model.

Outputs:

Probability of damage for forest areas under existing and potential tree cover/silviculture. Assessment of the value of the results by formal comparison with actual damage, and by user test through the forest managers.

Method:

The monitoring areas will be selected by Forestry Authority after consideration with MLURI of environmental dataspace (windiness from FA, wetness from MLURI) and management style ie current silvicultural techniques. Areas used elsewhere in model derivation ie task 5.1 will be excluded. Model from task 6 will be applied to the GIS datasets from Task 5.2. Output will form maps of current probability of damage, maps of predicted damage to date (for comparison with actual damage) and simulations of predicted damage occurrence over the next x years.

Uses:

This task will form an important part of the assessment of the value of the project work, and could be used to identify the potential for the model to be adopted by European foresters.

References:

Sub Task: 7.3

Participant: The Forestry Authority

MODELS USED FOR DATA ANALYSIS

I. Model Name: Task 6

II. Source of Model: This project

III. Computer language: To be determined

IV. Hardware needed: PC-based?

V. Data Inputs needed:

• Digital terrain model

• Crop data

• Wind climate

• Actual windthrow

• Predicted windthrow

VI. Model Outputs

• A. Tables/Statistics

• Correlation between actual and predicted damage

• B. Maps/GIS datalayers

• Probability of damage for current state of forest

• Probability of damage for alternative potential states

• Range of possible outcomes over the next x years

VII. Accuracy/Sources of Uncertainty/Method of Handling:

• To be tested

• Carry over of error from multiple sub-models

VIII. Existing linkages between data and software:

• To be determined

IX. GIS Software used:

• To be determined

X. Modelling Software used:

• To be determined

XI. Database Software used:

• To be determined

XII. Any other comments:

Prepared by C P Quine/B A Gardiner/ P D Bell

May 1995