Wind Risk


Wind is one of the largest causes of forest loss in the world, especially in temperate and boreal forests. For example, in the EU alone it accounts for more than half of all catastrophic damage by wood volume to forests (Schelhaas et al., 2003). Hurricanes affecting regions such as the US are particularly devastating. Hurricane Hugo in 1989 damaged more than a third of standing timber in South Carolina, and Hurricanes Katrina and Rita damaged over 2M hectares on the gulf coast (Beach, 2010 link). In Amazonia in 2005, a single squall from a convective storm destroyed 0.3-0.5m trees in the Manaus district, equating to 30% of the observed annual deforestation in that region in 2005 and the second highest deforestation rate in 15 years (Negron-Juarez et al., 2010).

Furthermore, a study following the 2005 Gudrun storm in Sweden, found that the impacts of storms on forests continue long after the actual event. The growth of Norway Spruce forests was found to be reduced by over 10% in the worst hit regions over the following 3 year period (Seidl and Blennow, 2012).

Whilst nothing can be done to reduce the likelihood of strong winds, forest management can have a strong bearing on the impact of such storms. Research to date has revealed that the level of impact depends on a number of key factors that determine the susceptibility of a forest to wind risk. These include: trees species; tree height (and height/diameter ratio); age; crown properties; stand edge; thinning characteristics; root properties; topography and soil properties (Hanewinkel et al., 2011). It is generally agreed that tree height is the most important variable governing susceptibility to wind. Other factors vary in importance depending on the study/model. Conifer species in general are more vulnerable than broadleaves but since local conditions vary and species adapt to their surroundings it is difficult to generalise on species vulnerability (Gardiner et al., 2011 Link).

Climate change adds an additional challenge to predicting wind loss. In 2012, the Intergovernmental Panel on Climate Change (IPCC) published a special report link  on managing the risks of extreme events, which concluded that it had very low confidence in the ability of the approaches used in the few studies that exist to predict strong and extreme winds with the exception of some predictions for cyclones. Tropical cyclones, were said to be likely to decrease or remain constant in number but to see an increase in mean maximum wind speed, whereas extra tropical cyclones were said to be likely to have moved pole-wards in the Northern and Southern hemispheres in the last 50 years and there was medium confidence that they would decrease in number and continue to move pole-wards (Seneviratne et al., 2012).

Approaches to assessing risk

To date, we have found that the only UK approach to wind risk assessment of relevance to our audience, is the widely used mechanistic model ForestGALES developed by Forest Research. We therefore focus on this tool.

ForestGALES  Version 2.5 (released October 2015)

Website: link  User Guide: link
For online demonstration, click on ‘Forest DSS’ tab (above right) and select ForestGALES 2.5 from the pull down menu
Primary Contact: Dr. Bruce Nicoll, Forest Research
Other: Professor Barry Gardiner (original developer / recent upgrade to version 2.5).


ForestGALES is a computer-based tool to assess forest wind risk. The tool assists forest managers in developing silvicultural practices to minimise wind risk, and provides return intervals of winds that can cause damage to stands either by uprooting or stem breakage. It does not require historical data and so can be adapted to new forest projects and locations lacking this information. Simplified versions requiring fewer inputs can provide lower resolution, regional scale wind risk assessments.

The latest web based version 2.5 (Nicoll et al., 2015) is available free from the Forest Research web site (link) . The full version is also available to download and costs £50 plus vat for commercial use, but it may be provided free of charge for academic use and for research in collaboration with Forest Research. A research version is also available which allows the adjustment of input parameters and a complete range of output data useful for research purposes. ForestGALES outputs can be read into ArcGIS to provide results in GIS format. An R version of ForestGALES has been developed with  Locatelli to allow easy incorporation in calculations developed in R. All versions are fully documented on the website.

Reducing wind risk through silvicultural practice
ForestGALES was originally developed to assist forest managers in developing silvicultural practices to minimise wind risk. Key data on forest projects can be input and GALES will advise on which stands are at the highest risk.

Identification of return intervals of windspeeds causing uprooting or stem breakage
GALES identifies the critical wind speeds at which trees are subject to uprooting or stem breakage, and provides the expected return period of such winds i.e. average number of years between wind speeds exceeding the critical level. Risk assessments can easily be derived from these outputs. If all of the information is readily available an assessment may be done in half an hour.
Note: Before ForestGALES was developed, the UK forest industry used a system called “Windthrow Hazard Classification” (WHC). WHC scores are provided as a ForestGALES output for comparison.

Impacts/Vulnerability assessment
ForestGALES provides the probability of an average tree being damaged within a stand. It does not, however, estimate the % of the stand that is likely to be damaged but damage to an average tree implies substantial damage within the stand that would normally require the whole stand to be cleared.
Gardiner and Locatelli are investigating whether vulnerability information can be added to the model in future, and are developing a probabilistic version of GALES with expressions to describe the variability. Contact: Gardiner

The full version of GALES typically requires data to be input from field measurements or yield models on a range of tree and stand characteristics as well as upwind edge effects, however, tree height, stem diameter, spacing, and species are the most important of these variables. Provided these data are available for a given investment a reasonable estimation of risk can be provided using default values for the other factors.
The model contains default wind values for the UK but it can be adapted to other countries provided ‘a’ and ‘k’ values for the Weibull distribution of local windspeed can be obtained (wind is assumed to be normally distributed and a is the mid-point of the distribution and k the standard deviation. In general, for different regions, a tends to move but k stays relatively constant).

Countries covered
The model is provided for the UK but has also been adapted for use in New Zealand, Canada (Quebec and British Columbia), Southwest France, Denmark and Japan. If wind data is available (see previous section) GALES could be adapted to other countries.

Lower resolution regional maps of wind risk
Lower resolution regional maps of wind risk to forests can also be derived from GALES. As part of the European MOTIVE project (link) , Nicoll worked with Gardiner and Dr Mart-Jan Schelhaas to produce European maps of critical wind speeds that would produce risk to forest stands across Europe (link to full report including maps). These outputs were derived from ForestGALES using forest structure from the Synthetic European Forest Structure Database, and soil information derived from the FAO soil maps. These results can be combined with past and future climates from the EU ENSEMBLES project to estimate future wind risk to forests (Gardiner et al., 2013). This work is intended to support adaptive forest management in the face of climate change. (Contact: Nicoll/Gardiner)
A simplified version of ForestGALES designed to work at large spatial scales (national or continental) with a reduced input dataset was developed by Dr. Ferenc Pasztor for use in a land-surface exchange model, ORCHIDEE. (Contact: Gardiner).

Single species vs mixed stands/age distribution
The web-based version determines wind risk for a single uniform stand of single-aged single species conifers. The full version, supports analysis of multiple stands and can be run over time to show how wind risk changes as a stand develops. It can support thousands of sub-compartments with different species but not mixed species, mixed age stands, or continuous cover stands although a version able to do this is in preparation. A basic assessment of the risk to mixed stands or continuous cover could be obtained by running GALES for each species separately to find which is the most vulnerable component, and this risk applied to the whole stand.

Yield models containing projected growth information can be input to GALES to model how wind risk will change over the rotation period. Forest yield models are included for the UK, but can be replaced by the user’s own yield models.

Nicoll is investigating the ability for GALES to work with complex stands including those managed for continuous cover for ForestGALES v3.4.
Gardiner is working on a research version of GALES for individual trees that accounts for local competition from neighbours and can handle mixed aged and mixed species stands.

Canopy level versus stand level analysis
ForestGALES normally works at forest stand level but it has been modified to run at individual canopy level too. The model can incorporate estimates of canopy dimensions produced by airborne LiDAR. The result is a more detail view of the variations of risk within a forest stand (Suárez et al., 2014). This new approach has been tested in the Cowal and Trossachs Forest District, using LiDAR data taken in 2008 and 2012. The model has been able to locate most of the areas affected by wind damage after the storm in January 2012. The identification of windthrow gaps allows a better estimation of timber loses and a more accurate production forecast.

Tree species covered
Conifers – GALES was originally developed to assess the risk to commercial conifers. Twelve conifer species are included: Scots pine; Corsican pine; Lodgepole pine; European larch; Japanese larch; Hybrid larch; Douglas fir; Noble fir; Grand fir; Sitka spruce; Norway spruce; and Western hemlock.
Broadleaves –Nicoll is exploring the future incorporation of broadleaf species. In the interim:
Locatelli has been working with GALES to assess Eucalyptus.
Gardiner has added Beech and hopes to add Oak soon.

Simplified GALES for additional species/ages
Wind risk is often similar for uprooting or breakage of trees in storms. Nicoll believes that using tree dimensions (if available), and manuals on the wood properties of trees (including the moduli of rupture and elasticity) around the world, it could be possible to determine critical wind speeds without having to undertake expensive research including mechanically overturning trees down to determine anchorage. In this way look-up tables could be provided which estimated the wind speed at which trees of a range of species, at a range of heights or ages, would be vulnerable. This project could be considered on request if funding is available. Contact: Nicoll

GALES linked to CAPSIS forest model
A version of ForestGALES in a Java library has been incorporated in the CAPSIS forest modelling framework with Celine Meredieu and Thierry Labbé at INRA. It allows calculation of wind risk to stands and individual trees. Contact: Gardiner

Accounting for variability of windspeed during a storm

A version that properly accounts for variability of wind speed during a storm and propagation of damage and storm duration is under development with Sophie Hale, Bruce Nicoll (Forest Research) and Sylvain Dupont (INRA). It is similar to the GALES-BC model developed by Ken Bryne and Steve Mitchell at UBC but uses new calculations of airflow over forests. Contact: Nicoll

Comparison of ForestGALES 2.5 with previous versions
The new version 2.5 reflects the latest science and has full documentation. A comparison of the old Windthrow Hazard Classification system with ForestGALES 2.1 and the new version, ForestGALES 2.5, based on a validation exercise using recorded wind damage in a recent winter storm (Hale et al., 2015) shows that the WHC is the most pessimistic, and ForestGALES 2.5 is the least pessimistic. Version 2.5 provided predictions of damage across a large forest area that were close to that observed. The difference between ForestGALES 2.1 and 2.5 is equivalent to reducing the DAMS (windiness) score by 2 points.


BEACH, R. H. S. E. O. L. T. P. S. 2010. The influence of forest management on vulnerability of forests to severe weather. In: Pye, John M.; Rauscher, H. Michael; Sands, Yasmeen; Lee, Danny C.; Beatty, Jerome S., tech. eds. 2010. Advances in threat assessment and their application to forest and rangeland management, Gen. Tech. Rep. PNW-GTR-802. Portland.
GARDINER, B., BLENNOW, K., CARNUS, J., FLEISCHER, P., INGEMARSON, F., LANDMANN, G., LINDNER, M., MARZANO, M., NICOLL, B., ORAZIO, C., PEYRON, J., REVIRON, M., SCHELHAAS, M., SCHUCK, A., SPIELMANN, M. & USBECK, T. 2011. Destructive Storms in European Forests: Past and Forthcoming Impacts. European Forest Institute, Atlantic European Regional Office – Efiatlantic.
GARDINER, B., SCHELHAAS, M. & NICOLL, B. 2013. Chapter V: Mapping the risk to European forests with a changing climate. In: FITZGERALD, J. & LINDNER, M. (eds.) Adapting to climate change in European forests – results of the MOTIVE project. Sofia: Pensoft Publishers.
HALE, S. E., GARDINER, B., PEACE, A., NICOLL, B., TAYLOR, P. & PIZZIRANI, S. 2015. Comparison and validation of three versions of a forest wind risk model. Environmental Modelling & Software, 68, 27-41.
HANEWINKEL, M., HUMMEL, S. & ALBRECHT, A. 2011. Assessing natural hazards in forestry for risk management: a review. European Journal of Forest Research, 130, 329-351.
NEGRON-JUAREZ, R. I., CHAMBERS, J. Q., GUIMARAES, G., ZENG, H. C., RAUPP, C. F. M., MARRA, D. M., RIBEIRO, G., SAATCHI, S. S., NELSON, B. W. & HIGUCHI, N. 2010. Widespread Amazon forest tree mortality from a single cross-basin squall line event. Geophysical Research Letters, 37.
NICOLL, B., HALE, S., GARDINER, B., PEACE, A. & RAYNER, B. 2015. ForestGALES: A wind risk decision support tool for forest management in Britain. Version 2.5. In: COMMISSION, F. (ed.). Edinburgh.
SCHELHAAS, M. J., NABUURS, G. J. & SCHUCK, A. 2003. Natural disturbances in the European forests in the 19th and 20th centuries. Global Change Biology, 9, 1620-1633.
SEIDL, R. & BLENNOW, K. 2012. Pervasive Growth Reduction in Norway Spruce Forests following Wind Disturbance. Plos One, 7.
SENEVIRATNE, S. I., NICHOLLS, N., EASTERLING, D., GOODESS, C. M., KANAE, S., KOSSIN, J., LUO, Y., MARENGO, J., MCINNES, K., RAHIMI, M., REICHSTEIN, M., SORTEBERG, A., VERA, C. & X. ZHANG, X. 2012. Changes in climate extremes and their impacts on the natural physical environment. In:. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation [Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, Tignor, and P.M. Midgley (eds.)]. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change (IPCC).. Cambridge University Press, Cambridge, UK, and New York, NY, USA, pp. 109-230.
SUÁREZ, J.C. 2014. ‘An individual canopy delineation algorithm based on Object-Oriented segmentation and classification’. Book chapter in ‘Challenges and opportunities for the world’s forests in the 21st century’. Springer.