This document is a report of discussions which were held on October 31st 1995. It also includes information from contacts who were not present at the meeting, the results of subsequent discussions and further information gained from the literature. I would appreciate some input from participants in the form of corrections, comments on misrepresentations, and additions if it is felt that I have overlooked important aspects. These changes and additions may be things which were discussed at the meeting, or may be the result of subsequent reflections.

1. What are the mechanisms by which snow causes damage to trees and forests?
2. Considering the data and knowledge available, what methods might be most appropriate for developing a model?

STRUCTURE OF THE DISCUSSION
The discussion focused initially on the mechanisms by which damage occurred due to snow and wind, and some concern was also given to the silvicultural aspects of this damage. Discussion then moved to the climatological parameters which might be important, to identify which needed to be understood and modelled. This discussion also considered the extent of damage incidence both temporally and spatially and the issues associated with scale and variability of climate events with respect to the meteorological records. The issues and conclusions summarised below encapsulate the general content and results of these discussions. FACE="CG Times"
CIRCUMSTANCES OF SNOW DAMAGE AND RELATED SILVICULTURAL ISSUES
Mechanisms of damage
Snow damage can be manifested both as breakage of tree stems and/or branches and by over-turning. Some of the information relating to this discussion has been supplemented by further communications from John Moncrieff and Douglas Malcolm (Forestry Department, Edinburgh University). There are several mechanisms operating which cause damage:

a) wet snow simply causing overloading of the tree, particularly in the case of unbalanced canopies which will produce preferential loading on the side of the tree with the most canopy, causing snap.
b) wet snow sticking to one side of the canopy caused by weather coming in from one direction and causing snap.
c) frost 'socketing' in which branches frozen by drifting snow are wrenched off under their own weight when the snow melts below them.
e) ice accretion on branches, again causing them to drop off prematurely.

Silvicultural Practices
These mechanisms have distinct links with silvicultural practices. For example the lop-sided growth of trees can result from a tree's position in a stand and will affect its vulnerability to damage since it will cause preferential loading on one side, even by vertically falling snow. Damage to branches only, may preserve the rest of the trunk and allow some quality wood from the tree to be salvaged. However, snapped branches expose the tree to attack by fungal growth and disease. There is experience of snow damage for nearly all ages of trees, although the type and severity of damage differs. Some species (e.g. lodgepole pine) are extensively damaged when young (< ten years old) but this is partly a feature of the very heavy crown/basal sweep and poor roots of a provenance of this species known as the 'South Coastal'. As regards silvicultural issues, stocking undoubtedly contributes to the situation and damage to older trees is of most sig nificance to managers. What really matters is the strength of a tree in relation to its height and the size of the crown. If a tree is growing with minimal competition it will produce a strong tapered stem with a large crown, and so trees are managed to reduce the crown by increasing stocking. Small crowns mean fewer branches and therefore an increase in wood quality with fewer knots. However trees with small crowns are less stable and stocking pre-disposes them to damage, therefore a balance has to be struck between stability and wood quality. Thus trees are thinned regularly and often to improve their stability. Most damage (mainly due to wind) occurs after thinning when trees are more vulnerable and less stable. Late thinning (close to winter) in particular increases the risk, but means that trees removed at thinning are larger and therefore of more monetary value.

Spatio-temporal extent of snow damage

Details about the amount of damage which has occurred due to snow and its spatio-temporal incidence is also unknown. This is partly due to the difficulty of distinguishing between damage caused by wind and damage caused solely by snow and may reflect the feeling that wind is of overriding importance and snow damage much less common. Evidence of snow damage is sparsely documented and often anecdotal. Therefore there is a difficulty in quantifying snow damage over time. However in the areas where snow damage has been known to occur the impact has been important. In Scandinavia most snow damage is a result of vertical stresses caused by direct weight of the snow, whereas in the UK, snow damage is invariably the result of wet snow plus strong winds. There are examples of very heavy vertically falling snow causing damage in the UK literature (e.g. The New Forest in the 1930's, Kielder Forest during the 1980's and in December 1990 12% of the forests on the North York. moors were damaged by wet snow and not particularly strong winds of around 25m/s). However this kind of snow damage tends to affect shoots and branches rather than trunks.
Summary of snow damage literature
More detailed information about snow damage has been obtained from document sources. Information about snow damage studies in Scandinavia and North America have also been included. Whilst the tree types and climatology of these areas is different to that of the UK, it is interesting to note their findings, and to consider how relevant this information might be to the UK situation. This summary is by no means exhaustive and probably not even representative of the total of snow damage literature consisting, as it does, of notes made from a few papers.
UK
Damage in North Yorkshire forests in December 1990 is discussed in a paper by Wright and Quine 1993. The storm was linked with the development of a low pressure system on a cold front which was moving south over Britain. The position of the depression, over the North Sea Coast, was different from systems that form in the Atlantic and then move across Britain from the west. An occluded front brought heavy snowfalls to North Yorkshire and a steep pressure gradient to the north of the low produced strong north/north easterly winds. The maximum gusts recorded were not remarkably high, but an estimated 75-80,000m3 of timber was lost. The pattern of damage was analysed. Most damage was thought to be consistent with wind factors, but an unexpected preponderance of damage in pines suggested that snow accumulation strongly influenced the pattern of damage. It is noted that although this combination of strong winds and snow is unusual in Britain it is not unique (e.g. Watson, 1936; Frank 1948; Anon. 1978; Wheeler 1991). Young 1938 and Bollard 1969 also discuss the occurrence of snow damage in the absence of strong winds in the form of branch and top breakage, but little overturning of trees. Cremer (1983) notes that snow damage is often observed to be worse on sheltered sites and on lee slopes than on exposed sites, however in a glazed frost storm the damage was also recorded as most severe on windward slopes (Sanzen-Baker and Nimmo, 1941).
Finland
In Finland snow damage to forests has been examined in relation to the meteorological conditions (Solantie 1994). It was found that the 40mm of snow load caused weak or moderate damage and greater than 60mm produced severe damage. Temperatures at the time of precipitation should be above 0oC and then below 0oC so that the precipitation attaches itself to the twigs and then freezes on. Temperatures exceeding 0.6oC prohibit damage by permitting the snow load to fall. Wind speeds above 9m/s (15m above ground) ensured damage reduction because snow was dislodged snow which was not attached. Solantie notes that there are few statistics either of snow damage or of the relation between the snow and heavy snow falls, but that there is a causal connection between snow damage and heavy snowfalls.
METEOROLOGICAL CONSIDERATIONS OF DAMAGING SNOW INCIDENCE
Magnitude Frequency Issues of Meteorological Phenomena
There are a number of significant issues involved in mapping variables in space and time. It is important to look specifically at what processes are being modelled with what data. Much of the meteorological data is presented as annual means measured at meteorological stations which are 70 to 100 km apart. The data is averaged across space and through time and extreme values and variability, which are the most important events to identify for snow incidence modelling, are lost. It may be that for any given weather event the spatial dependence is less than the distance between weather stations. Work in California using data disaggregated into shorter time periods has shown that there are major spatial dependence changes over a 24 hour cycle. Understanding the relationship between a variable of a particular magnitude and its occurrence will not enable prediction of the return period of the same variable of a different magnitude. It is important to understand the process and know something about its timescale in order to identify appropriate data from which to derive the relationship. For example a very heavy damaging snow fall may go unrecorded if the fall occurs over the space of a few hours and then melts and observations are made just before and just after the snowfall and melting.
There are a number of methods of analysing extreme values. Rodda (1967) has employed a method developed by Gumble (1954) to analyse the annual maximum rainfalls at each station and produce maps showing magnitude of rainfalls for various return periods in the United Kingdom. For data he has used the annual maximum rainfall and annual average rainfall in one day for each station (thus precluding the processing of copious amounts of individual data recordings), and the method shows there is a good relationship between magnitude and frequency even for stations with very large falls. However the method deteriorates where individual maximum falls are far in excess of the other measurements and a meaningful relationship was only obtained when the greatest fall was omitted from the analysis. These outlying extreme values may have an enormous return period not calculable using this data and represent an event so extreme as to be unpredictable. The remaining values were then used for constructing the maps of return periods for events of a particular magnitude. The relationships between average annual rainfall and one-day falls was found to be a good one, improving with a decreasing return period. This means that a minimum amount of information can be used to formulate a relationship between average annual rainfall and maximum one-day falls so that only average annual rainfall data (which is available for many more stations) can be used to construct maps of maximum one-day rainfall for any return period. This also means that recurrence intervals for particular events can be calculated using only data from the area of interest, thus reducing the cost of data and the time required to process large volumes. If trees are not likely to be planted in certain areas, or above certain heights, there is no need to acquire meteorological records for those areas. If it is possible to calculate the recurrence interval for any particular event in this way, there is an equation which can be used to work out the probability of an event of that magnitude occurring for a given time period. In addition to this is the issue that extreme events often happen together so that synoptic events may appear much more frequent if individual parameters are looked at separately.
Snow Damage Occurrence Compared to Synoptic Conditions
No comprehensive study is known to exist which gives an overview of damage events in the UK, and therefore the weather events, the frequency with which they cause damage and the spatio-temporal extent of damage is unknown. It is important to find figures and dates for snow damage in the UK to establish the extent of the problem and the synoptic conditions which gave rise to weather which caused damage.
It is not known exactly what combinations of wind and snow cause damage but it is thought that wet, extremely heavy snow falls cause the most damage. Therefore it may be better to identify whole synoptic damage events, and find a means of characterising a whole event rather than linking large numbers of measured meteorological parameters. It was suggested that this could be achieved using Lamb's classification of weather (Lamb 1950). By characterising the event in this way it may be possible to learn something about exactly what information will be required for a magnitude-frequency, probability analysis. It may be that the question is only that of how often this synoptic condition occurred, rather than looking at the details of the process and requiring temperature and precipitation data. For example heavy prolonged snowfall will probably cause the most damage and such an event can be produced by the progression of a front parallel to itself. It may be that this particular condition is produced by a limited number of synoptic scenarios.
Dan Cornford from Birmingham University contributed further thoughts concerning the snyoptic conditions. Snow and wind can result from two main synoptic situations: polar lows embedded in a northerly air mass, or depressions approaching from the Atlantic against colder air. In the latter situation, snow is likely to turn to sleet or rain as the warm air of the depression reaches further north and east.
Meteorological Circumstances of damage - Snow v Wind
Of particular importance in the UK for this project, is the relationship between snow and wind. Wind incidence is the most important factor causing damage in the UK and is being modelled by the Forestry Commission. Should these parameters be considered separately? Considering work done in other countries (e.g. Solantie 1994 below), and after some discussion, there was a general feeling that it may be possible to treat these two parameters separately for the case of snow damage to trees, if only for simplicity, and because of possible difficulties in determining the relationship between the joint occurrences of snow fall and wind. In the case of wind, the presence of snow would, in most cases, not increase the likelihood of damage. In other words, it would not change the critical wind speed at which damage occurs, so that wind speeds just below the threshold would not be considered more likely to damage trees if snow was also falling because an y snow would be dislodged from the branches by the very strong winds and thereby would not contribute significantly to the load on the tree. However a relationship exists because, in the case of heavy damaging snow falls, high winds might be considered damage reducing, dislodging snow from trees and thus reducing the applied load. There is, however, a counter-argument, that less heavy snow falls could cause damage when combined with wind because snow will stick to one side of the canopy producing a turning moment causing tree failure (mechanism (b) above). The threshold changes depending on the wetness of snow. For dry snow, an increase in wind will greatly reduce damage, but for wet snow accumulation is determined by the amount of wind passing through the tree. Therefore temperature is important.
During this project it will be difficult to undertake any data processing with a view to improving the understanding of the complex interplay between snow, temperature and wind over the timescale being considered given the obscurity, cost and quantity of UK Meteorological Office data. Specific issues include the irregular location of weather stations and the mode of data presentation by the Meteorological Office which is too generalised (through aggregation and averaging) to allow easy determination of the relationship between the incidences of wind and temperature and precipitation. Given that it would be much easier to consider wind and snow as separate climatological issues, and the lack of information concerning how wind and snow operate together to cause damage, it is important to ask meteorologists about the joint occurrences of severe wind and snow conditions. A brief consultation with meteorologists at Birmingham University revealed that alt hough not much information was to hand, the feeling was that severe snow and wind are not strongly correlated. They also suggested contacting Richard Harding at the Institute of Hydrology. In addition, more literature research should be undertaken to establish what is known about the joint occurrences of heavy snow and strong winds and the parameter thresholds determining the operation of these different mechanisms of damage. If their joint occurrence is unlikely, then they can be treated separately.
Approaches to developing a snow incidence model
There is a real issue of balance between the amount of time required to develop a comprehensive and sophisticated UK snow prediction model considering that long-term weather prediction in the UK is treated as a complex area of study by meteorological experts, and the relatively limited importance of snow damage for the UK forest industry compared to wind.
A literature review will be undertaken to consolidate the documentary evidence of snow damage to ascertain its extent, location, frequency and any information about causes and associated meteorological conditions. This will establish the importance and extent of snow damage in quantitative terms, provide an indication of the annual loss to the Forestry Industry in the UK, and give a more comprehensive picture of the circumstances of snow damage over time in the UK. If it transpires, as is expected from this meeting, that loss of trees through snow damage is very small, and that records of damage are too infrequent to allow any derivation of the occurrence of damaging meteorological events, then development of a model to predict snow incidence would be considered unnecessary and damage events too infrequent to be predictable.
Should snow damage prove to be a significant factor, modelling will be required, and if this was the case, it was thought best to adopt a method which was crude but complete, rather than a more refined method restricted to smaller areas and particular circumstances. The approach which is currently being pursued is that discussed above. That is, the identification of records of damage events in the literature, followed by an examination of the associated synoptic conditions from weather maps and reports (e.g. newspaper reports) to try to establish a relationship between particular synoptic scenarios and damage-causing weather. If further work is needed it may be useful then, to identify further examples of such synopses to test how likely the occurrences are, and how likely they are to be associated with damaging weather. If this proves successful, the spatial component could be introduced via parameters such as topographic (height versus temperatur e relationships) and geographic variables to produce multipliers and ways of mapping areas affected by snow, and of the snow causing damage. Approaches to this should incorporate the magnitude-frequency, scale and variability issues discussed above. Another approach might involve mapping snow lie (using, for example AVHRR data) for an extremely bad winter as a worst case scenario and using this information to map spatial variability of snow fall. This technique would provide a crude means of mapping between weather stations. The drawback with this technique would be the amount of time and data required to undertake this mapping for the whole UK. Lack of adequate coverage and cost are likely to be considerable issues for this approach. However, it may be possible to derive information from Met. Office Snow Reports which would give some idea of the range of locational and topographic variability. These reports give an indication of the number of days of snow lie for ten different stations. However, how much this may just reflect the process of ablation is uncertain and therefore they may not be useful.
Given the difficulties of modelling caused by a deficiency of relevant data, the infrequency of damage events, the difficulty of identifying events in meteorological terms and the lack of detailed records for events which are known to have occurred, a statistical or quantitative modelling approach may be inappropriate and unworkable. Therefore another approach could involve risk assessment in terms of expert opinions and rules which could be gleaned from the literature. These could include meteorological knowledge together with locational and silvicultural information, which would allow the linkage to silvicultural practices.
References
Andersen, K.F. Gales and gale damage to forests, with special reference to the effects of the storm of 31st January 1953 in the North-East of Scotland.. Forestry 27(2):97-121, 1954.

Anklin, M., Stauffer, B., Geis, K., and Wagdebach, D. Pattern of snow accumulation along a W. Greenland flowline - no significant change observed during recent decades. Tellus Series B-chemical and physical meteorology 46 (4):294-303, 1994.

Wind and Snow Damage. Forestry Commission 58th Report - 1977 - 1978 HMSO, London:16-17, 1979.

Bang, B., Nielsen, A., Sundsbo, P.A., and Wirk, T. Computer simulations of wind-speed, wind pressure and snow accumulation around buildings (Snow-Sim). Energy and Buildings 21 (3):235-243, 1994.
Barry, R.G. Mountain weather and climate, Routledge, 1992.
Baumgartner, M.F. and Apfl, G. Towards an integrated geographic analysis system with remote-sensing, GIS ad consecutive modeling for snow cover monitoring. International Journal of Remote Sensing 15 (7):1507-1517, 1994.
Bollard, W.A. Snow storm in Delamere Forest. Forestry Commission Journal 36:123-125, 1969.
Bonacina, L.C.W. Snowfall in the British Isles during decade 1926-35. British Rainfall 272-292, 1936.
Braun, L.N., Brun, E., Durand, Y., Martin, E., and Tourasse, P. Simulation of discharge using differential methods of meteorological data distribution, basin discretization and snow modelling. Nordic Hydrology 25 (1-2):129-144, 1994.
Brugge, R. Three years of warm weather over the British Isles. Weather 47:230-236, 1992.
Brun, E. and Martin, E. Modelling snow cover on different scales - application to avalanches, hydrology and climate. Houille blanche-revue internationale de l'eau 50(5-6):63-68, 1995.
Buisson, M.M.L., Good, W., and Vilaplana, J.M. Investigation of the spatial-distribution of the effects of snow carried by the wind -initial results of systematic measurements in-situ. Houille blanche-revue internationale de l'eau 50 (5-6):50-55, 1995.
Bultot, F., Gellens, D., Schadder, B., and Spreafico, M. Effects of climate-change on snow accumulation and melting in the Broye Catchment (Switzerland). Climate Change 28 (4):339-363, 1994.
Burns, F. Frequencies of snow depths for given places at selected stations in Scotland. Climatological Memorandum, Climatological Services (Met. O. 3), HMSO 40:25pp, 1964.
Burroughs, W.J. Why do cold Decembers in England come at the end of each century?. Weather 37:205-206, 1982.
Burt, S.D. Snowfall in Britain during winter 1978/79. Weather 35:288-301, 1980.
Carroll, S.S., Day, G.N., Cressie, N., and Carroll, T.R. Spatial modelling of snow water equivalent using airborne and groundbased snow data. Environmentrics 6 (2):127-139, 1995.
Carsey, F. Remote sensing of ice and snow - Review and Status. International Journal of Remote Sensing 13 (1):5-11, 1992.
Champion, D.L. The frequency and duration of snowfall in Great Britain. Weather 2:99-101, 1947.
Clarke, P.C. Royal Meteorological Society Discussion Meeting on Snowfall. Weather 29:73-75, 1974.
Cremer, K.W., Carter, P.R., and Minko, G. Snow damage in Australian pine plantations. Australian Forestry 46 (1):53-66, 1983.
Cremer, K.W. Snow damage in Eucalypt forests. Australian Forestry 46 (1):48-52, 1983.
Davis, D.T., Chen, X.Z., Huang, J.N., Chang, A.T.C., and Tsang, L. Retrieval of snow parameters by iterative inversion of a neural network. IEEE Transactions Geoscience and Remote Sensing 31 (4):842-852, 1993.
Davison, R.W. The supply of snow in the Eastern Highlands of Scotland: 1954-5 to 1983-4. Weather 42:42-50, 1987.
Davison, R.W. Winter weather and the supply of snow in the Eastern Highlands of Scotland: 1954/5 to 1983/4. Research Discussion Paper no. 21, Dept. Geography, Univ. Edinburgh , 1985.
Dey, B., Shanna, U.K., and Rango, A. Linear or non-linear covariance of seasonal snow melt and snow cover in Western Himalayas. Nordic Hydrology 23 (3):183-192, 1992.
Donald, J.R., Soulis, E.D., Kouwen, N., and Pietroniro, A. A land cover-based snow cover representation for distributed hydrologic models. Water Resources Research 31 (4):995-1009, 1995.
Dunsire, A. Frequencies of snow depths and days with snow lying in Scotland for periods ending winter 1970/71. Climatological Memorandum no. 70, Climatological Services (Met. O. 3) HMSO , 1971.
Dyumin, A.K., Anfilofiyev, B.A., Istrapilovich, M.G., Kuon, Y.D., and Matayeva, N.T. Strong snow storm, their effect on snow cover and snow accumulation. Journal of Glaciology 19 (81):441-450, 1977.
Erickson, M.C., Dallevalle, J.P., and Jensenius, J.S. Snow versus rain - looking beyond the magic numbers - comment. Weather and Forecasting 8 (4):542-548, 1993.
Ferguson, R.I. and Morris, E.M. Snowmelt modelling in the Cairngorms NE Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences 78:261-267, 1987.
Ferguson, R.I. High densities, water equivalents, and melt rates of snow in the Cairngorm Mountains, Scotland. Weather 40:272-277, 1985.
Ferguson, R.I. Magnitude and modelling of snow melt runoff in the Cairngorm Mountains, Scotland. Hydrological Sciences Journal 29:49-62, 1984.
Frank, H. Snow damage of the winter 1946-47. Forestry Commission Journal 19, 98 HMSO, London:, 1948.
Gary, H.L. and Walkins, R.K. Snowpack accumulation before and after thinning a dog-hair stand of Lodgepole pine. Research Note Rocky Mountain Forest and Range Experiment Station, USDA Forest Service RM-450:, 1985.
Gauble, S.L., Kochanski, W.W., and Irwin, P.A. Finite area element snow loading prediction - applications and advancements. Journal of wind engineering and industrial aerodynamics 42 (1-3):1537-1548, 1992.
George, D.J. The snowstorms of 4 March 1970. Weather 27:96-110, 1972.
Graham, J.N. An apparent relationship between the timing of lunar phase and severe winters. Weather 46:39-47, 1991.
Grayson, A.J. The 1987 Snow Storm: Impacts and responses. Forestry Commission Bulletin 87, HMSO, London 46pp, 1989.
Green, F.H.W. The transient snowline in the Scottish Highlands. Weather 30:226-235, 1975.
Green, F.H.W. Changing incidence of snow in the Scottish Highlands. Weather 28:386-394, 1973.
Green, F.H.W. Seasonal changes of snow cover in the Cairngorms. Weather 25:211-213, 1970.
Gregory, D. A consistent treatment of the evaporation of rain and snow for use in large-scale models. Monthly Weather Review , 1995.
Groisman, P.Y., Karl, T.R., Knight, R.W., and Stenchikov, G.L. Changes of snowcover, temperature, and radiative heat-balance over the northern hemisphere. Journal of Climate 7 (11):1633-1656, 1994.
Gumbel, E.J. Statistical theory of extreme values and some practical applications. U.S. Department of Commerce, Nat. Bur. Stds., Applied Maths. Series 33:, 1954.
Gutzler, D.S. and Rosen, R.D. Interannual variability of wintertime snow cover across the northern hemisphere. Journal of Climate 5 (12):1441-1447, 1992.
Hale, S.E. A simulation of the patterns of snow in the Scottish Highlands. 1992.
Harrison, S.J. Global Warming and winter road maintenance. Highways and Transportation in press:, 1992.
Harrison, S.J. and Harrison, D.J. Characterising winters: An index for use in applied meteorology. Journal of Meteorology (UK) 16:329-333, 1991.
Hottelet, C., Blazkova, S., and Bicik, M. Application of the Eth snow model to 3 basins of different character in Central Europe. Nordic Hydrology 25 (1-2):113-128, 1994.
Hulme, M. and Jones, P.D. Temperature and windiness over the United Kingdom during the winters of 1988/89 and 1989/90 compared with previous years. Weather 46:126-136, 1991.
IPCC Committee on Climate Change, Climate Change: The IPCC Scientific Assessment, Cambridge Univeristy Press, 1990.
Isaksson, E. and Karlen, W. Spatial-pattern and temporal-pattern in snow accumulation Western Dronning-Maud-land Antarctica. Journal of Glaciology 40 (135):399-409, 1994.
Jackson, M.C. Snow cover in Great Britain. Weather 33 (8):298-310, 1978.
Jones, R.H. and Vecchia, A.V. Modelling Snowmass Data. Proceedings of the 25th Symposium on the Interface: Computing Science and Statistics: Statistical Applications of Expanding Computer San Diego, CA.:1-10, 1993.
Karl, T.R., Groisman, P.Y., Knight, R.W., and Heim, R.R. Recent variations of snow cover and snow fall in N. America and their relation to precipitation and temperature variations. Journal of Climate 6 (7):1327-1344, 1993.
Kirnbauer, R., Bloschel, G., and Gutknecht, D. Entering the era of distributed snow models. Nordic Hydrology 25 (1-2):1-24, 1994.
Kounchowski, K. and Marciniak, K. Variability of mean monthly temperatures and semi-annual precipitation totals in Europe in relation to hemispheric circulation patterns. Journal of Climatology 8:191-199, 1988.
Lamb, H.H. Types and spells of weather around the year in the British Isles: Annual trends, seasonal structure of the year, singularities. Meteorological Office, Dunstable , 1950.
Leathers, D.J., Ellis, A.W., and Robinson, D.A. Characteristics of temperature depressions associated with snow cover across the North coast United States. Journal of Applied Meteorology 34 (2):381-390, 1995.
Lines, R. Stability of Pinus contorta inrelation to wind and snow.. Pinus Contorta as an exotic species. Proceedings of IUFRO Working party Meeting. 209-219, 1980.
Loth, B., Graf, H.F., and Oberhuber, J.M. Snow cover model for global climate simulations. Journal of Geophysical Research - Atmospheres 98 (D6):10451-10464, 1993.
Lowndes, C.A.S. Substantal snow falls over the UK 1954-69. Meteorological Magazine 100 (1188):193-207, 1971.
Lynchstieglitz, M. The development and validation of a simple snow model for the GISS GCM. Journal of Climate 7 (12):1842-1855, 1994.
Manley, G. Snowfall in Britain over the past 300 years. Weather 24:428-437, 1988.
Manley, G. Variations in the frequency of snowfall in E. Central Scotland 1708-1975. Meteorological Magazine 107:1-16, 1978.
Manley, G. The mountain snows of Britain. Weather 26:192-200, 1971.
Matson, M., Ropelewski, C.F., and Vanadore, M.S. An atlas of satellite-derived northern hemisphere snow cover frequency., Washington, D. C. :U.S. Department of Commerce, 1986.
Mayes, J.C. Contrasting weather in Northern Scotland 1988-90 in relation to regional airflow types. Weather 46:16-21, 1991.
McKay, G.A. and Gray, D.M. The distribution of snow cover. In: Handbook of snow, principles, processes, management and use, edited by Gray, D.M. and Male, D.H.,Pergammon, 1981,p. 153-190.
Meier, M.F. Application of remote sensing techniques to the study of seasonal snow cover. Journal of Glaciology 15:251-265, 1975.
Mellor, M. A Technical discussion of blowing snow, with application to snow drifting around structures and on highways. Explanation of the mechanisms of transport, with emphasis on turbulent suspension. Blowing Snow, Cold Reg. Sci. & Eng., US Army Corps Eng., Cold, Reg. Res. Part III - A3c:79pp, 1965.
Meteorological Office, Frequencies of snow depths and days with snow lying at stations in Scotland for periods ending winter 1981/82. Climatological Memorandum no. 144, Climatological Services (Met. O. 3), HMSO , 1983.
Moses, T., Kiladis, G.N., Diaz, H.F., and Barry, R.G. Characteristics and frequency of reversals in mean sea-level pressure in the North Atlantic sector and their relationship to long-term temperature trends. Journal of Climatology 7:130, 1987.
Rango, A. and Martinec, C.J. Areal extent of seasonal snow cover in a changed climate. Nordic Hydrology 25 (4):233-246, 1994.
Ratcliffe, R.A.S. Winter prediction. Weather 45:271-272, 1990.
Ratcliffe, R.A.S. Review of winter 1985/86 over the Northern Hemisphere. Weather 41:170-172, 1986.
Ratcliffe, R.A.S., Weller, J., and Collison, P. Variability in the frequency of unusual weather over approximately the last century. Quarterly Journal Royal Meteorological Society 104 (440):243-255, 1978.
Robinson, I.D.A., Newey, K.F., and Hein, R.R. Global snow cover monitoring - an update. Bulletin of the American Meteorological Soc. 74 (9):1689-1696, 1993.
Rodda, J.C. A country-wide study of intense rainfall for the United Kingdom. Journal of Hydrology 5:58-69, 1967.
Rodda, J.C. A study of magnitude, frequency and distribution of intense rainfall in the United Kingdom. British Rainfall , 1962.
Rowntree, P.R. Estimates of future climatic change over Britain. Part 2: Results. Weather 45:79-89, 1990.
Schaeffer, J.A. and Messier, F. Scale-dependent correlations of arctic and snow cover. Arctic and Alpine Research 27 (1):38-43, 1995.
Schmidtrogt, H., Frank, A., Konnert, M., and Deichner, P. Identification, growth and resistance to snow break of spruce provenances from the Black Forest. Allgemeine Forst und jagdzietung 163 (7-8):145-154, 1992.
Scott, P.A., Hansell, R.I.C., and Erickson, W.R. Influences of wind and snow on Northern tree-line environments at Churchill, Manitoba, Canada. Arctic 46 (4):316-323, 1993.
Sommerfeld, R.A. Classification outline for snow on the ground. USDA Forest Service Research Paper, Rocky Mountain Forest and Range Experiment, U.S. Dept of Agriculture, Ft Collins, Co. Colorado RM-48:, 1969.
Spink, P.C. A summary of summer snows in Scotland 1965-1979. Journal of Meteorology (UK) 5:105-111, 1980.
Sturm, M., Holmgren, J., and Liston, G.E. A seasonal snow cover classification system for local to global applications. Journal of Climate 8 (5):1261-1283, 1995.
Tout, D.G. The variability of days of air frost in Great Britain. Weather 42:268-273, 1987.
Troen, I. and Mortensen, N.G. WAsP: Wind Atlas Analysis and Application Programme: an introduction. Riso National Laboratory, Denmark , 1988.
Valinger, E., Lundqvist, L., and Brandel, G. Wind and snow damage in a thinning and fertilization experiment in Pinus-Sylvestris. Scandinavian Journal of Forest Research 2:129-134, 1994.
Valinger, E., Lundqvist, L., and Bondesson, L. Assessing the risk of snow and wind damage from tree physical characteristics. Forestry 66 (3):249-260, 1993.
Vames, S. Scientists sense of snow. Scientific American 272 (3):32, 1995.
Wallen, C.C. Impact of present century climatic fluctuations in the northern hemisphere. Geografiska Annaler 68A:245-278, 1986.
Wallis, H.S. On the snow storm of January 1881 (reprinted from the meteorological magazine, Feb. 1881). Meteorological Magazine 122 (1457):298-300, 199.
Watson, A., Davison, R.W., and French, D.D. Summer snow patches and climate in NE Scotland UK. Arctic and Alpine Research 26 (2):141-151, 1994.
Wright, J.A. and Quine, C.P. The Use of a Geographica Information System to Investigate Storm Damage to Trees at Wykeham Forest, North Yorkshire. Scottish Forestry 47(4):166-174, 1993.
Xu, H., Bailey, J.O., Barrett, E.C., and Kelly, R.E.J. Monitoring snow area and depth with integration of Remote Sensing and GIS. International Journal of Remote Sensing 14 (17):3259-3268, 1993.