A unit of Lasting Forests evolving since March 30, 1999
A Total Forest Management Plan and Wildland Management Decision Support System
[ HOME | The Trevey Home | Table of Contents | The Finder | Glossary ]

Snow

Crystal Any substance, usually solid, the atoms or molecules of which are arranged in an orderly array. Firnification Densification of the fallen snow cover. The density of randomly packed, uniform ice spheres is 580-600 kg/m3. Further densification is from melting, re-freezing, and pressure (as in glacier ice - 700-800 kg/m3 ). Grain The obvious subunit in snow on the ground. Precipitation Liquid or solid water that falls from the atmosphere to a surface on Earth. Qali The snow retained on tree limbs (It tends to compact the snow when it falls, not increase snow depth) Quality (of snow) Percent by weight that is ice (90% in frozen snow pack = 10% liquid water) Snow Crystal A single crystal of ice, usually of complex form, which grows by vapor deposition in the atmosphere. Snow Flake A polycrystal of airborne snow. An agglomerate of snow crystals. Sublimation The process of vapor forming directly from a solid. The opposite of deposition. Surface Hoar The deposition of water vapor directly on the surface without precipitation. Flat plates are common and there is no three-dimensional development. Texture The relationships among snow grains or crystals in a snow cover.from Sommerfeld 1976 Kilograms per cubic meter is the dependent variable for snow strength (related to avalanches)

The complexities of snow and its distribution and accumulation (snow packs) and its multiple factors make estimating it and its effects very difficult. Assuming it can be done precisely is unwise and accepting high variance is reasonable. The variables are, at least, albedo, vapor pressure, relative humidity, condensation factors, and cloud cover, all related to temperature, winds, elevation, aspect (thus solar radiation), main continental and oceanic storm patterns, snow fall intensity and sequences of rain, hail, and sleet before and after snowfall. Snow accumulations (snow packs) are variable.

For wildlife, depth (and energy costs of moving through it), crusts (upon which animals move easily), and drifts (as used by grouse) seem very important.

Nevertheless, knowledge of snow is important to and used by those interested in prospective runoff and conditions, reservoir operators, highway maintenance and safety, fishery staff, irrigators, power companies and line maintenance staffs, flood control operators, industrial, municipal, and other water users. It is of great interest to winter sport area developers and to tourism interests. Wildlife populations seem affected by it.

Compartment model of annual flow of precipitation. Evaporation and sublimation are shown for the Snow Pack

A Small Knowledge-Base Herein, I attempt summaries for modeling and for developing working hypotheses for rural life:

• Natural snow is formed in very complex shapes which tend to fuse together and re- crystallize in different shapes in snowpacks.
• Distributions of snow particle diameters in drifting snow with concurrent snowfall extend to larger sizes and exhibit a tendency to be bimodal compared to distributions in drifting without snowfall.
• Snow is precipitation with many characteristics different in the ecological system than rainfall. Snow is rain without soil impact, but it has many different ecological characteristics.
• The percent of snow typically intercepted by forest canopy cover is
  %Interception = 0.036 % Canopy cover.
• Snow aids nighttime cooling when skies are clear, radiating energy more effectively than bare ground."Snow reflects nearly 100 per cent of incident solar radiation on one hand, and radiates nearly as a blackbody in the infrared on the other."
• Relative humidity is maximum at night when temperatures are lowest.
• Snow insulates cooling air from relatively warm ground underneath.
• Cool air off of snow flows down-elevation, influencing animal conditions. Flow may be slowed by vegetation (little by grasses).
• Clean snow reflects 80% of solar radiation. Dirty snow reflects 40-50% of sun energy and heats up more, thus melts more than clean snow.
• Snow "preserves itself" by keeping air over it colder than bare ground nearby.
• After prolonged wet snowfall, runoff may likely be greater under a forest canopy than from an open site. This results from canopy drip.
• Avalanches are a major concern for recreation and highways.
• Changes in the aerodynamics of the forest canopy and harvesting can significantly increase snow accumulations in the forest opening and decrease it in the surrounding forest. Increased accumulations after harvests are usually due to reduced interception loss and redistribution of snow by wind.
• Snow in harvested portions of an area melt sooner than that in the remaining forest in the spring.
• Timber harvests reduces the evapotranspirational water losses but also change the aerodynamics and energy balance of the forest. Additional snow is dumped directly into small openings (gaps) or breaks in the forest canopy. Increased solar radiation in openings melts snow pack faster and earlier in the season than other snow.
• Redistribution of snow is one major issue. Changes from harvests, can remove the scant deposition from some areas and concentrate it in new locations.
• There can be significant increases in snow accumulation in forest openings and these are results of deposition and re-distribution. The size and shape of openings alters the trapping efficiency of the land. Trapped snows deplete down-wind snow packs in similar openings.
• Snow water equivalents peak toward the center of 1H openings, generally where there is minimum wind speed.
• A 5H opening will maximize the collection of snow(Rocky Mt) but others have found 2H and 3H openings to be maximum. Effects on re-distribution of snow may persist until the replacement stand is 3/4th or more of the height of the residual stand.
• Advance melting (earlier in season) resulting from the forest opening can cause a hydrograph to rise more quickly with little effect on on recession flows. There is usually small annual flow increase from treatment, only the rising side of the hydrograph.
• When 40% of area of a watershed is harvested (50% of timber volume) then 40% increase in flow can be expected.
• In studying Western US watershed work, the objectives of providing water for human use need to be clarified relative to Eastern interests in erosion control and flooding loss reductions.
• Timber harvesting in the snow zone reduces the evapotranspiration losses. This reduces soil moisture during the growing season. The timing and duration of soil moisture and stream flow are so significant to different objectives in the West that it is impossible to predict the actual effects of harvesting and snow re-distribution. Evaluating the full impact of different shapes of timber harvest openings seems a worthy challenge as soon as objectives can be clarified for all uses of the land for at least a 60 year period, perhaps longer. It may not be possible.
• Opening should not exceed 8 tree heights (8H) in width.
• If snowpack is more than 3.5 feet, the soil surface will probably not freeze.
• Snow is a good insulator. Wet soil has both higher heat conduction and capacity than does dry soil. Soil wet in the fall before freezing does not freeze as quickly or as deeply as a soil that is dry in the fall. Soil wet in the fall may take longer to thaw in the spring.
• A cold soil holds more water than a warm soil in that temperature influences both surface tension and viscosity.Clays carry heat down into the ground and do not warm as readily as sandy soils (poor conductors.)
• Rain-on-snow affects water content, runoff values, conditions of the snow pack but does not add enough heat to melt much snow. (See Sequences below)
• Much evaporates. Evaporation from snow is about 0.02 in. per day (Forbes 1955:10-10).
• Daily snow melt can be well estimated by degree days based on 0-degrees Fahrenheit
  M = KD
  =K ( T_a - T₀ )
  =K ( T_a - 32) ; for T₀ = 32 degrees F

  where
  M = daily snow melt in inches
  K = snow melt constant or degree day factor in inches / degree day
  D = (T_a - T₀ ) = total degree days above the reference temperature
  T_a = mean daily air temperature, or average of daily maximum and minimum
  T₀ = a base reference temperature, frequently taken as 30-degrees F, the melting point.
  K ranges from 0.02 - 0.11 with maximum of 0.30. SCS uses average of 0.06 and it seems reasonable that it is likely to vary over a snow-melt period.

  Equations have been developed and may be further refined for local work

  M = 0.0110 D^1.53 with R² = 0.88 and

  M = 0.214 + 0.0618D with same r-square.

• Over 60% of energy required for snowmelt is attributed to direct solar radiation, 20% to air/wind convection, and 20% to condensation loss. Thus estimated solar radiation to an Alpha unit may be an excellent relative index to snow depth, persistence, winter severity and effects on deer and other wildlife.
• Rain affects water content, runoff values, and conditions.
• Freshly fallen snow melts quickly in the forest, whereas it accumulates in nearby clearcuts. This produces uneven runoff from areas treated nearly identically.
• Fairly uniform in first deposition, it add waters to land area in very different amounts
• When early falling (before leaf fall), it may add enormous amounts of litter and limbs to the forest surface
• It offers protection to fauna in the soil and snow layer from deep freezes (e.g., ruffed grouse in snow banks)

Snow depths are taken with tubes. Costs and dangers of winter sampling are high. Here Giles was on Moscow Mountain snow patrol, with Mr. Chub, SCS, 1963

Sequences

Sequences include rain-on-snow, snow-on-frozen ground, and snow-on-wet (or dry) ground

The importance of sequence of events, often forgotten or rarely given comment, cannot be underestimated. Rain on snow can have runoff consequences of great importance. Rain on snow events have been observed to be generally less from forests than open areas but the amounts of difference is very difficult to quantify. Variables confounding the analyses are wind, net radiation, precipitation intensity under the canopy, and others.

Wildlife and Snow Relationships

A winter severity index can be developed from weekly air chill and snow depth.

Picton and Knight developed a winter condition index that might be mapped later where winter was November 15 to March 31 and

daily animals observed = a + b (32-degrees minus daily maximum air temperature) + c (snow depth)

Wildlife populations are especially affected by several bad years in sequence (cumulative effects) and populations can be studied (along with other forms) using the equation form:

Population = a + b (Snow _t + Snow _t-1)

where Snow is annual inches of water (snow equivalents). It may be that deer populations are a function of a sequence of three "bad" years as seen in hard mast production effects. Carrying capacity based on the concept of conditions at the minimum for a population for an area can result in a very low figure (no capacity) with occasional abundance and catastrophic losses.

On winter ranges, snow, more than any other factor, determines the availability of forage for deer and other animals. Snow may cover food supplies. Snow fences may influence drifts and allow deer to gain access to forage. Snow accumulations determine the area of potential winter range available to deer, moose, and elk early in the year. Winters of light snow are typically followed by population increases related to low adult deaths and high productivity from well-fed animals. A series of heavy-snow winters can result in greatly reduced populations.

Deer survive low temperatures better than deep snows but extended energy loss over a season by any and all means is the criterion for population success.

Crusting of snow results in high energy costs to animals.

Wind (and thus wind-chill) has an combined effect of robbing body heat (convective loss) and increasing crusting, thus required expenditure of energy.

Peek observed that the snow characteristics (hardness, density, and presence of conifer conditions) other than depth influence moose habitat selection during the winter. Hardness can be determined by a Rammsonde penetrometer.

For musk oxen, there is a preference for snow-free sites (in that relatively sparse vegetation will be heavily cropped when only a few meters away, more lush vegetation will be covered by only a few cm of light snow).

Urine of big game has been collected from frozen samples in snow to assess the health of the herd.

Artificial barriers can be used to protect livestock from winter wind. Barriers can reduce winds, produce snowdrifts. Drifts can produce vernal pools. All solid barriers cause snow deposition with initial accumulations being on the windward side and extending about 10 times shelter height, H upwind. Little snow accumulates on the lee side initially because snow particles in the rapid flow are moved up and over the barrier. The windward drift becomes deeper up to fence height, then particles are moved over the barrier and settle out. V-shaped snow and wind barriers can be used for wildlife and cattle. Pointed windward, snow accumulates at the sides of the structure or diversions. Small drifts can be expected at 4H inside the V.

When ponds are constructed in open areas, snow drifts can occur. To maximize snow accumulation place the pond embankment "pointed"downwind, then with a snow fence on the windward side of the pond.

A - - The density of snow (see Moen p. 68) 0.05 to 0.10 gm³

the maximum of o.34 is deposited during gale winds

the maximum increases to 0.4 as snow pools age

Wt = A + 0.0065 * C

where C is the temperature increase after snowing.

Relative to water evaporating from a dogs mouth , say at 25 degrees C = y calories lost ( heat removed by vaporization)

y = 595.95 - 0.5376 X where X is air temperature in degrees C.

Mapping Snow Distributions AT Journeys, November/December 2005 - snow distributions on the Appalachian Trail area

It seems likely that we can achieve an index to probable duration and presence of snow based on temperature, solar radiation, topographic shade, and land use classes. Solomon et al. 1976 developed a program to estimate snowmelt from intermittent snow packs with 4 daily input variables: maximum and minimum daily temperatures, precipitation, and short-wave radiation or percent cloud cover. Their programs for solar radiation incident on a given surface during a clear day may be well used in other analyses. Similarly, their cloud cover program may be useful.

Sartz (1973) described the complexities of snow and frost in relation to slope and aspect. Summarizing can be difficult but the models eventually developed to reflect, from a base of "South-facing slope conditions":

• Snow accumulation has been found greater in forest openings (clearcuts) than in adjacent forests and snow melt greater on rain-on-snow events because of the greater energy inputs (Berris 1984).
• Late winter snow packs are greater on north-facing slopes. (50%)
• Snow retards soil freezing by insulating the ground
• With a snow pack, insulating effect of snow outweighs radiation effect received at the ground surface
• With little or no snow, radiation is the dominating influence
• The net effect of snow on expected aspect influence on soil freezing and thawing can be zero, i.e., equal frost depth on north and south aspects ... thus delayed (or not) runoff.
• On cold, snowless winters, there is deep frost penetration and the difference in spring runoff due to aspect is unimportant. In winters when snow is great and frost penetration slight because of snow insulation, aspect differences can similarly be unimportant.

Toward New GIS Models

Studies of snow can become very specialized and isolating the objectives for the many different parts of the rural enterprise can be difficult. Needs vary with crop, livestock, wildlife, and recreationists using the areas. The studies and dimensions to be addressed include GIS images reflecting the multiple regression dimensions of weather stations location, elevations, slopes, aspects, unified slope and aspect, distance from coast or distance from inland weather recording stations (a unified aspect and elevation variable) , latitude, longitude:

• snow frequency
• snow depth
• first and last date of snow fall
• days under snow; snow-free days
• water equivalents

The needs for models and understanding for GIS display seem to me to be (2005):

1. probable number of days when snow on the ground will exceed 12 inches
2. probability of annual snow greater than 3 decided limits such as >15, >20, >25 inches.
3. probable snow water equivalent to total precipitation estimate for each alpha unit
4. probable maximum snow density (psi)
5. probable rain-on-snow events per year
6. probable 4 inch or greater snow event during "leaf-on" condition.
7. probable days in a year when snow >6 inches (wildland recreational potentials )
8. various ratios of (1) non-snow and less than 15-degree days and (2) days with snow
9. standard snow fence relations (x see below) along level terrain roadways, given high snow depth estimates

Snow Loads (Water Equivalents)

One calorie is the amount of heat that must be added to 1 gram of water (1 cubic centimeter) in order to raise the temperature 1 degree Centigrade. (Example: 100 calories must be extracted from 10 cubic centimeters of water to cool it 10 degrees C.) Six hundred calories are required just to evaporate 1 gram of water. (Thus, evaporation of 0.1 of 1 cubic centimeter of water will cool the remaining 0.9 cubic centimeters 60 degrees C - provided no other heat exchanges occur.) In fact, such great cooling would not be realized because of conduction of heat from the relatively warm air that surrounds the water. Eighty calories of heat must be extracted from 1 gram of water to change it from water at 0 degrees C (32 degrees F) to ice at the same temperature. (Thus, as much heat must be extracted from 1 gram of water in freezing it as is necessary to reduce its temperature by 80 degrees C.)

Loads on building roofs can be substantial and design requirements can be altered to fit local snow conditions expected. Water equivalents are needed for estimating runoff.

Portman et al. (1965:5-9) said "There is no satisfactory way of making a representative measure of the depth of a snow fall. When depth is measured, determining water equivalent is a problem. One estimate is 10 to 1 or one cm of snow is equivalent to 0.1 cm of water. In Vermont, water equivalence varied from 2.5 to 22 percent with a mode of 9 percent.

Drifted snow is more dense than snow that falls in a sheltered area.

Azuma (1985) found no relationship between elevation and snow load (weight on building roofs, etc.) on either a statewide or river-basin level in California, Lake Tahoe basin.

He used the following equation to get PSI or pounds per square inch

psi = predicted snow water equivalent in inches x 0.0361

or (1 ft³/ 1728 in³ x 62.4 lbs/ft³

Use of snow tubes tends to overestimate snow water content by as much as 12%. Obtaining good water equivalents is difficult due to drifting, slope, aspect, and other relations.

Freshly-fallen snow has density of 0.10 with a range of 0.07 to 0.15, thus 10 inches of snow contains 1 inch of water. in an acre-foot of water there are 3.26 x 10 ⁵ gallons of water or 27,167 gallons in an acre-inch.

1 gallon of water weighs 8.33 lbs.
Heavy snow accumulations can result in densities of 0.4 to 0.6 by spring.

In one state, the maximum water equivalent was 8.2 in snow of 25 inches ( density = 0.33)
A runoff of 2-3 inches ( rainfall or otherwise) can cause substantial flooding - thus a snow melt of over 20-30 inches can present a severe flood threat.

Managing Snow

Snow Fences

Snow fences are used to increase the snow pack, reduce sublimation, and regulate water for waterfowl and fisheries.

Snow fence relations are complex and very much a function of terrain shape and wind velocities. The general approximate equation (Tabler and Jairell 1980) with drift on the winward side of the fence:

y/H = 0.5 - 0.04 (x/H) , x/H is less than 12.5

where y is snow depth, H is the height of the fence in meters, and x is distance from the fence ( the fence is the Wyoming fence, with height from 0.8 to 3.8 m, 50% porosity, horizontal slats 15cm wide, and downwind inclination of fence).

Another model (Tabler, 1980) for approximate snow density for lee drifts, vertical slab fences, H in meters

p = 352H^0.18

q_lee = 5.1 H^02.18
where q is water-equivalent volume in m³ per meter of fence length

For common fences, lee-drift geometry is characterized by length proportional to fence height , H, and cross-sectional area proportional to to H² and water equivalent volume is proportional to H^02.18 due to density increases with snow depth. Thus there is a tall-fence advantage.

Snow Barriers Across Open Land

To collect water and reduce erosion and large drifts, make soil ridges and plant on the contour:

1. grasses, 18-inch stalks, strong, without lodging
2. in double rows
3. density to provide 65-74% air porosity

Snow Color

Regelin and Walmo (1975) studied the possibility of reducing snow depths in protected stands of forage by artificially darkening the snow surface. The surface increases adsorption of short-wave radiation which increased the rate of snow melt. (The procedure had been used in Russia and Japan to accelerate spring snow melt on airport landing strips.) They found the practice effective on southern slopes. reducing average snow depths by 92 percent compared to control plots. On easterly plots, reduction was 51%. Light snow falling on top of the black reduced its effectiveness. It can be applied aerially on big game management areas and increase the available forage. It seems reasonable that a combination of nitrogen fertilizer applied with carbon black could reduce costs and improve big game range conditions.

Further work on this unit may be in using GPS to delineate elevation zones of snow fall on local mountains, separating coastal and oceanic patterns. Snow removal issues related to highways may be of interest. Uses of salt and other melting substances may be added as these affect wildlife along roadsides and as they contaminate stream and groundwaters. See Geiger and Olgyay "The Climate Near the Ground." Snow coverage for segments of the Appalachian Trail may be of importance and will affect trail use rates as well as spring hiking conditions and trail maintenance. Snow prediction model was developed in 1970s for West Virginia.(?) As wind models improve, they may be added to drifting snow and snowfence considerations.

References

Azuma, D.L. 1985. Estimating snow load in California for three recurrence intervals, USDA For. Service Pacific SW Forest and Range Exp. Sta., Berkeley, CA. Res note PSW-379 , 6p.

Berris, S.N. 1984. Comparative snow accumulation and melt during rainfall in forest and clearcut plots in western Oregon, Oregon State Univ., Corvallis, OR 152 p.

Formozov, A.N. Snow cover and ecology of mammals and birds (see)

Gray, D.M. and D.H. Male, editors 1981.Handbook of snow: principles, processes, management and use, Pergamon Press, Elmsford, NY 796p

Haugen, A. editor. 1971. Snow and ice in relation to wildlife and recreation; Proc. of a symposium, Iowa State Univ. Ames, Iowa, Iowa Coop. Wildlife Research Unit.

Houghton, D.D. ed. 1985. Handbook of applied meterology, Wiley-Interscience, New York, 1461pp.

Jairell, R.L. and TR.D. Tabler. 1985. Model studies of snowdrifts formed by livestock shelters and pond embankments, Proc. Western Snow Conference, Bolder, CO, 1985, p. 167- 170.

Moen, A. and K. E. Evans, The distribution of energy in relation to snow cover in wildlife habitat, p.147-162 in Haugen 1971.

Regelin, W.L. and O. C. Walmo. 1975. Carbon black increases snowmelt and forage availability on deer winter range in Colorado, USDA Forest Service, Rocky Mt. Forest and Range Exp. Sta, Research Note RM-296, 4 p.

Rosenberg, N.J., B.L. Blad, and S.B. Verma. 1983 Microclimate 2nd ed., Wiley-Intersciece, New York, 495p.

Sartz, R.S. 1973. Snow and frost depths on north and south slopes., USDA For. Service, North Central For Exp. Sta., St. Paun. MN, 2p.

Solomon, R. M., P.F. Ffolliott, M.B. Baker, Jr., and J.R. Thompson, 1976. Computer simulation of snowmelt. USDA Forest Service, Rocky Mt. Forest and Range Exp. Station, Ft Collins, CO., Res Paper RM-174, 8p.

Sommerfeld, R.A. 1969 (revised 1976). Classification outline for snow on the ground, USDA Forest Service, Rocky Mt. Forest and Range Exp. Sta., Research Paper RM-48, SD-11-A522

Sommerfeld, R.A. 1974. A Weibull prediction of the tensin strength-volume relationship of snow. J. Geophysical research 79(23) : 3353-3356

Steinhoff, H.W. 1971. Planning study of ecological effects of artifically increased snow, p. 122-133 in Haugen, A. editor. Snow and ice in relation to wildlife and recreation; Proc. of a symposium, Iowa State Univ. Ames, Iowa, Iowa Coop. Wildlife Research Unit.

Tabler, R.D. 1980. Geometry and density of drifts formed by snow fences, J. Glaciology 26(94): 405-419.

Tabler, R.D. and R.S. Jairell. 1980. Studying snowdrift problems with small-scale models outdoors, Western Snow Conf Proc., April, Laramie, Wyoming 13p. (Rocky Mt. Forest and Range Exp. Station reprint)

Troendle, C.A. and C.F. Leaf. 1981. Effects of timber harvest in the snow zone on volume and timing of water yield in, Interior West watershed management, D.M. Baumgartner, compiler and editor, Symposium Proc., Spokane, WA April, 1980.

Go to top of page.

This Web site is maintained by R. H. Giles, Jr.
Last revision April 17, 2005.