U. S. DEPARTMENT OF AGRICULTURE
FOREST SERVICE

ALTA AVALANCHE STUDY CENTER
Project D
Progress Report No. 1

A TECHNIQUE FOR FORECASTING THE RATE OF SNOWFALL AT
ALTA,UTAH

Barry C. Nielsen
Meteorologist
Wasatch National Forest
(Department of Meteorology
University of Utah)
November 1966

INTRODUCTION

Forecasting and control of avalanche hazard is very dependent upon certain meteorological parameters. The most important of these are snowfall rate, total snowfall from a given storm, moisture content of the snow, air temperatures, and wind. For this reason, a study was initiated at Alta, Utah, in which rather detailed weather observations were taken during storms in order to study winter storm characteristics and their relationship to snowfall and other meteorological parameters at Alta. These relationships are described in this report.

A method to semi-objectively forecast snowfall rates and snowfall accumulations for periods of up to 24 hours, using a modification of Harley's method is also discussed. The results of this method are applied to the winter months of January through March 1965.


STORM CHARACTERISTICS AND WEATHER

The heavy accumulations of snow at Alta are more a function of the duration rather than intensity of the snowfall. Figure 1 illustrates this point. Snowfall rates are often moderate or heavy during and two to four hours following the passage of a front or trough, but at other times the intensity of the snowfall is usually classified as light. Therefore, it is not only important to be able to forecast the rate of snowfall, but also the duration.

During a storm period, it was observed that precipitation will be nearly continuous at Alta so long as certain meteorological conditions are present. The first of these is dynamic lifting of the air mass. This usually occurs when the curvature of the flow or wind shear is cyclonic, or there is positive vorticity advection, etc. When anticyclonic curvature is observed, precipitation seldom falls. One exception to this rule is when over-running moisture moves over the area from the Pacific Northwest well in advance of a disturbance. This may cause continuous light orographic snow to fall, even though other conditions do not favor precipitation. It can be shown that during a storm, precipitation generally will fall until the curvature of the flow becomes anticyclonic (or vorticity advection becomes negative), at which time precipitation will nearly always cease. The other factors of importance are wind velocity and direction, which determine the mechanical lifting. As expected, the more nearly perpendicular the wind direction is to the mountain range, and the stronger the winds are, the more pronounced is the orographic effect. At Alta, upper wind directions of 240 degrees to 300 degrees usually are associated with the heaviest precipitation. Light winds, winds nearly parallel to the Wasatch Mountains, or winds within an easterly component are generally associated with relative light or no precipitation. Examples of the latter conditions are found in storms in which a cold low moves directly over Northern Utah. The upper winds become light and variable, leaving only dynamic lifting to produce precipitation. In this situation, similar amounts of precipitation fall at both valley and mountain locations. Another example is storms originating over California which result in a strong moist southerly flow of air. This situation often gives heavy precipitation to east-west oriented mountain ranges such as the Uintahs, but relatively small amounts at locations such as Alta, since the winds are nearly parallel to the mountain range.

Another important factor which influences precipitation at Alta is the moisture content of the air. A warm moist air mass will produce more precipitation than a dryer one, provided, of course, all other factors remain the same. The above factors must all be considered when forecasting snowfall rates and accumulations at Alta.

Temperature and wind are also important in estimating the avalanche hazard. During storm periods, air temperatures at the three sampling locations in the Alta area (Tanner's Flat, 7710 ft.; Alta Guard Station, 8760 ft.; and Germania Pass, 10,300 ft.) were found to be very close to the free air temperatures taken from the Salt Lake City radiosonde data for each respective elevation. See figures 2,3,and 4. During fair weather, the effects of radiation are evident, with all stations at Alta having higher temperatures during the day, and lower temperatures during the night than those indicated on the Salt Lake Sounding.

The wind information obtained during this study was unreliable due to the instrumentation problems.


A TECHNIQUE FOR THE PREDICTION OF PRECIPITATION AT ALTA

Precipitation can be considered a function of two parameters, namely (1) vertical motion and (2) available moisture. The vertical motion is a function of the prevailing synoptic conditions, the release of latent heat by condensation, the topographic conditions, and frictional effects. Since the main interest here is the orographically induced precipitation, only vertical motion due to topographic influences is considered in this report. It is known that the majority of the precipitation which falls at Alta is orographically induced.

The vertical velocity for air flowing up a slope can be computed by the well-known expression

where w_m is expressed in millibars per 3 hours, g is the force of gravity. P_o is the density of the air at ground level, P_o is the pressure at ground level, and VH is the slope of the ground. For the elevation of Alta, the term 700/P_o is nearly unity and is neglected.

Since little is known about the wind flow in and over mountainous regions, a modified value of the 700 Mb wind is used for VO. An advantage of using the wind at this level is that current and prognosticated data are readily available at this level. It was found that when the wind direction was from the 240 degree to 300 degree interval, the 700 Mb wind speed could be used directly for V_o. For other westerly directions, it was found that it was necessary to modify the 700 Mb surface wind velocity as

where 0 = (240 degrees - 700) and 0- = (700 - 300 degrees) for southwesterly and northwesterly flow respectively, where o is the 700 Mb wind direction. This factor decreases the orographic lifting effect of the wind as it becomes more and more parallel to the mountain range until VO becomes nil for wind directions of 360 degrees or 180 degrees. For wind directions with any easterly component, orographic lifting is considered to be small, and for computational purposes a value of V_o= 5 mph is assumed. From topographic charts of the Alta area, VH was approximated as 0.095.

The available moisture, or precipitable water may be obtained by the standard equation

This value may be conveniently obtained by use of a plastic overlay (developed by the Weather Bureau) used in conjunction with the Stuve diagram.

Harley (1965) developed a tabular method of computing precipitation rates using the computed vertical velocity and precipitable water. See Table 1. His method is used in this study to compute the precipitation rate which is orographically induced. This accounts for most of the precipitation at Alta.

During each storm period, the topographically induced vertical velocity and precipitable water were computed for each 12 hourly sounding taken at the Salt Lake City Airport, and interpolation was used to obtain the intermediate data at the 6 hourly intervals. Using the Harley method, precipitation rates were computed (in units of inches/6 hours) and combined with the duration of snowfall to give a 24-hour precipitation computation. A 24-hour period was used since precipitation data for Alta is only given for that interval of time. The results are shown
in the scatter diagram of Fig. 5 in which the computed precipitation total (horizontal axis) is plotted against the observed precipitation (vertical axis). It is seen that the computed value agrees quite satisfactorily with the observed for most cases--especially when the inherent errors of the data are considered (such as large time intervals between sounding and map data, and inaccuracies in the observed snow depth due to drifting).

Other errors in excess of 0.25 in. can be related to non-orographic effects such as heavy snowfall associated with active frontal systems, and heavier than computed precipitation associated with very unstable conditions and pronounced dynamically induced upward motion due to synoptic conditions. These situations are usually associated with substantial precipitation amounts at the Salt Lake City Airport where precipitation depends more upon dynamic lifting and not mechanical lifting. On the other hand, cases where precipitation was less than that computed can usually be identified with the situation where there is sufficient moisture and orographic lift, but synoptic conditions (such as anticyclonic curvature, wind shear, vorticity, etc.) tend to inhibit the orographic effect by subsidence. This Ts sometimes the case with over running moisture from the Pacific which is advancing over the western United States which is dominated by an upper air ridge.

An analysis of errors greater than 0.25 in. in shown in Table 2.


USE IN PREDICTION

Numerical weather analysis has made possible fairly accurate prediction of the upper air features, surface conditions and winds aloft for periods in excess of 24 hours. Also, even though the actual precipitable water is not a parameter which can easily be directly predicted, it can be estimated by indirect means. It is known that the 500 - 1000 Mb thickness is proportional to the precipitable water. The thickness pattern can be predicted by numerical techniques. Changes in the thickness can be estimated for a given locality and inferred changes in precipitable water amount can be determined. The predicted values of 700 Mb or 10,000 ft. winds and precipitable water can then be used to compute the orographic precipitation rate, and this, combined with the expected duration of the orographic precipitation (based on the curvature of the 700 Mb flow), can be used to estimate the total orographic precipitation for a given period. This may be combined with expected non-orographic effects (such as fronts, etc.) to estimate the total precipitation rates and amounts.


REMARKS

At the time this study was made, radiosonde data were available at 12 hourly intervals only. Thus, considerable interpolation was necessary in order to obtain data for 6 hourly intervals. This source of error should be reduced now, inasmuch as radiosonde data are available at 6 hourly intervals. This should improve the accuracy of the method. Also, radar data are now available, which should help considerably in forecasting precipitation duration, time and intensity, especially over a short time interval.

This study and forecast method is only an exploratory investigation into various relationships between storm characteristics and what actually happens at Alta. Much work remains to be done in better substantiating,extending and refining precipitation forecasting in a mountainous area.

*******************************