Weather Forecasting: The Nature and Prediction of Winter Storms
Knowing how to interpret a winter forecast means knowing when to send out your crews.
Winter storms in the continental United States can be responsible for significant disruption to our commerce. In tragic instances, these storms can result in loss of life and property. To minimize the impact of such disturbances, we must accurately forecast the various meteorological elements that accompany them. It is important that public works officials, engineers and facility managers who are charged with providing for public safety on our roads understand these forecasts.
The nature of winter storms
Winter storms in the continental United States are called “cyclones” because the horizontal winds circulate counter-clockwise (cyclonically) around them. Each individual cyclone has its own specific characteristics, yet each shares general characteristics as well. Cyclones are often of “synoptic” scale (possessing a radius of 1,000 to 2,000 km) and can simultaneously affect millions of square kilometers. Thus, a single winter cyclone can affect the weather in Minneapolis, Minn., and Amarillo, Texas, at the same time, although the weather elements that it brings to each city will be quite different. These cyclones often have lifetimes of about five days during which time they move and evolve through a life cycle; first gaining intensity and finally lessening in fury. It is important to note that some portion of the large region affected by a given cyclone can experience hazardous weather at any stage of this life cycle.
Cyclones are not usually circularly symmetric vortices of swirling air. They have considerable structure on spatial scales smaller than the previously mentioned synoptic scale. Perhaps the most well known examples of this smaller-scale (or mesoscale) structure in cyclones are the frontal zones (the cold and warm fronts of the cyclone), which are characterized by the largest windshifts, the most intense horizontal temperature differences and often the most intense precipitation in the storm. The horizontal structure of a textbook cyclone is shown in Figure 2, below, along with vertical cross sections through the cold and warm fronts of the storm. Concentrated upward vertical air motions, which produce the precipitation in cyclones, occur in the frontal regions of the storm.
Because the frontal zones separate air masses with distinctly different temperatures, the distribution of precipitation type in winter storms often resembles the one depicted in Figure 1, page 6. Notice that much of the frozen precipitation in winter storms is associated with the warm front of the cyclone. However, in particularly strong storms, the heavy rain that immediately precedes the cold front is followed by plunging temperatures behind the front which can lead to rapid road freezes. It is important to note that, despite the fact that the mesoscale frontal zones are often about 100 km wide, the precipitation bands that develop within these frontal zones are often only tens of km wide. These small-scale features, in turn, are embedded within the massive synoptic-scale cyclone itself.
As you can imagine, the type of precipitation that falls in a given area is crucially dependent on the exact path the cyclone takes. For instance, a difference in cyclone track of less than 100 nautical miles can easily change a prospective blizzard, say, in the Boston metropolitan area, into nothing more than an annoying winter rain event or a slight dusting of snow (see Figure 3, facing page). This dependence on cyclone path is one of the most decisive factors influencing the impact that an impending winter storm will have.
There are several climatological paths for cyclones in the continental United States (see Figure 4, page 12). This accounts for a good number of observed winter cyclone paths; but, of course, not for all of them. The cyclone paths labeled A, B and C in Figure 4 often generate large amounts of precipitation as a result of plentiful moisture availability from the Gulf of Mexico or the Atlantic Ocean. The so-called Alberta Clippers (path D) usually remain well to the north of the Gulf of Mexico and are associated with lesser amounts of precipitation as a result. Their high northern latitude, however, ensures that frigid arctic air accompanies these storms and can often result in life-threatening, blizzard conditions through a combination of blowing and drifting snow that can continue long after the actual precipitation has ended. In addition, significant “lake effect” snows can occur downwind of the Great Lakes in association with, or in the absence of, a significant synoptic-scale cyclone. In fact, a great deal of lake-effect snow falls in the wake of the associated cyclone as it exits the region. With all of these spatial scales, path dependencies and precipitation-type issues to consider during a winter storm, you might be wondering, “How are forecasts of such complicated natural phenomena made?”
Predicting winter cyclones
Although the meteorologist on television traditionally portrays weather systems on two-dimensional maps, these weather systems are fully three-dimensional. Wherever a cyclone exists at the surface, you can be sure that there is a related disturbance at upper levels (up to 6 miles above the surface). In fact, it is almost always the case that the surface cyclone (the feature with which we all have to directly contend) is nothing more than a reflection of the much more vigorous cyclone at upper levels (see Figure 5, page 12). Surface cyclones form to the east of upper-level cyclones because the dynamics of the upper system compel the air to rise in that region, removing it from the atmospheric column there. The rules that govern this behavior can be described by mathematical equations. These equations are solved on high-speed computers using initial conditions supplied by observations. Synoptic-scale forecasts made in this manner have become extremely accurate in the past 15 years.
The hard part is forecasting on a smaller scale. Numerical forecast models rely on converting the unevenly spaced observations from around the world into evenly spaced, gridded data sets from which calculations can be made. The horizontal and vertical distances between these computer-generated data points are a major factor in determining the practical limits of forecasting ability. The National Center for Environmental Prediction (NCEP), a division of the National Weather Service, uses current numerical weather prediction models with horizontal grid spacings of 80 km. Winter cyclones are large enough to cover an area entailing many 80 km-square grids. This enables meteorologists to accurately map the current location and forecasted path of the cyclone. However, if a prediction places an otherwise perfectly forecasted cyclone just one grid point east or west of where it actually goes, huge differences in forecast accuracy result. Recall the example in Figure 3 concerning the sensitivity to cyclone path.
In addition to cyclone path errors in forecasts, other characteristics of cyclones conspire against accurate precipitation forecasts. Remember that most of the precipitation falls in frontal regions of the storms and that these frontal regions, while synoptic-scale (1,000 km) in length, are mesoscale (100 km) in width. If a forecast model has a horizontal resolution of 80 km, it may well miss the important air motions and precipitation production that is actually occurring in these frontal regions. Thus, many precipitation-producing processes that occur in frontal zones are “invisible” (unresolvable) to the numerical forecast models.
Numerical forecast models with smaller horizontal distance between grid points can overcome some of these limitations. Such models have been developed, resulting in more accurate forecasts. But, because they employ more grid points to cover the same area, they are much more expensive to operate. The explosion of computer technology in the past decade has brought the science of numerical weather prediction to a new level of complexity and accuracy. It is anticipated that this rapid advance will continue in step with further advances in the computer industry.
Sources of forecast information
Despite the fact that some of the most important forecast elements (such as precipitation and precipitation type) are among the least skillfully predicted by numerical forecast models, the picture is not so bleak. You can obtain useful information concerning precipitation probability, amount, location and time of onset from a number of sources in the public and private sector. The primary source of such information is the National Weather Service (NWS), which employs highly qualified scientists who treat the numerical forecast data as “guidance” with which to make their more accurate forecasts for local areas. Considerable value is added to the numerical product when it is interpreted by an experienced, knowledgeable forecaster who has access to up-to-the-minute satellite, radar and aircraft observations with which to gauge the accuracy of the numerical guidance. Potential cyclone path errors and precipitation intensity in frontal regions can be tracked in the observations, thus allowing the forecaster to make accurate short-term (6 to 12 hour) forecasts even in the face of occasionally poor numerical guidance. The primary responsibility of the NWS in winter cyclones is to issue advisories and warnings. Winter weather advisories are notices that the conditions are correct for the development of hazardous conditions. A warning (such as a heavy-snow warning or blizzard warning) is issued when the threat from winter weather is imminent and usually unavoidable (in other words, such conditions are already occurring a short distance away).
A large number of private sector forecasting firms and other businesses also provide weather information (a list of these can be found in the monthly publication, Bulletin of the American Meteorological Society). You can also use the Internet for weather information. Gather information while the weather is benign for a list of sites that you can reference quickly when the weather turns foul. Each of the more than 100 National Weather Service Forecast Offices (which you can access at http://www.nws.noaa.gov/ ) has its own web site with a voluminous amount of local climate and forecast information.
Optimum roadway maintenance during winter storms results from prediction of adverse weather conditions, not simply reaction to them. Our knowledge and predictive skill concerning winter storms has never been higher. This fact, coupled with your increasing access to information technology, can help provide you with a solid base in the development of effective roadway maintenance strategies.
Dr. Jonathan E. Martin is an assistant professor in the Department of Atmospheric and Oceanic Sciences at the University of Wisconsin (Madison, Wis.).
Want to use this article? Click here for options!
© 2017 Penton Media Inc.