The Tools and Concepts for Forecasting Winter Precipitation


This informative forecast writing will focus on the weather forecasting tools and concepts that are used to predict winter weather events. There are various types of winter weather events that are analyzed and forecasted including mid-latitude cyclones, ice storms, lake-effect snows and sleet storms. Each of these will be discussed separately. In addition, bust concerns and nowcasting aids will be covered. Many of these analysis and forecasting tools are available online and will be referenced within this writing. This essay is a general overview of issues and data sources related to forecasting winter precipitation. This writing will be handy to refer to as an introductory guide when analyzing and forecasting a winter precipitation event, especially when examining North American winter storms. A winter weather forecast is not easy to forget by the public because it will drastically alter their plans. If the public see winter precipitation in the forecast, it is a guarantee they will remember if it does not happen. When predicting rain, it will be in the liquid form but when predicting winter precipitation it can be snow, wet snow, sleet, freezing rain, just a cold rain, freezing drizzle, a wintry mix, a change from one form of frozen precipitation to another, etc. which adds to the complexity. Accumulation is also much more scrutinized than with just rain. A winter weather forecast is a big adrenaline rush for a forecaster. So much is on the line with this forecast and arguable the greatest variety of weather data has to be analyzed when putting this forecast together. Let’s get started!

Topic A. Mid-Latitude Cyclone Winter Storm

With a winter mid-latitude cyclone (deep cyclone throughout the troposphere), the heaviest snow is typically 100 to 300 kilometers to the left of the cyclone track. The heaviest snow is often in a narrow band. This region has the greatest combination of cold air and dynamic lifting. Snow in association with upper level lows (700 mb and higher), however, tend to be more directly below the low track. Watch mid-latitude cyclones that tilt with height. Often the upper level low will be displaced hundred(s) of kilometers to the west or northwest of the surface low. Upper level lows can bring unexpected heavy snow. Model vertical resolution is poorer in the upper levels as compared to the surface. The intensity of an upper level low can be misleading from the models. Upper level lows are notorious for their explosive development or rapid decay.

There are several common source regions that mid-latitude cyclone winter storms can have for their initial development. Although these are common source regions, if conditions are right, then a mid-latitude cyclone winter storm can develop pretty much anywhere in the mid-latitude and polar region. Common source regions of winter cyclones include: Colorado lows (above or adjacent to Colorado Rockies), Canadian lows (these can be generated from the Canadian Rockies and are often called Clipper Storms since the snow they generate tends to be lighter due to less moisture availability), CISK (Convective Instability of the Second Kind: large latent heat release that can result in, for example, a low pressure advecting rich moisture into the low such as from the Gulf of Mexico or Atlantic waters, in addition a general name given to the more severe cases are Nor’easters), Southwest U.S. trough in and adjacent to the 4-corners region and California (often induced by a jet streak as it rotates around a trough), Gulf of Mexico low (often form along pre-existing thermal gradient and can influence the Southern U.S. and evolve into Nor’easter storms).

There are several important questions that a forecaster asks when examining a potential winter storm set-up. The BIG 7 questions to ask yourself when looking at the models will be discussed in this portion of the essay. It is a good idea to keep these questions handy since reminders of what to look for are critical in order to avoid over or under-analyzing a particular element of the storm. Differences in the answers to any of these questions can cause the actual forecast to evolve differently than the forecasted one.

QUESTION 1: What will the track of the low pressure most likely be?

This is critical for determining areas that will receive rain, sleet, snow, a cold rain, nothing or a mixture of these either at the same time or over time. The boundary between rain and snow is analyzed most successfully if the track of the low is also forecasted successfully. Temperature advection location is determined by the position of the low as it evolves over time. Colder air tends to be west, north and under the low track while warmer air tends to be to the south and east of the low track. This is due to the cyclonic nature (counterclockwise motion) of air is it generally advects closer to the low pressure system. Comparing model tracks is important. Examine the primary forecast models and your favorite models to note the similarities and differences between the low tracks on each model. As the event evolves closer, such as a day and hours before the event, it is important to examine a wider range of forecast models such as mesoscale and high resolution models.

QUESTION 2: How will the low evolve through time (speed, intensity)?

Not only is the storm track important, but the speed and intensity are also important considerations. The speed will influence the length of time that each precipitation type(s) occur and thus the amount of precipitation that could occur. The intensity of the low will influence precipitation amounts also. A stronger low and a slower moving low will tend to generate more winter precipitation. If the low is an upper level low, be wary of rapid development or decay that can occur. Pay attention to lifting mechanisms on the forecast models. This is especially true for situations in which a significant low pressure is not present. There may not be enough lifting to generate winter precipitation even though all other factors are in place.

QUESTION 3: How much will the PBL (Planetary Boundary Layer) temperatures change through time?

In many winter weather situations, the ground temperature, surface air temperature and near surface temperature are important considerations. In a freezing rain situation, locations on the ground surface that are at or below freezing will support freezing rain but the rain will stay wet on above freezing surfaces. The ground temperature will influence how quickly, if at all, the precipitation will melt once on the ground. If the ground surface is above freezing, expect some accumulation to be lost to melting especially if the sleet or snow is light.

PBL stands for “Planetary Boundary Layer” and this is the layer of air that is most influenced by the surface (such as surface friction and surface energy transfer). It generally ranges from the ground surface to 50 mb to 150 mb above the ground (several hundred feet to around 5000 feet in thickness) depending on conditions. This is the last layer that precipitation falls through before reaching the surface and thus this layer influences what the final precipitation type will be that is observed at the surface. It is important to time when precipitation occurs and when the temperature is cold enough to support a certain precipitation type. A common forecasting mistake is to not match up precipitation liquid equivalent with the temperature profile as the precipitation falls. If cold air moves in too late, much of the accumulation can be lost to rain instead of being accumulating sleet or snow. For locations relatively close to sea level, the heaviest snow tends to fall in the region where 850 millibar temperatures range from -2 to -6 C (1000 to 500 millibar thickness between 5,310 and 5,370 m). In this zone there is a good overlap of available moisture and temperatures being cold enough to support snow when lifting is significant. CAA (Cold Air Advection) will cause the air to sink. CAA and drier air advecting into the area can overcome UVV (Upward Vertical Velocity) generated by a shortwave or another lifting mechanism. This can end a winter precipitation threat when lifting is marginal.

QUESTION 4: How much moisture will be advected toward the low?

Available moisture is one of the necessary conditions for precipitation generation. What is really meant by this is that there has to be enough moisture in the air so that when air is lifted, it will cool the air to 100% relative humidity and condense out moisture to generate precipitation. Dry air working into a low pressure system can result in forecast busts on the precipitation amount. What happens is that a significant portion of the condensed moisture and ice is lost to evaporation and sublimation. When air with a low relative humidity advects into the area, some to all of the precipitation can be lost to evaporation and sublimation aloft. This can vaporize precipitation before reaching the ground and/or can also prevent precipitation from developing. For significant winter precipitation, it helps to have a saturated temperature profile. This means the temperatures aloft are at or near 100% relative humidity through a deep layer of the troposphere (especially where lifting occurs). High dewpoints in the warm sector of mid-latitude cyclone can wrap moisture into the cold sector, producing more ice crystal generation once the air is lifted, enormous amounts of latent heat release and significant precipitation generation. The 1000 to 500 millibar average relative humidity generally needs to be greater than 90% to maximize heavy winter precipitation.

QUESTION 5: How do the temperature profiles (forecast soundings) evolve around the low?

A sounding profile shows how the temperature, dewpoint, and wind speed/direction change with height. From this, a variety of additional information can be determined such as the thickness between pressure surfaces, wet-bulb temperature and index values. Temperature, dewpoint and wind information is directly measured by rawinsondes (instrument packages tied to weather balloons) that sample the atmosphere as they rise. While these samples are done at only specific locations, the data that they give combined with other data allow forecast models to generate these profiles above many more locations of interest. The data is typically analyzed in 2-dimensions above a specific location, in which the dimensions are height (pressure) and time. The weather forecaster will examine how temperature, dewpoint and wind change through time above a location. Often overlapped on this data in a winter weather situation by the weather forecaster will be forecasted precipitation and wet-bulb temperature. For a 4-dimensional analysis the weather forecaster will examine these profiles for many locations. The additional 2-dimensions are length and width (distributed on a constant pressure surface) which gives a portrayal of weather conditions over a constant pressure surface. The combining of data can be used to generate analyses of vorticity, lifting and convergence. The use of a constant height surface is examined, for example, when looking at forecast model panels (i.e. 700 mb model analysis/forecast).

Often, a forecaster will be interested in the forecast at one particular location and will thus focus most on the model soundings and model data for that location. Questions that a forecaster will ask when examining soundings include: What factors are influencing the precipitation type?, How much wet-bulb cooling potential is there?, When precipitation is generated, what precipitation type is supported? How much accumulation will there be at the surface? How will the precipitation type(s) change over time?

Be on the lookout for this forecast pitfall: uplift mechanisms moving downstream just as freezing temperatures are advected in. This is especially true in the Southwest quadrant of the mid-latitude cyclone. It is critical that the vertical motion that generates precipitation is occurring at the same time there is a temperature structure capable of producing winter precipitation. Also, a one or two degree difference can greatly impact the winter forecast. Polar air often modifies due to warm soils. In other situations, WAA (Warm Air Advection) can raise temperatures above freezing just as lifting mechanisms move into place. Also, soil temperatures may be too warm for accumulation when the precipitation is light.

QUESTION 6: What is the potential for dry air to wrap into the low?

Even a shallow layer of dry air advecting into the region can stifle a winter weather precipitation event. In the past, this tended to be more of a problem when the models had a lower resolution. Even so, be on the lookout for dry air advecting in and ending a precipitation event (especially at the surface and at the height around where precipitation is being generated). Satellite data such as water vapor imagery is a good tool to use to look for monitoring dry air advecting into a low pressure system since you are looking at actual data with a high resolution. In some cases, air will become too dry in the cold sector to support snow. Although the cold air advection brings in cold temperatures that can support snow, it will often also drag in dry air that limits precipitation from reaching the ground. Dry air can also sneak in from an arid or semi-arid source region. This air can cut-off moisture rich air from entering the forecast area and can thus end a winter precipitation threat. Even if precipitation does occur, evaporation can prevent it from reaching the surface. Often dry air wraps into the mid-latitude cyclone, cutting off precipitation in the cold sector.

QUESTION 7: What mesoscale influences are the synoptic models not going to pick up?

Mesoscale influences can be described as big weather differences over a relatively small area. These are well known in severe weather situations in which one neighborhood can experience severe damage from wind and hail while an adjacent neighborhood has little damage. A similar type of issue can come about in winter weather situations. Elevation changes can cause precipitation amounts to vary greatly over a small distance. Convective elements or banding may occur that cause some locations to receive heavy snow while surrounding locations have less snow. Distance from major bodies of water can also have a significant influence on snow accumulation. Each forecast city has unique mesoscale challenges. Forecasting in an area for a couple of years helps a forecaster learn these challenges. It certainly helps to interview forecasters that have had experience forecasting in a certain area as a starting point to learning about these mesoscale challenges.

Topic B: Freezing Drizzle/ Freezing Rain/ Light Mix

Many winter weather situation do not require a mid-latitude cyclone. These events can still be significant since a small accumulation of ice can cause travel problems. What these events have in common is that the ground surface temperature or wet-bulb temperature is at or below freezing and, in addition, liquid precipitation is expected to fall in the form of drizzle, rain or a mix. An example for a set-up for this situation is for a polar air mass to be in place with the addition of the advection of moisture moving in aloft. The increasing temperature and moisture aloft can be caused by Warm Air Advection (a.k.a. Isentropic lifting, return flow) from the Gulf or Atlantic. Slang used is “overrunning” since differential advection with height is occurring (cP air in PBL, mT air above this layer). If return flow aloft (Warm Air Advection) is too strong OR the surface temperatures are at freezing, freezing drizzle/rain will change to all rain shortly since the warmer air aloft will mix out the colder temperature at the surface. A Topic B set-up can also occur when a shortwave associated with the subtropical jet transverses Polar air at the lower levels. If temperatures are cold enough, light snow or winter precipitation can result. A Topic B situation can lead to historic winter storms when the freezing rain is heavy and lasts for hours. These events can shut down cities for a significant period of time and can cause severe damage and travel headaches. The following image shows a general freezing rain profile:


Topic C: Lake Effect Snow

See the following two excellent writings:

Topic D: Sleet (Ice Pellets)

Sleet (ice pellets) is a rather interesting precipitation type. Sleet results when a snowflake partially or mostly melts aloft but then refreezes before striking the ground. The end result looks like a frozen water drop that makes a distinct pinging sound when striking the ground. This precipitation type often occurs in the transition from rain to snow or vice versa. Thus, sleet will occur with mid-latitude low pressure winter storms in typically a narrow band in the transition from rain to snow. In some situations though, the entire storm can be sleet from beginning to end. A common situation in which the entire storm can be sleet is to have a layer of polar air at the surface, with moist air advecting over this layer. A shortwave or upper level low can also instigate the precipitation development. There has to be an above freezing layer of air aloft. Without this layer, the snow would not be able to partially melt. The layer is not warm and deep enough though to allow the precipitation to completely melt. An ice nucleus must be left in order for the drop to freeze when it falls back into a subfreezing layer. If the snowflake completely melts, then what will be experienced at the surface is freezing rain (when ground is at or below freezing). If the snowflake does not melt much then what will be experienced are snow pellets. Snow pellets are crusty snowflakes. So little melting occurs that the snowflake becomes a little icier instead of looking like a frozen raindrop. Snow pellets also appear whiter than ice pellets. A snow pellet has characteristics of both sleet and snow. If the layer of warm air aloft (by warm air it is meant a layer aloft that is above freezing) has the right depth and temperature profile, sleet will be generated that can last for hours. In some cases, there can be vigorous dynamic lifting in place from a shortwave and warm air advection that can cause convection to occur and the generation of thundersleet. These situations can produce thunder and heavy accumulations of sleet.

Glazed Tree Branch After Winter Ice Storm, Snow and Frozen Rain, Icicles

Tools Used to Forecast Winter Precipitation

Tool A: Overview of Analysis Charts

Examine the latest analysis charts to get a feel for the weather pattern you will be dealing with for the event. On the Internet, a basic chart for each level is adequate. These include the surface, 850 mb, 700 mb, 500 mb and 300/250 mb analysis. This gives you a feel for the important synoptic influences on the weather.

Tool B: Graphical Synoptic Models

A good first step is a general overview of the synoptic short term (12 to 48 hr) graphical synoptic models. Several of the most widely used graphical models are the NAM, GFS, ECMWF, CANADIAN, and RUC. A location to view the output from each of these models is given below:

Here are some important questions to answer while putting together your winter weather precipitation forecast:

1. What types of thermal advection are taking place at the various levels in the atmosphere?

Discussion: Thermal advection at various levels will change the temperature structure of the atmosphere over the forecast region. This is important to a precipitation type forecast. WAA will lead to rising air while CAA will lead to sinking air. Other lifting or sinking mechanisms will also have to be studied to access their significance.

2. Is the atmosphere likely to produce precipitation that will reach the ground over the forecast region?

Discussion: See the precipitation prognostications from the various models and compare them. Infer what the risk is of precipitation evaporating before reaching the ground (can occur if UVV is small and PBL is dry). A forecasted accumulation of 0.50 inches is much more noteworthy than a 0.01 inch accumulation, especially in a short term forecast. Light accumulations are much more bust prone than heavier accumulations since it is easier to eliminate light accumulations than it is heavy accumulations. In either case, it is critical to examine changes in each model run just before the event occurs.

3. Are surface temperatures below freezing during the precipitation event?

Discussion: Keep in mind that precipitation prognostications show accumulated precipitation over a 6 or 12 hour period. Since the temperature occurs at the valid time but precipitation is cumulative, freezing temperatures at the valid time does not mean the precipitation that occurred during the previous 6 or 12 hours has to be of the frozen variety. Use the surface and 1000-mb prognostications for low level temperatures and compare the models. This is a common forecast mistake when this is missed: to have winter precipitation, the precipitation must occur when the temperature profile supports winter precipitation.

4. Will the precipitation fall as all snow?

Discussion: If the surface and 850 mb temperatures are below freezing, snow is the most likely precipitation for locations near sea level. Compare the models.

5. Is there a potential for freezing rain or sleet?

Discussion: An 850 millibar temperatures of 4 C or more above freezing but with 1000 mb and the surface temperatures below freezing are an indication freezing rain or sleet is more likely than rain or snow. Freezing rain and sleet events occur when a shallow polar air mass differentially advects under warmer air or when warmer air differentially advects over a subfreezing boundary layer. Sleet also occurs in the transition between rain and snow. The depth of the PBL subfreezing temperatures, and the depth / temperature structure of the above freezing layer aloft differentiate freezing rain from sleet. Compare the models. The following images show the general sleet profile and the general freezing rain profile and the thickness/temperature characteristics:



6. What is the evaporative cooling potential?

Discussion: Precipitation falling into a dry layer will result in a cooling and moistening of that layer. An above freezing surface can become below freezing when melting and evaporation from precipitation cool the air at the surface. This cooling can also change the precipitation type. Cold rain can change to freezing rain or sleet while freezing rain or sleet can change to snow when there is the potential for further evaporative cooling.

7. What is the timing of the precipitation event with the thermal structure of the atmosphere over the forecast region?

Discussion: Timing of precipitation with the vertical temperature structure of the atmosphere is critical because it determines the precipitation type that will fall. Compare the models.

8. What are some problems the models may have?

Discussion: Surface snow/ice cover and lack of data over ocean regions can distort model output. See if the models are moving toward a consensus or are diverging. If they are diverging, deduce what important process is complicating the weather situation. Also, the models may not handle mesoscale influence such as elevation changes and mesoscale gradients of surface snow/ice cover.

Tool C: Numerical Output


Online location:

MOS stands for “Model Output Statistics”. Basically what it consists of is forecast data for a point location and gives data in 3-hour forecast increments. The data includes information such as temperature, dewpoint, cloud cover, wind information, probability of precipitation, precipitation type and visibility. Examine forecasted temperature, dewpoint and wind trends. See how closely the actual observations are matching to the MOS numbers as the forecast period goes by. From the MOS, notice the similarity and differences in precipitation type, precipitation chance, precipitation amount, temperature trend, dewpoint trend, and other relevant factors.

Tool D: Forecast Soundings

Forecasts sounding use the synoptic forecast models, thus any issue impacting the forecast models will also impact forecast soundings. The power of forecast soundings is that it allows for examining detailed parameters for a specific location beyond what is given in MOS data. For example, a detail of temperature, dewpoint and wind information is given with height along with wet-bulb temperature values, an ability to examine temperature/moisture/and lifting in the snow growth region aloft, examination of thickness criteria for precipitation type, and the ability to see a graphical (a more visual) display through time. An example of software that is used for forecast soundings is BUFKIT:

Tool E: NOAA Forecasts and Public Forecasts

Forecasting knowledge and accuracy will not improve unless time is spent EACH DAY studying relevant meteorological data. Uninformed readings of MOS (Model Output Statistics) data and “copying” NWS (National Weather Service) forecasts will not develop forecasting skill since a person is depending on someone else’s forecast. In this case it would be NOAA (National Oceanic and Atmospheric Administration) employees and computers. It is best to use MOS and the NWS as forecasting TOOLS, but they should not be the sole basis of a forecast.

Time should be spent looking at meteorological data BEFORE reading NWS, public forecasts or MOS data. MOS data and NWS data should be examined AFTER you have come to your own conclusions of what weather to expect. The forecast can then be adjusted if a trend in the MOS data or forecast information covered by the NWS differs from your initial “trial” forecast.

Time spent looking at data will vary with the type of weather in the area and the forecast responsibility incumbent on a person. Some of the most important data to study before formulating the forecast include: (1) Analysis charts for each level in the atmosphere, (2) Thermodynamic and forecast soundings nearest the forecast region, (3) Comparing model initiation to the analysis charts, (4) 4panel NAM/GFS and other model data, and (5) current synoptic scale surface charts (temperature, dewpoint, pressure, wind, frontal position, and so forth).


There comes a stage when forecasting the event tapers down because the event is starting. Nowcasting is monitoring minute by minute data to observe, adjust expectations and communicate what is happening and what could happen shortly.

Satellite and Radar are a powerful nowcasting tool since they are updated frequently and include data that is currently happening or will shortly happen. Pay close attention to the latest satellite and radar when a winter storm is in the developing stages. Radar intensity trends can give an indication whether the storm is strengthening or weakening. For example, radar can be used to note where the heavier snow bands are setting up.

When observing upper level lows, it is of great benefit to use satellite data. As soon as the clouds become brighter white in association with the upper level low, then the low is deepening. Also use satellite data to see if moisture or dry air is wrapping into the upper level low. Since upper level lows are cold cored, they can produce snow in the winter even when the surface temperatures are initially well above freezing. Locations away from the low are more likely to see rain instead of snow when temperatures are well above freezing at the surface.

Satellite and Radar data are tools to use in locating where the heaviest winter precipitation is falling. These tools can give an early indication if the forecast is going according to plan. If not, be prepared to tweak the forecast. Look for mesoscale processes that can increase precipitation intensity (upslope flow, orographic lifting (Rockies, Appalachians, higher elevation regions east of the Great Lakes), elevated moisture, temperature and moisture mesoscale boundaries).

Monitor the surface temperatures and dewpoints hour by hour for temperature advections, moisture advections, wet-bulb cooling, diabatic heating and cooling (radiational cooling, daytime heating). Monitor hour by hour wind speed and direction for dry air advection which can increase wet bulb cooling, and since wind speed helps determine the magnitude of temperature advection.

Read National Weather Service special discussions / watches /warnings and keep up with minute by minute changes that can occur in the weather data such as frequently updating the surface chart and radar/satellite output. Examine new data as it becomes available such as analysis charts, model panels, and Skew-T’s.


This writing focuses on concepts to keep in mind when forecasting winter precipitation and discusses several tools that forecasters use when developing a forecast and monitoring a winter storm. Forecasting winter storms is very challenging due to the complexities involved with forecasting precipitation type, characteristics of the precipitation, accumulation and the variety of factors that influence precipitation location and amount across the forecast region. These forecasts are highly visible to the public thus forecasting mistakes are much more prone to be noticed, especially if they are very different than anticipated. Each location across the country has a variety of mesoscale influences on a winter weather event, thus knowing the local area such as elevation variation, latitude variation, and large water body influences are important to forecaster success. Each event will have a unique set of challenges but forecaster experience and the proper use of forecasting tools will increase the likelihood of forecaster success. It is a good idea to skim this article before each winter weather event to keep all the questions, concepts and tools in mind that are mentioned in this writing. There are unique concepts and tools that you can add to the information in this writing when making your forecast. There is no one right way to make a forecast but the ability to learn from mistakes, challenge yourself and learn new skills/technology will only be of long term benefit to your forecasting success.

Similar Posts:

Tagged , , , , , , .