Road Weather: Pavement Forecasting

Jon Tarleton
Marketing Development Manager and Meteorologist
Quixote Transportation Technologies, Inc. - SSI
St. Louis, Missouri

The weather impacts us all, whether it is in our personal lives or while at work. Consciously or not, we know that every time we step outside the weather is going to affect the way we feel. Our miles of asphalt and concrete, which are our main concern, are also affected by the weather on a minute-by-minute basis. In the winter when we are concerned about pavement temperature the most, we spend most of our time trying to guess how the weather will impact our pavement, and thus, impact driving conditions.

Unfortunately, the pavement does not always behave like we expect. How can a bridge deck get colder than the air temperature? The temperature of a bridge or road surface is one of the most critical pieces of road information, so therefore we must learn how and why it changes. Why does frost form sometimes, but not always? We need to learn what to look for in the formation of frost and how it gets on our roads. Why does the pavement temperature sometimes drop dramatically at the beginning of a snow event, and why does it drop only slightly at other times?

Once we know how things change, can we forecast pavement temperature? We all know how inaccurate a weather forecast can be—so why would any meteorologist try to predict what is going to happen to an object in that weather?

The Effects on Pavement
As we all know, anything that is exposed to the weather is also somehow affected by the weather. Nothing says it better than the classic tourist trinket—a small rock on a string with the instructions, "If the rock is wet it is raining, if the rock is moving it's windy, and if the rock is gone it has been stolen." The human body and the bodies of animals are the only things that can attempt to regulate their own temperature when the weather changes. All other objects are incapable of this, and therefore change as the weather changes. Our roads and bridges are no different. The four major things that affect our pavement are the air, sun, moisture, and the amount of heat beneath the pavement.

Before we had devices to measure pavement temperature we relied heavily on air temperature as our gauge to when the pavement temperature might change. However, we have since learned through such devices as Road Weather Information Systems (RWIS) and vehicle-mounted pavement sensors that trying to gauge pavement temperature using air temperature alone is not accurate. The temperature of the pavement can be dramatically different in many cases caused by other weather parameters.

Our pavement surface is exposed on one side to the air, which means the air does impact our pavement temperature. To understand how the pavement is impacted we must have a very basic understanding of the balance of heat, otherwise known in science as Thermodynamics. The basic understanding is that every object is in a constant state of temperature change. For example, think of a table in a room. This table is constantly gaining and losing heat. Now if the room is at a constant temperature, the table eventually reaches the temperature of the room. If we lower the temperature of the room dramatically the table will lose heat until it reaches the new room temperature; and because the table is exposed to the air on all four sides the temperature change happens more quickly than, for example, a book laying flat on the floor in the room.

Surface Area: The greater the exposed surface area the faster the change. A bridge is similar to a table in that it is exposed to the air on all four sides; therefore, this is one reason why bridges and other elevated roadways can cool more quickly than pavement with ground underneath.

Another consideration for the temperature change of an object is the density and composition of that object. For example, if you have two objects—a metal one and a rubber one—outside on a cold day, and if you touch the metal handle with your bare hands, it feels colder than the rubber object even though both are at same temperature. Why? Because metal conducts heat better than rubber; therefore, the metal object draws heat out of your hand more quickly than the rubber handle, making it feel colder.

Density/Material: Certain objects gain and lose heat more quickly than others depending on density and composition of the object. A bridge that contains metal and concrete will gain and lose heat faster than a road surface. Is it possible for a bridge deck to get colder than the air temperature? Yes, because of the density of the bridge and the fact that it can lose heat on all sides—it allows for the air to act like a vacuum and draw the heat out of the bridge.

One of the biggest misconceptions with pavement temperature change is the effect that wind chill has on the pavement: Wind chill only affects the human body and has no effect on roadways, bridges, etc. Wind chill is a value developed by meteorologists to give people an idea as to how cold it feels outside. Wind does affect the pavement, but only when the air temperature is changing. As cooler or warmer air blows across the pavement it eventually changes the temperature of the pavement. If the air is getting colder because of the wind, then the stronger and/or colder the wind the faster the pavement cools.

Solar Radiation (Sun)
Regardless of the time of year, the sun is one of the biggest controllers of pavement temperature. The sun has such a huge impact on the pavement that even on a partly cloudy day, the temperature of the pavement can fall as much as 10 degrees when the pavement is shaded by clouds. In fact, even in the middle of the winter, during the day and under heavy cloud cover, the pavement is still being impacted by the sun's radiation. Elevation is another important consideration with respect to solar radiation, because the higher the elevation the higher the solar radiation.

Moisture affects the temperature of the pavement through the amount of moisture in the air, and by the precipitation falling from the sky onto the pavement. Moisture appears in the atmosphere in all three forms: as a gas in the form of water vapor; as a liquid in the form of clouds and rain; and as a solid in the form of snow and ice. The most difficult of the three to understand is the water vapor form because it is invisible to us.

Water vapor is reported by weather measurements in two forms: relative humidity and dew point. Now, while most people think they understand relative humidity, it is really quite a confusing measurement. To better understand the amount of moisture in the air and to see its effects on the pavement, dew point is the preferred measurement to use.

When the pavement temperature falls below the dew point it will develop condensation on the surface. If that pavement temperature is below freezing then frost will form on the surface. To form bridge frost we have some key conditions to watch for:

  • Clear skies overnight
  • Low-level moisture from melted snow or recent rains
  • Light winds
  • Pavement temperature below freezing
  • Pavement temperature less than the dew point temperature

If all these things occur we may still not always have frost. Why? There are two key issues we need to watch for which may hinder our frost formation, or idea that frost could exist. The first issue is, where are we measuring our pavement temperature and dew point? If we have the luxury of an RWIS system then we are measuring the pavement temperature from a sensor in the bridge and the dew point from a device on the RWIS tower. But where is our RWIS tower? How close is the RWIS tower to the bridge? Remember that an RWIS tower could be as much as 5,000 feet from the bridge, which is long enough to cause changes in dew point measurements. If we do not have an RWIS site then the most convenient location to obtain the current dew point is from a nearby airport. As you can imagine this is okay, but many times this is miles from our bridge. Most times during frost conditions the bridge temperature only drops a degree or two below the dew point, so we have little room for error.

The second issue relates to how quickly frost forms. When the bridge temperature only drops a degree or two below the dew point, condensation begins to form slowly over several hours. The farther the pavement temperature drops below the dew point the quicker the frost formation.

Condensation can also occur in a road surface as well, which sometimes is called road frost or a type of black ice. This typically happens when the ground underneath a road has been allowed to get very cold. Then when a sudden warm-up occurs (not caused by the sun) the road surface cannot warm quickly enough, because it is sitting on a "block of ice." These events typically occur late in the winter or early spring in areas of the country where cold weather has been persistent.

As you can see, if we know the current dew point and pavement temperature we can determine if and when frost or black ice will form on the pavement surface. What about moisture in the form of rain or snow? How does this affect the pavement temperature? Much like when colder air moves across our pavement surface and cools the pavement, rain and snow will typically cause cooling of the pavement. The harder the rain or snow is falling the more it increases cooling, and therefore the heat currently being held by the pavement cannot keep up with the cooling effects of the rain or snow.

The dew point has another important relationship with falling precipitation most people do not realize. We learned earlier that the farther apart the air temperature and dew point are, the drier the air. As soon as it begins raining or snowing some of the water/ice begins to evaporate back into the atmosphere, even while it is still raining or snowing. The process of evaporation requires heat from the surrounding area, so in the case of our pavement it takes the heat from the pavement. The drier the air, the faster evaporation occurs and thus, the faster the cooling process on the pavement. The best effects of this can be seen during the winter when the air is drier than it is in the summer.

Heat Beneath the Pavement
As we discussed earlier a road surface is affected on only one side by the atmosphere. Its other side is affected by the ground beneath. In an effort to predict the pavement temperature, meteorologists know that to accurately predict the pavement temperature we would need to know how much heat is beneath the pavement. As RWIS were installed across the country, a temperature sensor was added beneath the pavement to measure this heat. It turned out that subsurface temperature was one of the most important pieces of information needed to determine how the pavement would react to the weather above. When the subsurface temperature is warm, snow may still accumulate on the pavement above; but as soon as the snow stops, the heat beneath the pavement warms the surface back to above freezing, allowing for melting. When the subsurface temperature is cold, rain can fall on the pavement causing the last drop in temperature needed to freeze rain on the road. Subsurface information can only come from RWIS installed by highway and street departments.

Forecasting the Pavement Change
As we have learned, there are many variables that go into the changing of pavement temperature. But even though these processes may seem simple enough it is still difficult to try predicting these changes on your own. Don't forget your own likely impression of weather forecasting, "They never get it right!" Forecasting the weather is challenging enough; trying to predict the air temperature at 9:00 p.m. this evening can prove to be a difficult task. Now try forecasting the temperature of something out in the weather at 9:00 p.m., such as the pavement. You have to accurately predict the weather to have any chance of predicting the temperature of something out in the weather. This alone makes forecasting pavement temperature a difficult task, but certainly not impossible.

Now, since the pavement temperature can change so quickly over a given area because of the factors we discussed (air temperature, precipitation, clouds, thickness of the pavement, etc.), pavement forecasts are forced to be site-specific; for example I-80 at Exit 22. So, all good pavement forecasts are site-specific.

To create a pavement forecast a meteorologist must first create a forecast for the atmosphere above the pavement, because as we learned the weather above the road is so important to pavement temperature change. This atmospheric forecast is then input into a computer program that has an understanding of all the pavement influencers we have discussed. This computer program, known as a pavement model, then must know the current pavement temperature and the trend (is it getting warmer or cooler). The last value required is the current subsurface temperature, so that this model now has a complete picture of all the forces affecting the pavement. Now an accurate prediction of the pavement temperature can be made. Typically these forecasts go out 24 hours into the future, but could go further out if ever required by the industry. So to get one of these pavement forecasts do we need an RWIS system? Not necessarily.

Of course, having an RWIS system does make the forecast more accurate, but it is not necessary. A pretty good pavement forecast can be generated using the nearest RWIS site (typically owned by a DOT, city or airport). The meteorologist then makes an atmospheric forecast over your area of concern, and then the nearest RWIS site is used to gather the other current information; start times of rain/snow are still for your location, not the nearby RWIS site. In addition, current RWIS conditions typically only impact the first 6-8 hours of the pavement forecast; after that the atmospheric forecast has a bigger impact.

As you can see, there are a lot of forces impacting our pavement temperature and, thus, making our decision as to what will happen is difficult...but not impossible! Trying to do all this on your own is not worth your time. Knowing whether the pavement will freeze is one of the fundamental questions that you must answer to determine how you will deliver the best service. Simply trying to guess the answer may work sometimes, but it is a lot more efficient if you have a much more reliable solution. State DOTs rely heavily on these pavement forecasts to determine when to pre-treat their roads, when to schedule crews, and how much chemical will be required throughout the duration of the event. "But we are a small agency with a limited budget." Doesn't that also mean you have less room for error?

Jon Tarleton began his career with seven years of experience forecasting for RWIS clients. His weather emphasis has been in Midwest and mountain west forecasting, but has forecast for the entire United States and portions of Canada. He has been responsible for all RWIS user training and weather training for Quixote Transportation Technologies, Inc. - SSI for the past two years. His training sessions focus on using RWIS as a decision tool, understanding key winter weather information, and how important pavement forecasts are in the decision process. Tarleton obtained his bachelor's degree in Atmospheric Science from the University of Missouri. He can be reached at (314) 872-0509 or