Once ice forms on an airplane's wings or horizontal stabilizer, all of the work that went into the original, careful aerodynamic design of the airplane begins to count for less and less. Ice adds roughness and can change the shape of airfoils, reducing the lift that the wings are generating or the downward force that the horizontal stabilizer is creating. Since patterns of ice growth are random, the aerodynamic changes can be uneven and can cause rolling or pitching movements.
Obviously, you don't have to worry about icing if you stay out of clouds and falling precipitation when the temperature at your altitude is below freezing. It's also obvious that aircraft wouldn't be a practical means of transportation if they had to stay on the ground whenever clouds covered a chilly sky. That's why pilots, weather forecasters, and researchers spend so much time and effort trying to master the complexities of icing.
And icing is complex. While the basic cause of icing is simple, many factors cause some clouds to contain huge amounts of supercooled water droplets, while other clouds that appear to be similar contain mostly ice crystals. These complexities are the reasons why fore- casting icing is notoriously difficult and why even the most experienced winter-weather pilots are sometimes surprised by ice.
Researchers are learning more and more about the details of what goes on in cold-weather clouds. This research is being turned into new kinds of experimental forecasts that should do a better job of helping pilots to avoid icing. (See below.) There is even an expectation that perhaps, 10 or so years from now, most general aviation aircraft could be equipped with on-board detectors that will help pilots to avoid icing in much the same way that on-board radar and lightning-strike detectors now help pilots to avoid thunderstorms. Some large, high-end aircraft-including airliners-already have ice-detection systems.
Over the last decade, researchers have been learning more about what goes on inside icy clouds. As often happens in science, increased knowledge has shown that things are more complicated than anyone had realized. The research is also showing that some of the things that pilots think they know about icing could be wrong. For instance, says icing researcher Marcia Politovich, pilots might believe that they don't have to worry about icing if they are flying through snow. If a cloud is made of ice crystals, or if the precipitation falling from the cloud is snow, there should be no icing danger because ice crystals don't stick to an airplane. Over the years, many pilots have flown through snow or clouds of ice crystals with no problems. The assumption has been that once the supercooled water droplets in a cloud begin turning to ice, they all turn to ice fairly quickly. The cloud is said to have glaciated. But, Politovich says, "re-search flights through icing by aircraft in different parts of the world have confirmed that most of the time when you have icing conditions, you also have ice particles. Don't let your guard down just because you have seen some snow hitting the windshield."
Politovich, who has been conducting icing research at the National Center for Atmospheric Research since the 1980s, says that researchers have even seen some cases of mixed freezing drizzle and snow falling from clouds. In other words, snow isn't a guarantee that you won't run into supercooled water drops either inside or below a cloud.
Researchers have learned that up and down movements in clouds help to determine the icing danger. Places in the cloud where the air is rising at a rate of about three to 10 feet per second can be icy, while nearby areas could be free of icing danger. This is why icing can be patchy; that is, within a single cloud, icing can be heavy in some places but absent in others.
Icing is often worse at the tops of clouds. "The rule of thumb is don't fly too close to the cloud top," Politovich says. "If you go up or down even 500 feet, it can take you out of danger." If you are climbing through a cloud and you see it beginning to get light above, be more aware of the chance of icing.
Winds blowing in different directions within a cloud and right above a cloud can encourage the growth of supercooled water droplets in the cloud. Such wind shear also creates a wavy look to the top of the cloud. "If you're flying above a cloud deck and see little rolls in the cloud, you might want to stay away from it," Politovich says.
Even the number of tiny particles of sea salt, dust, or various kinds of pollution in the air can help to determine the amount of supercooled water and ice in a cloud. Some kinds and sizes of particles in the air, such as sea salt, encourage water vapor to condense onto them to form water droplets. These are called condensation nuclei. Other sizes and kinds of particles, including some kinds of dust and pollution, encourage droplets to freeze when they hit them. These are ice nuclei.
Politovich says that ice nuclei are generally more numerous around urban areas because they are generated by human activities. In theory, this means that if two clouds were otherwise alike in terms of temperature and the amount of water they contained, the one in an urban area would offer less icing danger than the one in a remote area. That's because the ice nuclei in the urban cloud would turn more of its water drops into ice crystals. But, Politovich adds, this doesn't mean that pilots should assume that there is less danger of icing flying in the heavily populated Northeast than flying over, say, North Dakota. Too many other things go into the equation. It does mean, however, that as researchers refine the computer models used to forecast icing, they have to realize that "the model that works in Florida might not work in Bismarck (North Dakota) because of the salty atmosphere in Florida rather than dust from the wheat fields around Bismarck," Politovich says.
"I'm starting to compile information about nuclei to see if I can make any sense out of them geographically or seasonally," she says. Such data could become part of the computer models used to forecast icing.
To work well, computer models need lots of data. The newest models do a much better job of predicting the large weather patterns-storms and fronts and movements of air masses-that set the stage for icing. But none of today's ordinary weather observations attempt to directly measure the amounts of supercooled water or ice crystals in clouds.
Such measurements are the next step and are an important part of the research that Politovich is conducting. In the spring of 1999, she led scientists from several organizations in a research project to study icing around Mount Washington, New Hampshire. The research focused on testing several kinds of devices that show promise for measuring the amounts of water and ice in clouds. Mount Washington was selected as the location for the study because the various devices could measure clouds as they moved past overhead on the way to the mountain where they would engulf the weather research station on top. Direct measurements from the mountain-top station were compared with those by the devices making remote measurements.
Radar is a useful tool in analyzing airborne moisture. Politovich explains that water drops in clouds more easily absorb very short radio waves than they do longer waves. This means that radar using different energy wavelengths can measure the amount of water in a cloud. This is done by comparing how much energy at different wavelengths the cloud scatters back to the radar.
One possibility that's being studied is making dual-wavelength radar a part of the new National Weather Service Doppler radar system or the Terminal Doppler Weather Radar that are being installed at airports to watch for low-level wind shear. Radiometers, which measure the natural microwave radiation emitted by water droplets and ice crystals in the air, are also being tested. The emission characteristics of ice and water are different, Politovich says. "If we use some smarts about it, we might be able to use this, maybe in conjunction with radar, to measure supercooled water in clouds."
One complicating factor is the simultaneous presence of ice and supercooled liquid in clouds. Measuring whether clouds are made of supercooled water or ice would be a lot easier if they were almost all one or the other.
Once methods of detecting amounts of supercooled water and ice in clouds have been developed, they will likely be used first in ground-based instruments. "What we are all eventually working for is an airborne system," Politovich says. "But we need to make sure that these things work on the ground before we put them on airplanes." As with other kinds of devices, a lot of work will be needed to make this measuring equipment small enough for airplanes.
"The dream," says Politovich, "is that you have airplanes equipped with an instrument to detect icing, say, five miles ahead and at the same time transmit this information to the ground where it's compiled and combined with other data to produce icing maps, which can then be transmitted up for display in the cockpit." Once such a system is in place, she says, "if you're a pilot flying around in the goo and want to know what might be the best route, you turn on your icing display and get a vertical cross-section of your route and then put it on plan view to see if you should divert, say, 30 degrees, to get around the icing."
She says that some work has started on collecting data from the icing sensors that are now on most airliners. This data would be added to the more conventional weather data used in icing forecasting models.
But until icing information becomes available in every cockpit, Politovich has a couple of pieces of advice for pilots who have to fly in icing conditions. First, if there is a danger of icing, you should turn off the autopilot and hand-fly the airplane. She has studied icing accidents in which the autopilot kept adjusting the trim to offset aerodynamic forces created by the ice. Eventually the autopilot's limits are exceeded, it automatically disconnects, and the pilots suddenly find themselves trying to hand-fly an airplane that wants to perform some unusual maneuvers. Second, she says, "a general rule of thumb is if you are in icing, one of the first things you should be doing is to figure out how to get out of there. There is no reason to be poking around in that stuff. You might as well make life easier and get out of it."
Official icing forecasts from the National Weather Service's Aviation Weather Center (AWC) in Kansas City and various kinds of experimental forecasts from the AWC and the Research Applications Program at the National Center for Atmospheric Research are available on the Internet.
The experimental forecasts offer a glimpse into the possible future of icing forecasting and can give you a second opinion if you are planning a flight into an area of potential icing. As with any weather information from the Web, you need to carefully check the day and time that the forecast or observations were made. This is especially important with the experimental forecasts.
In addition to the forecasts, both the AWC and NCAR sites have links to material explaining the forecasts, how they are made, and what they show. Visit the AWC Web site (www.awc-KC.NOAA.gov/awc/Aviation_Weather_Center. html). From the home page, click on "AWC Official Forecast Products" and then go to "Airmets" in the "Adverse Conditions" section. Here you'll find links to both text products and helpful graphics.
Let's imagine that you are planning to fly from Dulles International Airport (IAD), near Washington, D.C., to Burlington, Vermont (BTV) and that you want to fly at around 8,000 to 10,000 feet above sea level. You could start on the AWC site with the following Boston airmet for icing:
ZCZC MKCWA1Z
WAUS1 KBOS 152045
BOSZ WA 152045
AIRMET ZULU UPDT 3 FOR ICE AND FRZLVL VALID UNTIL 160300
.AIRMET ICE...ME NH VT MA CT NY LO PA OH LE
FROM 40N PQI TO YSJ TO CON TO BDL TO 25S JST TO AIR TO CLE TO YYZ
TO YQB TO 40N PQI
OCNL MOD RIME/MXD ICGICIP BLW 100 NRN HLF AREA AND BTN 025 AND
070 SRN HLF AREA. CONDS CONTG BYD 03Z THRU 09Z.
FRZLVL...SFC-040 THRUT EXC 040-080 OVR SRN HLF VA.
You can see right away that you will have to worry about occasional, moderate rime or mixed icing in clouds or precipitation below 10,000 feet in the north and between 2,500 and 7,000 feet in the south, which would be over Pennsylvania in this case. Figuring out the exact location of the icing danger from the text is difficult. Fortunately, the AWC also supplies maps outlining the areas covered by airmets and other weather alerts.
The next step in planning your flight might be to go to the AWC's experimental forecasts page to get a couple of second opinions on the icing that you might encounter. This page is linked from the main AWC page or you can reach it directly (www.awc KC.NOAA.gov/awc/ experimental.html). Here you can scroll down to the icing forecasts. One of your choices is the neural net forecasts. These are created by an artificial intelligence computer program that can quickly sort through more data than any human forecaster could hope to and then make forecasts. The National Center for Atmospheric Research Web site also has some experimental forecasts, which are being prepared in cooperation with United Express. At the Web site (www.rap.ucar.edu/rapdemos/iidademo/index.html), you'll see a map of the United States with different air routes marked. If you click on one of these, you'll go to a cross-section showing what icing conditions to expect along your route. There are also maps that show icing dangers at various altitudes.