|
|
|
An interview with leading icing expert Ben C. Bernstein Interview by Scott C. Dennstaedt (March 2006) In a blink of an eye convective SIGMETs are not stalking us any more and our memories of dancing around thunderstorms have turned into visions of AIRMETs for icing as the cold air from Canada plunges south. Over the next five or six months, we'll be looking directly into the countenance of winter. Not to worry, the hottest new icing product on the web, called CIP Severity, became operational on December 6, 2006. Former National Weather Service meteorologist, instrument flight instructor and IFR magazine contributing editor, Scott Dennstaedt provides a sneak preview of this newest National Weather Service product in the December issue of IFR magazine. “An analysis of icing intensity has been long overdue," says Scott who teaches aviation weather to instrument rated pilots all over the country. While AIRMETs and SIGMETs still provide the official icing forecasts, the Current Icing Product or CIP provides a better spatial and temporal resolution than the traditional advisories produced by the Aviation Weather Center in Kansas City. Scott further adds, “It is refreshing to see an icing intensity product that attempts to highlight the regions of airspace that instrument pilots should be sure to avoid.” During the Spring of 2006, Scott had the pleasure to visit with meteorologist and icing expert, Ben C. Bernstein. At that time, Bernstein was one of several meteorologists at the National Center for Atmospheric Research or NCAR in Boulder, Colorado who was responsible for the research and development behind CIP. “Icing is a complex problem. There are so many factors that play a role in what makes or breaks an icing situation,” says Bernstein. He further emphasized, “I am not totally fond of rules of thumb with respect to icing. There are some that are descent, but there are so many exceptions to every rule of thumb that they can be dangerous.” Like many icing researchers, Bernstein prefaces many of his statements with “it depends.” He further explains that icing comes down to a competition. Once you get into subfreezing temperatures and there are clouds present, there’s a competition between the production of ice crystals and the production of supercooled liquid water. He offers this simple model. If there is sufficient lift and a cloud is formed, there are two possible outcomes; it will either become a cloud containing mostly supercooled liquid water or will contain mostly ice crystals. The latter is a glaciated cloud and does not represent a serious icing threat. Depending on the specific scenario, both ice crystals and liquid water can coexist in the same cloud. As the cloud continues to develop there’s a competition between the existing ice crystals trying to deplete the liquid water present in the cloud. In other words, the ice crystals may grow at the expense of supercooled liquid water droplets. Lower liquid water content in the cloud means a lower risk of icing in that cloud. However, “The devil is in the details,” Bernstein acknowledges. Most of the time precipitation falling from a cloud also tends to deplete the liquid water content in it. Snow is one of the best examples. Snow falling from a cloud or into a cloud tends to deplete liquid water in that cloud or in clouds below it. Therefore, if snow if observed at the surface, it tends to diminish the icing potential in the clouds producing the snow, but does not eliminate the possibility according to Bernstein. Scott asked him, “If it’s snowing at the surface, is it safe to launch into the clouds?” Bernstein emphatically says, “No, that’s not a good choice.” Ice pellets (sleet) and snow grains or graupel falling at the surface tends to be a good indicator of icing aloft. Both ice pellets and snow grains are produced by liquid water “riming” on the snowflake or ice pellet. Bernstein says, “If there’s riming, there’s liquid water in the cloud.” “Rain, on the other hand, is a mixed bag,” says Bernstein. “On some days, rain at the surface means there’s nasty stuff around and other days, rain might be depleting the liquid water content in the cloud.” What about a plain overcast sky with no precipitation whatsoever? Bernstein indicates, “Now you’ve eliminated the depletion mechanism, but you may have also eliminated production because those clouds are not getting enough kick to get much going. Even the most harmless looking stratus layer may be loaded with water or it may be rather weak and very thin with not much going on…in other words, it depends.” So the $64,000 question Scott posed to our expert is, “Are there any weather products a pilot can use to make this determination between a harmful cloud loaded with ice and a harmless cloud with little or no icing potential?” Bernstein answers, “That’s what we try to do in CIP. We first try to identify that a cloud exists and then we try to get a handle on what’s happening in that cloud. What are the mechanisms that relate to -- ‘it depends.’ What temperature is it occurring at? How deep is the cloud? What’s the cloud top temperature? Is there precipitation falling and what type is it? How intense is it? How strong is the lift in the cloud?” All of these things add up to how likely it is for supercooled liquid water to be in the cloud and how much is likely to be there. “That’s what CIP tries to do,” Bernstein continues. “CIP tries to digest all of these things to give you an answer that is somewhat credible, ergo, there’s a 30 percent chance of icing as opposed to a zero percent chance.” Even though CIP provides some help, pilots need to understand that icing prediction is a complex problem. Even rather benign-looking situations can turn out to be problematic. Bernstein watches pilot reports every day during the fall, winter and spring and is drawn to severe pilot report to monitor CIP to determine what needs to be improved. Scott asked him about an icing accident earlier this year near Birmingham, Alabama where a Cirrus SR22 ATP pilot had to activate the Cirrus Airframe Parachute System or CAPS after he lost control of his airplane while trying to get on top of cloud layer. Bernstein commented, “There were clouds with supercooled liquid around, but it didn’t look spectacular, meaning a lot of supercooled liquid. The cloud layer didn’t look that thick or particularly wet.” For this particular accident, a strong cold front with thunderstorms moved through the Birmingham region about 18Z ushering in much colder air. At 18Z, the freezing level was about 9,000 feet. Just before his departure around 21Z, the freezing level had dropped to 4,000 feet and there was an AIRMET for icing from 3,000 to 8,000 feet. The accident occurred inside the AIRMET, but right on the edge of the AIRMET boundary. According to Bernstein, post-cold frontal icing can occur when there is a low level destabilization of the atmosphere that can kick up the water content in the swaths of stratocumulus clouds behind the front. This is all assuming that the temperature is just right. The temperature in the icing layer near Birmingham was about -4 degrees Celsius which is a perfect for icing, especially clear ice. Bernstein then pointed out, “The instability helps to generate condensation or liquid water if the cloud has some depth. Upslope flow tends to give the clouds a little “juice” or boost to the clouds as well. However, the clouds in Alabama on that late afternoon did not look like it had a lot of depth.” Sometimes it call comes down to how well the pilot prepares before the departure. We don’t know exactly how this pilot prepared prior to launching into known icing conditions, however, the NTSB did state that the probable cause in part was “the pilot's inadequate preflight planning and failure to obtain a current weather briefing.” Experienced or not, knowing how to do a complete self briefing is paramount. Scott’s next IFR article, “Pump Up Your Internet Weather Brief” will provide some great tips for even the most seasoned instrument pilot. Determining that icing exists in a cloud is quite complex; determining the amount of icing in the cloud is either incredibly difficult or fundamentally impossible in some cases. Nevertheless, the sole purpose of the new CIP Severity is to provide pilots with a way to assess the location and altitude of supercooled liquid water that represents a serious icing threat. CIP Severity also offers a masked variant that combines the CIP probability field with the intensity so pilots can quickly assess those regions that have the highest probability of moderate or greater icing. This product masks out those lower probability events to only show areas that have a high likelihood of containing supercooled liquid water. Additionally, the icing intensity levels of trace, light, moderate and heavy are preserved. The term “severe” is not used since it is reserved for how the aircraft reacts to the meteorological conditions, not the meteorological conditions themselves. Icing severity is primarily a function of three elements. Factors of severe icing include the liquid water content in the cloud, temperature in the cloud and drop size. While temperature is fairly easy to predict, the other two factors are not. Bernstein says, “Small errors in timing, vertical velocity in the cloud, stability and moisture initiation could mean the difference between a benign or dangerous event.” There are classical freezing rain events that remain easy to identify, but are less common than most other large supercooled water droplet scenarios. According to Bernstein, these other non-classical freezing rain events are about 9 times more common and have a more complex structure and therefore are more difficult to identify. What about the Forecast Icing Potential? Will there be a severity product available soon? Bernstein says, “It is one thing to try to determine icing severity based on observations like we do in CIP. It is a totally different problem trying to forecast the potential for severe icing based purely on model data.” Bernstein mentioned that there is a plan in place to have a FIP Severity product operational by December 2008 with an experimental product now available on the experimental ADDS web site. All of the CIP products, including the severity products, can be found on the Aviation Digital Data Service web site at http://adds.aviationweather.gov/ or on their experimental site at http://weather.aero/icing/ . Currently the operational version of CIP has a resolution of 20 kilometers. At the National Center for Atmospheric Research, Bernstein runs a research version of CIP that is 400 times the resolution of the operational version. A higher resolution version would allow pilots to see small pockets of heavier icing. At the moment, there are no plans to release this higher resolution version any time soon. Recently, Ben C. Bernstein has left his research position at NCAR. This past year, he has started his own company, Leading Edge Atmospherics and his current focus is working with aircraft manufacturers from around the globe to help them achieve certification for flight in known icing conditions. "It’s really challenging and a terrific experience working so closely with the test pilots and flight test engineers for aircraft ranging from helicopters and small prop aircraft on up to regional jets," says Bernstein. If you are still not convinced that Ben Bernstein knows his icing, see his long list of publications. Scott Dennstaedt and Ben Bernstein have teamed up recently and will be producing a one-of-a-kind training CD on icing. You can preorder this CD by visiting http://chesavtraining.com/ice_is_not_nice.htm. |
Send e-mail to
|
or call 803-802-2591 with questions about the services and/or products offered by Chesapeake Aviation Training,
Inc..