Here is some study done by Charles (Dutch) Owen on alt.games.microsoft.flight-sim a couple years ago:"Where does ice come from in the simulator? From the 'icing' attribute thatis assigned to each atmospheric layer in a given simulation setup. The otherattributes you might think would bring on icing like cloud cover andprecipitation don't have anything to do with it. If the icing attribute isset (whether by hand in your own set-up or automatically by the "realweather" option in MSFS or ActiveSky for example) there'll be icing wheneveryour airplane is in that layer (as long as the temperature at altitude isbelow freezing), and if it's not set there won't be any icing regardless ofhow likely icing looks like it should be from all the other parameters.It's crucial to understand this: the icing attribute governs icing. And itgoverns the entire layer it's defined in. The only other thing that mattersat all is temperature -- for icing to occur the temperature at the altitudeof the airplane must be below freezing."The icing attribute has several levels of severity: none, trace, light,moderate, and severe. This intensity level controls how fast ice build-upwill occur, and nothing more. It has nothing to do with how likely you areto encounter ice -- you'll do that whenever you're in any layer than has anintensity above "none" and a temp below zero, and you'll accumulate ice theentire time you're in that layer. Whether you'll notice anything or notdepends on how high above "light" the intensity is set, and thesusceptibility of your airplane to ice effects (they are not all the same!)."This is the first part of the problem. As you know, "real weather" (whetheror not it's built in or third party) sets each weather station based on thecontents of its most recent METAR, downloaded from servers on the internet.These METARS do not normally include icing information. So, the "realweather" programs make logical deductions based on the temperature andmoisture conditions and set the icing attribute based on a combination ofeducated guesses and random factors. It seems from observation of thebuilt-in "real weather" system that it shies away from assigning the iceattribute at all; and when it does, it rarely assigns moderate or severeicing. The result is that while flying the pilot simply isn't likely toencounter serious structural icing hazards. This is not entirely wrong:serious icing isn't a common everyday occurrence in real life either. Thisrarity wouldn't be a problem if the actual effects of the relativelyunusual moderate or severe icing areas were realistic."scott s..

bj,I can reproduce it, so I'll file a bug report.Thanks,