WRAL WeatherCenter Blog

What happened yesterday to make the weather report so wrong? I live in Greenville and up until about 6pm you were predicting 1-3 inches. If we got anything it was only flurries.

Posted January 22, 2008

MIKE MOSS SAYS:        Nancy,      The principal issue that turned this weekend's winter weather event into such a minor one was that cold air surging into the region from the west and interacting with the moisture and lift mechanisms that produced precipitation arrived about 4-5 hours later than initially forecast. While the nature of the weather features involved (a sharp upper level trough and strong upper level disturbance to our west, along with an intense upper level jet streak just to our northwest and a surface low developing south of us and tracking along the southeastern coast), the onset time, duration and quantity of precipitation (which ended up from around .2 inches to around .8 inches across the region) all worked out very much as they had been forecast for about two days prior to the event. However, snow accumulations that were generally forecast in the 1 to 4 inch range depended on a transition from rain to snow that would occur shortly after daybareak in western sections of the viewing area to early or mid-afternoon in your direction did not occur until a significant fraction of the eventual total precipitation had already been partitioned into liquid rather than frozen from at the surface. In the end, snowfall accumulations ranged from atrace on the eastern fringes of the viewing area to around a half inch through the Triangle area to around an inch on the western fringes of our region.

That certainly was enough to result in a lot of disappoinment on the part of everyone who was hoping for a nice snowfall, not to mention the more serious consequences of some unnecessary cancellations and changes of plans in anticipation of the event. In central North Carolina, we live in an area where forecasting wintry precipitation, while it has improved considerably through the years, is and will remain subject to significant uncertainty because in most of our potential snowfall situations, there is a delicate balance between the availability of moisture, lifting mechanisms, and the temperature and dewpoint profiles existing through the lower atmosphere across the area, along with the manner in which all these factors are modified by the proximity of the Appalachians to our west and the Atlantic to our east. This balance often results in the line (sometimes sharp and sometimes diffuse) between frozen and liquid precipitation existing in our vicinity. In the case of this weekend, there was about a 4-5 hour period in which the temperature of the lowest 3000 feet or so of the atmosphere stayed about 2-3 degrees warmer than prior computer model forecasts had indicated would be the case. On a rainy day with temperatures in the 50s, or an unusually cold (for us) snow event with temperatures in the mid 20s, this kind of temperature error is also not uncommon but has little consequence - when the error crosses the threshold region that determines whether snow or rain reaches the surface, though, it can result in the kind of forecast miss we experienced this weekend. We do try to stress that wintry weather forecasts are often subject to change due to these magnified impacts of rather small errors in temperature structure (of course there can also be errors in precipitation forecast amounts - this wasn't much of an issue with the scenario just passed, but considering the 10 to 1 average ratio of snow to liquid precipitation, an error in precipitation amount that would seem insignificant if it were only raining can make for a very large difference in snow on the ground), along with the forecast speed, direction and intensity of the features that lead to precipitation. Once a "most probable" scenario is selected, though, it is all too easy for ourselves as well as those of you using our forecasts to give less attention to the full envelope of potential outcomes.


This blog post is closed for comments.

Oldest First
View all