On January 7th, 2015, an unrelenting moisture plume combined with the right synoptic setup to produce an orographic cirrus event in the lee of the Appalachians. The cirrus shield featured a stationary back edge spanning from Delaware southward through North Carolina.
To preface, we reference an orographic cirrus forecasting guide made for the staff here at WFO Raleigh based on local office research as a primer as well as an example of what a typical atmospheric sounding looks like during an orographic cirrus event.
The 500 mb analysis for 1200 UTC on January 7th, courtesy of the Storm Prediction Center, shows a nearly classic pattern supportive of orographic cirrus consisting of low pressure to the northeast and a ridge to the west. While the location of both the low and the ridge are not exactly ideal, (the low is further north and the ridge is further west than desired) the resulting westerly flow over the spine of the Appalachians is good enough to support cirrus.
In addition to the synoptic pattern, the analysis shows us strong winds aloft (>100 kts in most places) which is usually enough to satisfy the criteria of wind speed increasing with height. In order to know for sure we look at the 1200 UTC observed sounding from RNK and see that not only is speed increasing with height from the inversion up to the tropopause but, it is fairly unidirectional as well with some slight backing with height which is a common occurrence in most orographic cirrus cases.
In addition to the winds, the sounding also shows us a strong temperature inversion near 800 mb extending up to 700 mb which is in the target height for the inversion near mountain top level that is necessary to create the standing wave pattern needed to produce orographically induced cirrus. Finally, the sounding shows sufficient moisture aloft to produce cirrus clouds. A quick check of the IAD sounding shows similar features to those described on the RNK sounding, thus confirming the proper environmental conditions needed to produce cirrus.
With the synoptics and environmental conditions in place, an animation of the GOES infrared satellite imagery (top of the page) from the morning of January 7th does indeed show an orographic cirrus event setting up around 0700 UTC and lasting through the morning and into the early afternoon before the stationary back edge of the cirrus begins to break down. An animation of the GOES water vapor imagery shows clearly that there is a steady plume of moisture west of the mountains that is gradually progressing southward through the event, which resulted in the southern edge of the cirrus eventually ending up in North Carolina instead of just being confined to the Mid-Atlantic region. The cirrus overspread most of the Delmarva Peninsula, a good portion of VA east of the mountains including Roanoke, and the Piedmont Triad Region of North Carolina.
So how did the cirrus affect the forecast? From the map below we can see that Greensboro, NC (GSO) was the only one of four major forecasting points (GSO, RDU, RWI, and FAY) in the WFO Raleigh county warning area to be completely under the cirrus shield.
Temperature verification statistics from 1200 UTC 7 January 2015 for the Point Forecast Matrix (RAH forecast), the MET (NAM MOS), the MEX (Extended GFS Guidance), and MAV (GFS MOS) guidance for the four forecasting sites mentioned above shows that GSO was the only site to experience a cold forecast bias in four runs prior to verification. In the case of the MEX and the MAV it was 4 degrees too cold 24 hours out. This makes sense with the unforeseen extra cloud cover. The MET guidance seemed to handle the situation the best and this is logical because of the better resolution in space and time. Data from RDU and RWI, which were on the edge of the cirrus, show that there is still a cold bias in the GFS guidance products, but not as pronounced. FAY, which was the only station clearly removed from the cirrus, did not have any cold bias in the guidance whatsoever and in fact, the MEX and MAV had a perfect forecast 24 hours out.
The interesting take from this event for me personally is that prior to the enhanced cirrus development, the MAV forecast was already calling for broken to overcast skies 24 hours out. Despite that, the development of the more widespread and more opaque cirrus shield over the area contributed to the four degree cold bias despite an already cloudy forecast. It would have been interesting to see how much of a cold bias there would have been if the forecast had called for clear skies. Prior case studies have shown differences in temperature of up to ten degrees under an orographic cirrus shield with otherwise clear skies.
For more on the orographic cirrus research conducted at WFO RAH, please see the poster located here.