Pennsylvania Ice Storm 08-09 January 1999
1. Introduction:
On 07 January 1999, model forecasts suggested a wintry mix of snow; sleet and freezing rain would strike central Pennsylvania on 8-9 January 1999. The first snow bands developed in northern Maryland around 0500 UTC (midnight) and slowly moved northward into Pennsylvania. A series of three snow bands would form along the Maryland border and move northward into Pennsylvania during the morning of 8 January. After the last of these bands moved northward and weakened, there was a lull in the snow.
During the afternoon and early evening hours, a second surge of clouds and precipitation moved northeastward toward Pennsylvania. Sleet arrived in southwestern Pennsylvania around 1900 UTC (2 PM) and spread northeastward. The precipitation changed to rain west of the mountains. By 5 PM the sleet had reached the Williamsport area. The precipitation began to mix with freezing rain and changed to all freezing rain at most locations shortly after midnight on 9 January. Temperatures briefly went above freezing early Saturday, however, cold air moved from west to east across the Commonwealth, re-freezing the slushy mess.
Most locations received between 2-4 inches of snow. A few locations, mainly in the higher elevations of the Cambria and Somerset Counties received 4-8 inches of snow. In addition to the snow, many locations received 1 inch of sleet and 0.25 to 0.50 inches of freezing rain.
This page will show how the event was forecast and how it evolved. Currently, WSR-88D radar imagery and Eta model forecast soundings are used to describe the event.
We plan to add more imagery over the next few days.
Method:
All data were collected in real-time at the National Weather Service Office in State College. All data, included WSR-88D data have been archived to 8-mm tape. In addition to standard meteorological data, 5-minute data were retrieved from ASOS sites to analyze the gravity wave.
Model soundings, showing the warm intrusion and favorable profile to support gravity wave development will be added to this page. Due to computer problems, several model times were not available.
The observed snowfall as of 1200 UTC 8 January 1999 is shown (note all values are inches *10 so 35 is 3.5 inches). Snowfall amounts were light across most of Pennsylvania. The heaviest snow was observed along the border of Pennsylvania and Maryland. Snowfall totals were in the 1-4 inch range in this area with 5 inches in northeastern West Virginia. Most of this snow was observed between 0600 and 1200 UTC. The corresponding liquid equivalent rainfall is also shown. Again, the heaviest liquid equivalent was in near the Pennsylvania, Maryland and West Virginia borders.
We lost most of our data on 9 January and have only limited observations for the time period from 1200 UTC 8 January through 1200 UTC 9 January. The few observations available show 3-4 inches of snow over southern Pennsylvania.
b. Snow band phase 0600-1500 UTC (0100 AM-1000 AM 8 January)
A reinforcing arctic air mass moved across the northeastern United States on 7 January. Under clear skies and light winds, temperatures plummeted into the single digits and low teens across almost all of central Pennsylvania early Thursday night, 7 January 1999. A few locations saw the temperatures fall below zero before midnight. However, clouds began to rapidly approach from the southwest. Temperatures began to rise before midnight as clouds rapidly increased and thickened.
The first snow band developed over Maryland and West Virginia and by 0600 UTC, the band was moving toward Johnstown, Bedford, and Chambersburg. By 0720 UTC this band was in southern Pennsylvania. By 0809 UTC, a band of steady light snow extended from Bedford to Chambersburg. This band weakened as another band formed to the south around 0857 UTC. By 0946 UTC (0446 AM) this second band was moving toward Johnstown and reached the Altoona area by 1036 UTC. This band weakened considerably by the time it reached southern Clearfield and Centre Counties around 1125 UTC (0625 AM). Despite this, a quick 0.5 inches of snow was observed in the State College area between 6-7 AM.
A third snow band developed near the Maryland border around 1200 UTC. By 1214 UTC (814 AM) this band was moving toward the Johnstown and Altoona areas. The band had already reached Somerset and Bedford. It intensified as it moved northward. Moderate to heavy snow must have fallen in northern Bedford County around 1303 UTC (903 AM) as the band reached up to 30dBZ intensity. The western edge of this band weakened after it moved through Altoona around 1352 UTC. Some moderate snow must have fallen from Bedford eastward to Gettysburg and Carlisle as this band moved northward. By 1442 UTC, as the remnants of the band moved through State College, all banded structures began to dissipate. Snow rapidly tapered off to light snow and flurries.
By 1649 UTC, no echoes were observed over the southwestern portions of the Commonwealth. Only weak unorganized echoes showing light snow and flurries were observed over central Pennsylvania. For the next several hours, radar echoes continued to weaken and diminish. However, satellite imagery showed a second area of precipitation moving northeastward up the Ohio Valley toward southwestern Pennsylvania. Timing tools suggested the leading edge of this precipitation would reach the Johnstown area between 2000 and 2100 UTC (3 PM and 4 PM). Observations showed mainly sleet and freezing rain on the northern edge of this band and rain farther to the southwest. When this area moved over central Pennsylvania, it produced primarily sleet and freezing rain. Most of the significant snow with this event occurred with the snow bands between 0600-1500 UTC.
Not text yet some imagery of the radar and bright band effects include:
Western PA mixed precipitation echo at 2309 UTC plan view and cross section
0008 UTC plan view and cross section
0107 UTC
plan view and cross-section.
In these images, the cross-sections show a melting layer around 5000 ft across central Pennsylvania.
Eta mean-sea level forecasts from the 1200 UTC cycle 08 January 1999 are shown. These data show that the model forecast a surface low to move up the Ohio Valley between 1200 UTC 08 January and 1200 UTC 09 January 1999 and be located over central Pennsylvania by 1200 UTC 09 January 1999. This track favored the intrusion of warm air aloft prior to the erosion of the low-level arctic air.
The corresponding upper air forecast valid at 1200 UTC, 1800 UTC, 0000 UTC and 0600 UTC are shown. These data show the rapid increase in the 850-mb temperatures in the face of strong southwesterly flow. The confluent 250-mb heights and strong jet entrance region probably played a significant role in maintaining the low-level cold air.
The 850 mb frontogenesis forecast for the period are also shown. These forecasts show the strong frontogenesis over Pennsylvania at 1200 UTC that was forecast to move northward with time. The initial area of frontogenesis at 1200 UTC was closely oriented along the observed WSR-88D bands at this time. The strong warm advection (not shown) suggested that the low-level frontal surface was acting to force ascent, which may have lead to the snowbands across southwestern Pennsylvania at this time. EPV fields from the 0000 UTC cycle showed some weakly negative EPV air in this same region in the 1000 to 850 mb level between 0600 and 1200 UTC 08 January 1999.
ii. Eta model Soundings
This was the second event of 1999 where a surface low was forecast to remain to the west of Pennsylvania. This track does not normally favor heavy snow, due to the intrusion of warm air above the surface relatively early in the event. Model soundings from the Eta captured this intrusion of warm air. Similar to the event of 2-3 January 1999, the precipitation came in to phases. A light snow phase during the initial warm frontal overrunning and than a period of moderate sleet and mixed precipitation phase. The two significant periods of precipitation were separated by minor lull in the precipitation.
The 0000 UTC 7 January 1999 Eta forecasts for Altoona, (AOO) Harrisburg, (CXY) and State College (UNV) showed that snow would arrive during the day Friday. The precipitation as forecast to change to ice pellets and freezing rain. In Harrisburg, the precipitation was forecast to change to all rain toward the evening hours 9 January. The reason for this change over can be seen in the Altoona temperature time section. The Eta forecast above freezing temperatures near 830 mb by 1500 UTC Friday, this caused the model to produce ice pellets by 1500 UTC. The lowering of the warm air caused the model to forecast freezing rain by 1800 UTC. Similar temperature time sections were made for UNV and CXY.
The 1800 UTC run of the Eta on 7 January showed the same trend. At UNV, the model forecast an elevated warm layer by 1800 UTC.
Closer to the event, during the morning of 8 January, the Eta continued to forecast a wintry mix at AOO, UNV, CXY, and Williamsport (IPT). The Eta forecast the precipitation to change to all rain at AOO and CXY while producing a large threat for freezing rain at UNV. Snowfall was forecast to be between 2-4 inches (using a 10 to 1 ratio). The cause of this remained the intrusion of warm air aloft as seen in the UNV and IPT temperature time section. (IPT section has AOO label will be fixed).
The final forecast cycle, available during the evening hours of 8 January1999 from the 0000 UTC 9 January 1999 run of the Eta continued the previous trend. The precipitation profiles for AOO, UNV, CXY, and IPT are available. It is interesting to note how little snow was forecast and how much ice was forecast at IPT and UNV. These forecasts were very successful.
It is unclear where forecasts on radio broadcasts during the day Saturday for 2-4 inches of snow as the precipitation changed back to snow originated. Perhaps from users of the older NGM and its attendant model output statistics data.
How we forecast this event:
A winter storm warning was issued during the afternoon of 7 January 1999 for a large portion of central Pennsylvania. The warnings called for snowfall accumulations of 4-7 inches of snow before mixing with sleet and freezing rain. Several counties in the lower Susquehanna Valley were in an advisory for a wintry mix.
