Section 1: Project Objectives and Accomplishments
The project objectives were divided into two main goals: 1) Gain a deeper understanding of the dynamics of mesoscale band formation with extratropical cyclones, and 2) Improve Quantitative Precipitation Forecasts (QPF) for banded precipitation events with extratropical cyclones. The specific goals were as follows:
Dr. John H. E. Clark of the Pennsylvania State University (PSU) Meteorology Dept. focused more on the first main goal of understanding the dynamics of mesoscale band formation. Dr. Clark is an associate professor with the Pennsylvania State University (PSU) Meteorology Dept. The main responsibilities of the PSU Meteorology Dept were as follows:
David Nicosia focused on the second main goal of this project: improving QPF forecasting for banded precipitation events. David is a senior forecaster with the National Weather Service (NWS) in State College, PA. The NWS in State College contributed the followed toward this project:
1.2.1 Dynamics of mesoscale band formation
The results of our research suggest that mesoscale precipitation bands are formed by a vigorous mesoscale dynamical process that is initiated by meso-alpha scale mid-troposphere frontogenesis. It is hypothesized that this mesoscale process eventually leads to a down scale energy transfer and results in the formation of the snow bands. The proposed conceptual model which explains this vigorous mesoscale process responsible for mesoscale band formation is as follows (a detailed description appears in Nicosia and Grumm, 1999):
1.2.2 Applications to forecasting
The AVN (grid spacing 80 km), Eta (grid spacing 48 km), mesoscale-Eta (grid spacing 29 km) and fine scale MM5 (grid spacing 12 km) model simulations were used to perform our analyses. A key finding in this project was that the mesoscale models (mesoscale-Eta and MM5), while not resolving snow band formation, simulated the meso-alpha environment that was conducive to band formation. Significantly coarser scale models, like the AVN and synoptic Eta, did not simulate these meso-alpha features mainly because they could not resolve the strong horizontal temperature gradients associated with intense mid-troposphere frontogenesis. In fact, the higher resolution models produced relatively accurate QPF, which was helpful in determining where the heaviest snowfall occurred in this snowstorm. The coarser resolution models did not produce enough QPF to the north and west of the low and mid-level cyclone track toward the cold side of the cyclone. These models erroneously spared areas from heavy snow that were hard hit. This is an important point to remember when forecasting heavy snow during East Coast cyclogenesis.
An examination of model fields revealed that the coarser resolution models maintained, little if any, CSI or CI, as measured by the equivalent or moist potential vorticity (MPV) fields, to the north or northwest of the cyclone track. The mesoscale models maintained a banded region of negative MPV northwest of the cyclone track that was close to the region where the heaviest snow fell. This region was also where mesoscale band formation was most prevalent. A MPV generation term was examined for each model. It was found that the mesoscale models had 3 to 4 times more negative MPV tendencies when compared to the coarser resolution models. This was related to the more realistic thermal and moisture gradients produced by the mesoscale models. These results confirm the importance of higher resolution models for accurately simulating the processes responsible for mesoscale band formation in snowstorms. This, in turn, is critical for more accurate QPF and determining where the heaviest snow will occur operationally.
Our findings suggest that model-produced regions of mid-troposphere frontogenesis coupled with negative values of EPV are harbingers of potential snow band formation in east coast storms provided the ingredients for feedback, i.e., enhanced differential moisture advection occur. Without the latter condition, the potential for the rapid reduction of EPV does not exist.
Since frontogenesis and EPV are readily available in NWS field offices both in plane view and also in cross-sections from mesoscale models, forecasting the potential for mesoscale snowband formation is possible operationally. From our project, we developed a forecast methodology for anticipating mesoscale band formation and heavy precipitation in an extratropical cyclone. This forecast methodology is as follows:
1) Use a mesoscale model to examine frontogenesis and EPV in the lower to mid
troposphere during cyclogenesis.
2) Take a cross-section perpendicular to the midlevel isotherms of frontogenesis and EPV.
3) A deep layer of negative EPV in conjunction with a midlevel frontogenesis maximum signals the potential for mesoscale band formation.
4) Where the deep layer of negative EPV and intense midlevel frontogenesis persists the longest is where the heaviest snow or rain will fall.
This forecast methodology has been tested only on two major snowstorms over this COMET project and it worked fairly well, operationally, in anticipating where the heaviest snowfall would fall. Although we have gained considerable understanding of the processes leading to the formation of mesoscale precipitation bands, we still feel that additional work is needed on the mesoscale precipitation band mechanisms before we can be assured that our conceptual model can be transformed into an operational tool successfully.
Section 2: Summary of University/NWS exchanges
Two teletraining sessions were given to up to 10 NWS Eastern Region offices regarding the results of our research in December 1998 and January 1999. It is estimated that over 100 forecasters listened in on our presentation that lasted for 1 hour.
Preliminary results and some of our objectives were given to a PSU graduate level mesoscale meteorology class. This presentation was also attended by NWS personnel. Most of our final results were presented to a PSU graduate level mesoscale meteorology class in December 1998.
Section 3: Presentations and Publications
Nicosia, D. J., J. H. E. Clark and R. H. Grumm, 1999: The Role of Frontogenesis and Negative Equivalent Potential Vorticity in a Major Northeastern U.S. Snowstorm. Oral presentation and abstract at the 2nd AMS Isentropic Conference at Millersville University, Millersville, PA in April 1999.
Nicosia, D. J., J. H. E. Clark and R. H. Grumm, 1999: Quantitative Precipitation Forecasting for East Coast Snowstorms: Forecasting the Location of the Heaviest Snowfall. Proposed presentation at the upcoming Northeast Regional Operational Workshop to be held in Albany, NY in Sept. 1999.
Clark, J. H. E. and D. J. Nicosia, 1999: Mesoscale band formation in extratropical cyclones. Pre-print AMS 8th Conference on Mesoscale Processes in Boulder, CO in June 1999.
Section 4: Summary of Benefits and Problems Encountered
4.1 University's perspective
The Department of Meteorology benefited from the opportunity to increase our interaction with the NOAA/State College Forecast Office personnel. The Department is already using some of the software, such as GEMPAK, that was used in this project for instructional purposes. In addition, a number of valuable data sets produced in this project have been saved for use in our synoptic and mesoscale meteorology courses.
An undergraduate student, Adam Breo, participate in this project even though there was no monetary compensation involved. Adam carried out much of the GEMPAK analyses of the banded precipitation structures.
Finally, the close relationship initiated with Richard Grumm and Dave Nicosia is continuing with a new COMET Partners project.
4.2 NWS perspective
This COMET partner's project has been very useful in improving forecasts of heavy snowfall in central PA. The forecast methodology developed from this project was used successfully during two of Pennsylvania's most intense winter storms of the past two winter seasons. This forecast methodology was employed using the mesoscale-Eta model during the snowstorm of 29-30 December 1997 and 13-14 March 1999. Each of these storms produced in excess of one foot of snow over a significant part of our county warning area. Concepts gained through this project have helped forecaster's better understand mesoscale band formation which in turn has helped them utilize the mesoscale model data more effectively.
For the 29-30 December 1997 snowstorm, this forecast technique (in its infant stages of development at the time) successfully identified the potential for heavy snow from central PA to central NY (See the appendix for the forecast discussion issued by a NWS forecaster from central PA for details). For the 13-14 March 1999 snowstorm, this forecast technique was used to accurately pinpoint where the heaviest snowfall would occur. Several forecaster's received NOAA performance awards for there accurate forecast for this snowstorm despite inconclusive model guidance.
Overall, this project has improved knowledge and forecaster awareness on the processes responsible for mesoscale band formation. This applies not only to the central PA office but also to other adjacent offices that have absorbed some of this research through various means (teletraining workshops, forecast coordination between offices and conference presentations).
TTAA00 KCTP 300036
CENTRAL PENNSYLVANIA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE STATE COLLEGE PA
738 PM EST MON DEC 29 1997
HANG ON TO ALL WRNGS AND LEAVE TOTAL SNOW AMNTS AS IS.
ADVY LOOKS GOOD IN SRN LWR SUS VLLY.
SNOW HAS BEEN SLOW TO SPREAD INTO C-PA AND WE REALLY ARE IN A
HOLE ON THE RADAR COVERAGE HERE. HOWEVER... ECHOES FILLING IN
QUICKLY AND THIS STORM HAS A HISTORY OF HVY SNOW HEADING RIGHT UP
THE APPLCNS. UP TO 13 INCHES IN NE WV AS PER WBCPNSWBC. WV
IMAGERY SHOWS THAT THE MAIN UPR LVL LO WAS STILL BACK OVR ERN
TENN AND ERN KY. ANTHR LOW-LVL WAVE WAS RIPPING NE UP THE COAST
AND SUPPORTG MOD-HVY RAINS ACR NJ/ERN PA INTO SRN NEW ENGLAND W/
SNOW ON THE NRN/WRN FRINGE OF THIS PRECIP SYS. THIS LOW-LVL WAVE
WAS WELL NE OF THE UPR LVL LO AND BELIEVE THAT IT WILL RUN IT/S
COURSE DETACHED FM THE UPR LVL LO. CYCLOGENESIS MEANS THAT THE
WESTWARD TILT OF THE LOW W/ HGHT DECREASES W/ TIME AS THE UPR AIR
LO CATCHES UP TO THE SFC LO. IF THE CYCLONIC SPIRAL OFF THE
E-COAST BECOMES THE "MAIN" SFC LO... THAN THE UPR LVL LOW OVR
KY/TN WILL HV TO START MOVG MUCH FASTER E REAL SOON WHICH IT IS
NOT (AS PER THE WV IMAGERY). THE MESO-ETA/ETA... NGM AND AVN ALL
INDICATE THAT THE 500 MB LO TRACKS TO WV/WRN VA BY 0600 UTC AND
TO NE PA BY 1200 UTC DEEPENING IT TO ARND 515 DM OR SO. THE STORM
REALLY HASN/T COME YET FOR PA... IT WILL ARRIVE W/ APPRCH OF THE
UPR AIR LO LATER TONIGHT AND IT/S SUBSEQUENT DEEPENING.
ANTHR SFC LOW WAS FORMING OVR NE NC AND SE VA. THIS LO I BELIEVE
WILL LINK UP WITH THE ABV MENTIONED UPR AIR LO AND SUPPORT THE
MAIN SFC LO N THRU NJ... INTO WRN NEW ENGLAND/ERN NY STATE BY TUE
MORN. THE 850 AND 700 MB LOWS WILL SPIN-UP AS A RESULT AND TRACK
ACRS ERN PA... PUTTING CNTRL PA (ROUGHLY JST/AOO NE TO BGM/ELM
AREA) IN THE DEFORMATION ZONE. THE MESO-ETA SOLTN SUPPORTS A BAND
OF INTENSE FGEN FRCG IN ASSCTN W/ DEEP LAYER CSI DVLPG TO THE N
AND W OF THE 700 MB LO TRACK. THIS FRCG SUPPORTS ALMOST 40 MB/SEC
ACRS C-PA FOR SEVERAL HOURS LATER THIS EVE AS THE DEFORMATION
BAND OF HVY SNOW APPEARS TO BE PIVOTING IN THE LOCAL AREA.
ALSO... THE MESO-ETA QPF HAS 1.4 INCHES IN C-PA. THUS... I REALLY
WOULDN/T DROP AMNTS AT THIS POINT EVEN THOUGH LITTLE SNOW HAS
FALLEN YET IN THE CWA. THE MESO-ETA ACTUALLY BRINGS IN ALL THE
HVIER PRECIP TO CWA BTWN 0000 AND 0300 UTC AND CONT/S IT UNTIL
ABT 0900 UTC-12 UTC.