An Investigation of Model-Simulated Band Placement and Evolution in the 25 December 2002 Northeast U.S. Banded Snowstorm David Novak NOAA/ NWS Eastern Region Headquarters, Scientific Services Division, Bohemia, New York Stony Stony Brook University, State University of New York, Stony Brook, New York Brian Colle Stony Brook University, State University of New York, Stony Brook, New York Daniel Keyser University at Albany, State University of New York, Albany, New York
Previous Work Compare Eta, MM5, and WRF forecasts to observations –Models initialized with EDAS at 0000 UTC 25 Dec 2002 –36/12/4 km one-way nest for MM5/WRF ModelSSTConvectionPBLMicro- physics Eta BMJMYJFerrier MM5 v3.4.0 NavyGrellMRFSimple Ice (3 class) WRF v2.0.3 NavyGrell–DevenyiMRFWSM-3
MSLP Time Series
12 km MM5 12 km WRF Simulated Radar Reflectivity (shaded, dBZ) 700-hPa height (thick solid, m) 700-hPa 2D Miller Frontogenesis (thin solid, °C 100 km -1 h -1 ) 1800 UTC
12 km MM5 12 km WRF Simulated Radar Reflectivity (shaded, dBZ) 700-hPa height (thick solid, m) 700-hPa 2D Miller Frontogenesis (thin solid, °C 100 km -1 h -1 ) 2000 UTC
12 km MM5 12 km WRF Simulated Radar Reflectivity (shaded, dBZ) 700-hPa height (thick solid, m) 700-hPa 2D Miller Frontogenesis (thin solid, °C 100 km -1 h -1 ) 2200 UTC
12 km MM5 12 km WRF Simulated Radar Reflectivity (shaded, dBZ) 700-hPa height (thick solid, m) 700-hPa 2D Miller Frontogenesis (thin solid, °C 100 km -1 h -1 ) 0000 UTC
4 km MM5 4 km WRF 700-hPa 2D Miller Frontogenesis (shaded, °C 100 km -1 h -1 ) 700-hPa temperature (thick solid, C) 700-hPa wind barbs 2000 UTC
Motivation Why did the MM5 and WRF models forecast the band too far to the southeast? –Is the deformation/frontogenesis farther northwest? Can the modeled sharp 700-hPa trough and attendant intense frontogenesis be verified? What accounts for the different band evolution forecasts in the WRF and MM5? –MM5: one single band that dissipates early –WRF: correct event length but two separate bands
Analyses and Observations RUC and EDAS used for analysis, with supplemental tropospheric observations DatasourceVariablesInstrument Error NOAA ProfilesWind1 kt; 3 degrees WSR-88D VADWindSituationally dependent MDCRSWind, Temp3–5 kt, 5 degrees AnalysisResolutionTechnique RUC20 kmOI EDAS12 km3-D VAR
Analyses and Observations 18 UTC RUC 700 mb Height (red, 15 m) 700 mb Temp (shaded, 2°C) Analysis Winds (white barb) Observed Winds (black barb)
RUC vs. EDAS 18 UTC RUCEDAS
Analyses and Observations 19 UTC RUC 700 mb Frontogenesis (red, °C 100 km -1 h -1 ) 700 mb Temp (shaded, 2°C) Analysis Winds (white barb) Observed Winds (black barb)
Analyses and Observations 22 UTC RUC 700 mb Frontogenesis (red, °C 100 km -1 h -1 ) 700 mb Temp (shaded, 2°C) Analysis Winds (white barb) Observed Winds (black barb)
Analyses and Observations 00 UTC RUC 700 mb Frontogenesis (red, °C 100 km -1 h -1 ) 700 mb Temp (shaded, 2°C) Analysis Winds (white barb) Observed Winds (black barb)
RUC vs. EDAS 00 UTC RUCEDAS
MM5 and WRF 19 UTC WRFMM5
MM5 and WRF 22 UTC WRFMM5
MM5 and WRF 01 UTC WRFMM5
Features of Note Sharp 700-hPa trough, attendant winds and frontogenesis can be verified Trough and associated frontogenesis farther northwest than models forecast Easterly flow forecast in WRF run over CT was not observed
Potential Vorticity High values of PV associated with –Cyclonic flow –High static stability –Low tropopause –Upper trough Low values of PV associated with –Anticyclonic flow –Low static stability –High tropopause –Upper ridge PV is the product of the –Absolute vorticity –Static stability Figures from Thorpe (1985) for Northern Hemisphere Slide courtesy Dr. Mike Brennen (NCSU)
Dynamic Tropopause 12 UTC MM5WRF Pressure and winds on the PV=2 PVU surface (shaded)
Dynamic Tropopause 15 UTC MM5WRF
Dynamic Tropopause 16 UTC MM5WRF
Dynamic Tropopause 17 UTC MM5WRF
Dynamic Tropopause 18 UTC MM5WRF
Dynamic Tropopause 19 UTC MM5WRF
Dynamic Tropopause 20 UTC MM5WRF
Dynamic Tropopause 21 UTC MM5WRF
Dynamic Tropopause 22 UTC MM5WRF
Dynamic Tropopause 23 UTC MM5WRF
Dynamic Tropopause 00 UTC MM5WRF
Dynamic Tropopause 01 UTC MM5WRF
Dynamic Tropopause 02 UTC MM5WRF
PV generated below level of maximum heating –Warming increases static stability –Pressure falls convergence increases absolute vorticity PV+ PV- PV and Latent Heating Opposite occurs above level of maximum heating where PV is reduced PV growth rate determined by vertical gradient of LHR Slide courtesy Dr. Mike Brennen (NCSU)
12 UTC Model PV - Reflectivity Comparison MM5WRF Pressure/winds on the DT (shaded) and reflectivity contoured > 32 dBZ
15 UTC Model PV - Reflectivity Comparison MM5WRF
16 UTC Model PV - Reflectivity Comparison MM5WRF
17 UTC Model PV - Reflectivity Comparison MM5WRF
18 UTC Model PV - Reflectivity Comparison MM5WRF
19 UTC Model PV - Reflectivity Comparison MM5WRF
20 UTC Model PV - Reflectivity Comparison MM5WRF
21 UTC Model PV - Reflectivity Comparison MM5WRF
22 UTC Model PV - Reflectivity Comparison MM5WRF
23 UTC Model PV - Reflectivity Comparison MM5WRF
00 UTC Model PV - Reflectivity Comparison MM5WRF
PV Cross Sections 21 UTC MM5WRF
mb PV 21 UTC MM5 WRF
PV Findings Model-simulated bands appear downwind of PV filaments PV filaments appear to be created by diabatic processes occurring in southeast sector of cyclone Simulated band evolution was particularly sensitive to diabatically-generated lower- tropospheric PV anomaly over Long Island
Conclusions and Implications Southeast band position error appears to be due to a misplacement of the sharp 700-hPa trough and associated frontogenesis Although both the MM5 and WRF successfully predicted band formation, respective band evolution appears to be sensitive to convection occurring in the southeast sector of the cyclone Suggests the likelihood of banding may be more predictable than exact timing, location, and evolution
18 UTC Radar Observations
19 UTC Radar Observations
20 UTC Radar Observations
21 UTC Radar Observations
22 UTC Radar Observations