Mesoscale Precipitation Structures Accompanying Landfalling and Transitioning Tropical Cyclones in the Northeast United States Jared Klein, Lance F. Bosart, and Daniel Keyser University at Albany, SUNY, Albany, NY CSTAR II Grant NA04NWS David Vallee NWS Weather Forecast Office, Taunton, MA M.S. Thesis Seminar 5 July 2007

Objectives Examine the distribution of rainfall in relation to tropical cyclone (TC) track and identify smaller- scale areas of enhanced rainfall accompanying landfalling and transitioning TCs in the Northeast U.S. Identify key synoptic- and mesoscale processes that impact the precipitation distribution for these TCs. –Upstream thermal trough and downstream thermal ridge–jet interactions –Upper-level jet (ULJ) and lower-level jet (LLJ) interactions –TC-induced coastal frontogenesis –Orographic precipitation enhancement

Motivation Timing and location of mesoscale features is difficult to predict. Inland flooding is responsible for nearly 60% of fatalities from landfalling TCs (Rappaport 2000). There has been a recent increase in frequency of TC-related flooding events over the Northeast. –1950–2003: Average of 1 event every year –2004–2005: 10 events in 2 years

NPVU QPE Total precip (in.) vs. TC track: Total Precip (in.)–10 Storms Max Rainfall: 35 in.

Data and Methodology Identify TCs that produced ≥ 100 mm (4 in.) of rainfall in the Northeast U.S. for 1950– Able 1950 Dog 1952 Able 1953 Barbara 1953 Carol 1954 Carol 1954 Edna 1954 Hazel 1955 Connie 1955 Diane 1955 Ione 1958 Helene 1959 Cindy 1959 Gracie 1960 Brenda 1960 Donna 1961 Esther 1962 Alma 1962 Daisy 1963 Ginny 1969 Gerda 1971 Doria 1971 Heidi 1972 Agnes 1972 Carrie 1976 Belle 1979 David 1985 Gloria 1988 Chris 1991 Bob 1996 Bertha 1996 Edouard 1996 Fran 1997 Danny 1998 Bonnie 1999 Floyd 2001 Allison 2002 Isidore 2002 Kyle 2003 Bill 2003 Isabel 2004 Alex 2004 Bonnie 2004 Charley 2004 Frances 2004 Gaston 2004 Ivan 2004 Jeanne 2005 Cindy 2005 Katrina 2005 Ophelia 2006 Ernesto

Construct a climatology of precipitation distribution vs. TC track. –2.5° NCEP–NCAR reanalysis for synoptic diagnostics –0.25° NCEP 24 h daily (1200–1200 UTC) UPD –Higher resolution precipitation analysis produced by Ron Horwood (NERFC) –10 km RFC NPVU archived QPE –NHC best-track data Diagnose synoptic- and mesoscale processes associated with heavy precipitation for Ivan (2004) and Ernesto (2006). –Upper-air analyses and Q vector (geostrophic wind) diagnostics using 1.0° GFS dataset –Surface analyses and F vector (full wind) diagnostics using ~0.6° dataset created from GEMPAK Data and Methodology

Climatology Results LOT = left of track ROT = right of track

Upper-level downstream ridge and jet development. –Occurred in nearly every case –Placed Northeast U.S. in equatorward jet-entrance region –Amplified LLJ and positive θ e advection Enhanced precipitation as TC interacts with a pre- existing mesoscale boundary or coastal front. –Occurred in almost every case –Heavy precipitation region along and in cold sector of coastal front (CF) –Stronger θ gradient when interacting with a upstream midlatitude trough during extratropical transition (ET) Climatology Results

Possible orographic enhancement of precipitation. –Occurred in almost half the cases –Track far enough inland so that low-level easterly flow ahead of storm was upslope on the eastern sides of the Appalachian Mountains Climatology Results

Preferred Areas of Possible Orographic Precipitation Enhancement in the Northeast U.S. Blue Ridge Catskills Berkshires White

Q vector: Time rate of change of Q div–con: QG forcing for descent– ascent Q s : Time rate of change of direction of Q s div–con: QG forcing for descent– ascent within thermal trough–ridge Q n : Time rate of change of magnitude of Q n div–con: QG forcing for descent– ascent on cold–warm side of frontal zone θΔ Adapted from Martin (1999) Q Vector Partitioning in Natural Coordinates θΔ

Case Study 1: Ivan September 2004

LOT Precip Distribution NPVU QPE 09/17 09/19 09/18 Dates denote 0000 UTC positions Total precip (in.) vs. TC track: 1200 UTC 16 Sep–1200 UTC 19 Sep 2004 Ivan Case Study 1: Ivan

300 hPa h (dam), wind speed (m s −1 ), and div (10 −5 s −1 ) 300 hPa Analyses: 1200 UTC 16 September hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) 1.0° GFS

Confluent flow in equatorward jet-entrance region 300 hPa Analyses: 1200 UTC 17 September hPa h (dam), wind speed (m s −1 ), and div (10 −5 s −1 ) 300 hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) 1.0° GFS Frontogenesis in jet-entrance region

300 hPa Analyses: 1200 UTC 18 September 2004 Strengthening downstream ULJ and ridge 300 hPa h (dam), wind speed (m s −1 ), and div (10 −5 s −1 ) 300 hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) 1.0° GFS Strong frontogenesis in jet-entrance region

925 hPa Analyses: 1200 UTC 16 September hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) WSI radar, 925 hPa θ e (K) and wind barbs (kt) Pre-existing baroclinic zone Symmetric reflectivity structure 1.0° GFS

925 hPa Analyses: 1200 UTC 17 September hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) WSI radar, 925 hPa θ e (K) and wind barbs (kt) Northeastward extension of precip field along baroclinic zone 1.0° GFS Band of frontogenesis along baroclinic zone

925 hPa Analyses: 1200 UTC 18 September hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) WSI radar, 925 hPa θ e (K) and wind barbs (kt) Highest reflectivity near nose of LLJ/θ e ridge axis 1.0° GFS Strong frontogenesis along warm frontal zone

925 hPa Q Vector Diagnosis : 0000 UTC 18 September 2004 Radar at 1200 UTC 17 September 2004 QQnQn QsQs Radar at 0000 UTC 18 September 2004 Highest reflectivity near strongest QG forcing for ascent Radar at 0000 UTC 18 September 2004 WSI radar Q s div–con couplet within thermal trough–ridge Q vectors (10 −10 K m −1 s −1 beginning at 2.5 × 10 −11 ), θ (K) contoured in green, and Q div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors Q n div–con bands within frontal zone 1.0° GFS

Radar at 1200 UTC 17 September 2004 QQnQn QsQs Radar at 0000 UTC 18 September 2004 Highest reflectivity near strongest QG forcing for ascent Radar at 0600 UTC 18 September 2004 WSI radar 925 hPa Q Vector Diagnosis : 0600 UTC 18 September 2004 Q vectors (10 −10 K m −1 s −1 beginning at 2.5 × 10 −11 ), θ (K) contoured in green, and Q div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors 1.0° GFS

Radar at 1200 UTC 17 September 2004Radar at 0000 UTC 18 September 2004Radar at 1200 UTC 18 September 2004 QQnQn QsQs WSI radar 925 hPa Q Vector Diagnosis : 1200 UTC 18 September 2004 Q vectors (10 −10 K m −1 s −1 beginning at 2.5 × 10 −11 ), θ (K) contoured in green, and Q div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors 1.0° GFS Highest reflectivity near strongest QG forcing for ascent

Cross Section of F n Magnitude: 0000 UTC 18 September 2004 Deep frontogenesis tilting toward cold air w/height 925–500 hPa layer-avg F n vectors (10 −10 K m −1 s −1 ), θ (K) contoured in green, and F n div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors F n magnitude [K (100 km) −1 (3 h) −1 ] shaded, θ (K) contoured in gray, wind barbs (m s −1 ), and ω<0 (µb s −1 ) contoured in red 1.0° GFS

Cross Section of F n Magnitude: 1200 UTC 18 September 2004 Deep frontogenesis tilting toward cold air w/height 925–500 hPa layer-avg F n vectors (10 −10 K m −1 s −1 ), θ (K) contoured in green, and F n div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors F n magnitude [K (100 km) −1 (3 h) −1 ] shaded, θ (K) contoured in gray, wind barbs (m s −1 ), and ω<0 (µb s −1 ) contoured in red 1.0° GFS

Heaviest 6-h precip along and on cold side of surface boundary 6-h precipitation (in) ending at 0600 UTC 18 September UTC 18 September 2004 NPVU QPE Flow of tropical air into surface boundary F n vectors (10 −10 K m −1 s −1 beginning at 1.0 × 10 −10 ), θ (K) contoured in green, streamlines contoured in black, and F n div–con (10 −14 K m −2 s −1 ) shaded in cool–warm colors ~0.6° surface data 0600 UTC 18 September 2004

Heaviest 6-h precip along and on cold side of surface boundary 6-h precipitation (in) ending at 1200 UTC 18 September UTC 18 September 2004 NPVU QPE Flow of tropical air into surface boundary F n vectors (10 −10 K m −1 s −1 beginning at 1.0 × 10 −10 ), θ (K) contoured in green, streamlines contoured in black, and F n div–con (10 −14 K m −2 s −1 ) shaded in cool–warm colors ~0.6° surface data 1200 UTC 18 September 2004

Case Study 2: Ernesto August–September 2006

ROT Precip Distribution NPVU QPE Dates denote 0000 UTC positions 09/01 09/02 09/03 Total precip (in.) vs. TC track: 1200 UTC 31 Aug–1200 UTC 1 Sep 2006 Case Study 2: Ernesto

300 hPa h (dam), wind speed (m s −1 ), and div (10 −5 s −1 ) 300 hPa Analyses: 1200 UTC 31 August hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) Jet much farther downstream than with Ivan 1.0° GFS

300 hPa h (dam), wind speed (m s −1 ), and div (10 −5 s −1 ) 300 hPa Analyses: 1200 UTC 1 September hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) 1.0° GFS

925 hPa Analyses: 1200 UTC 31 August hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) WSI radar, 925 hPa θ e (K) and wind barbs (kt) 1.0° GFS

925 hPa Analyses: 1200 UTC 1 September hPa frontogenesis [K (100 km) −1 (3 h) −1 ], θ (K), and wind barbs (kt) WSI radar, 925 hPa θ e (K) and wind barbs (kt) Highest reflectivity near nose of LLJ/θ e ridge axis Strong frontogenesis along warm frontal zone 1.0° GFS

Radar at 0000 UTC 1 September 2006 Q n div–con bands within coastal boundary as Ernesto nears landfall Strong forcing for descent–ascent associated with Q n and Q s div–con 925 hPa Q Vector Diagnosis : 0000 UTC 1 September 2006 QQnQn QsQs WSI radar Q vectors (10 −10 K m −1 s −1 beginning at 2.5 × 10 −11 ), θ (K) contoured in green, and Q div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors 1.0° GFS

Radar at 0600 UTC 1 September hPa Q Vector Diagnosis : 0600 UTC 1 September 2006 QQnQn QsQs WSI radar Highest reflectivity near strongest QG forcing for ascent Q vectors (10 −10 K m −1 s −1 beginning at 2.5 × 10 −11 ), θ (K) contoured in green, and Q div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors 1.0° GFS

Radar at 1200 UTC 1 September hPa Q Vector Diagnosis : 1200 UTC 1 September 2006 QQnQn QsQs WSI radar Highest reflectivity near strongest QG forcing for ascent Q vectors (10 −10 K m −1 s −1 beginning at 2.5 × 10 −11 ), θ (K) contoured in green, and Q div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors 1.0° GFS

Cross Section of F n Magnitude: 0000 UTC 1 September –500 hPa layer-avg F n vectors (10 −10 K m −1 s −1 ), θ (K) contoured in green, and F n div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors F n magnitude [K (100 km) −1 (3 h) −1 ] shaded, θ (K) contoured in gray, wind barbs (m s −1 ), and ω<0 (µb s −1 ) contoured in red Strongest frontogenesis focused near surface 1.0° GFS

Cross Section of F n Magnitude: 1200 UTC 1 September –500 hPa layer-avg F n vectors (10 −10 K m −1 s −1 ), θ (K) contoured in green, and F n div–con (10 −15 K m −2 s −1 ) shaded in cool–warm colors F n magnitude [K (100 km) −1 (3 h) −1 ] shaded, θ (K) contoured in gray, wind barbs (m s −1 ), and ω<0 (µb s −1 ) contoured in red Strongest frontogenesis focused near surface 1.0° GFS

Heaviest 6-h precip along and on cold side of surface boundary 6-h precipitation (in) ending at 1200 UTC 1 September UTC 1 September 2006 NPVU QPE F n vectors (10 −10 K m −1 s −1 beginning at 1.0 × 10 −10 ), θ (K) contoured in green, streamlines contoured in black, and F n div–con (10 −14 K m −2 s −1 ) shaded in cool–warm colors 1200 UTC 1 September 2006 ~0.6° surface data

Heaviest 6-h precip along and on cold side of surface boundary 6-h precipitation (in) ending at 1800 UTC 1 September UTC 1 September 2006 NPVU QPE F n vectors (10 −10 K m −1 s −1 beginning at 1.0 × 10 −10 ), θ (K) contoured in green, streamlines contoured in black, and F n div–con (10 −14 K m −2 s −1 ) shaded in cool–warm colors 1800 UTC 1 September 2006 ~0.6° surface data

Upper-level jet streak jet streak low-level Q n LLJ Q s div Q n div Q s con Q n con low-level θ Summary of Case Studies: Conceptual Model 1 Heavy rainfall sfc boundary

LOT Precipitation Distribution ROT ColdWarmZθ ColdWarmZθ Strongest frontogenesis focused near surface Deep frontogenesis tilting toward cold air w/height Summary of Case Studies: Conceptual Model 2

Acknowledgements Special thanks to: –Lance Bosart and Dan Keyser –David Vallee –John Cannon and Dan St. Jean - WFO GYX –Kevin Tyle and Alan Srock –The rest of the grad students for keeping me sane for the past two years! –My family –Adrienne

Summary of Case Studies Heaviest precipitation occurs in the presence of strong surface F vector convergence and upper-air Q vector convergence. –Q n forcing for descent–ascent bands located within low- level frontal zone beneath equatorward jet-entrance region –Q s forcing for descent–ascent couplet located within upstream–downstream thermal trough–ridge over eastern U.S. Heaviest 6-h precipitation occurs along and on cold side of mesoscale surface boundary.

Both environmental circulation of TC and downstream LLJ induce the poleward transport of high θ e air into a pre-existing low-level baroclinic zone. LOT and ROT precipitation distribution is related to vertical structure of frontogenesis. –Ivan: LOT precipitation distribution with deep frontogenesis tilting toward cold air with height –Ernesto: ROT precipitation distribution with strongest frontogenesis focused near the surface Summary of Case Studies

Introduction –Objectives –Motivation Data and Methodology Results –Climatology –Case Studies Conceptual Models Outline