Flash Flood Composite Analysis in Vermont and Northern New York

1. Flash Flood Composite Analysis in Vermont and Northern New York John M. Goff Gregory A. Hanson NWS Burlington, VT

2. Impetus for study • Since 1990, 20 of 25 federally declared disasters in Vermont were attributed to flooding. • Flash flood detection remains a challenging process in the Burlington, VT CWA.

3. Overview • North American Regional Reanalysis (NARR) composites • Atmospheric sounding composites • Distribution analysis of key sounding parameters • Spatial distribution of events

4. Study Area

5. NARR Analysis Methodology • Flash flood cases between 1981 and 2010 were analyzed (30-year period). - from archived AWIPS data and a local study (Breitbach 2009). • Data availability and temporal restraints narrowed dataset to 51 total cases for NARR analysis.

6. NARR Analysis - Isotachs * Weak flow less than 6 m/s below 850 mb. * Modest flow at mid and upper levels. * Coupled jet structure at mid and upper levels.

7. NARR Analysis – Height/MSLP * Modest front or trough evident at surface. * Broad weak troughing aloft from 850 mb through 250 mb.

8. NARR Analysis - Moisture * Distinct moisture plume extending north/northeast along surface trough into study area. * Enhanced boundary layer moisture convergence from eastern NY into central/northern New England.

9. NARR Analysis - Stability * Only modest instability present with LI values generally from -2 to 0°K and CAPE from 500-800 J/kg. * Little to no convective inhibition present. * Supports idea of modest though not excessive updraft strength.

10. Sounding Composite Methodology • Used RAOB software • 17 of original 51 cases • Limited by local archive of RUC Bufkit data • Events classified by discrete convective character. - Type A: isolated storm - Type B: classic training storms - Type C: MCS/MCC

11. Sounding Composite Methodology • Sounding composites for entire dataset and for each case type • Distribution analysis of key sounding parameters performed

12. Mean Composite Sounding * Light winds generally 20 kt or less in the boundary layer. * Modest winds aloft in the 25 to 45 kt range. * Modest veering/WAA boundary layer profile. * Light to modest values of shear.

13. Mean Composite Sounding * Deep WCD of 3.6 km indicative of warm rain processes at work. * Modest CAPE/N-CAPE values support only modest updraft strength. * Low LCL values suggest inhibition of moist downdraft production and limited cell movement.

14. Classification of Sounding Events * a) Isolated storm, b) classic training cells, c) MCC/MCS

15. Distribution of Events

16. Type B Profile • Classic training cells… Plainfield, VT 05.27.2011

17. Type B Profile * 71% of cases (12); sounding parameters thus similar to the mean sounding. * Slightly less wind but more marked veering in the boundary layer. * Deepest WCD value of the three plots indicating at least the potential for Type B events to produce the highest rainfall rates.

18. Type A Profile • The stationary single cell… Bristol, VT - 08.28.2004

19. Type A Profile • * 17% of cases (3); limited dataset limited overall conclusions. • * Highest CAPE/N-CAPE and LCL heights imply a deeper/warmer boundary layer. • * Lightest and most westerly low to mid-level wind profile of the three types.

20. Type A Profile * Mean westerly flow at mid-levels orthogonal to western slopes of Green Mountains, suggesting a mechanism by which Type A storms could become anchored to terrain.

21. Type A NARR Composite Heights * While limited, data suggests Type A cases arise when a weak surface trough is bridged aloft by weak to modest ridging.

22. Type C Profile • Mesoscale Convective Systems… Hancock, VT 08.06.2008

23. Type C Profile * 12% of cases (2). * Minimal instability and coolest thermal profile among the three types. * Light winds throughout the sounding, though most strongly veered implying more marked WAA signature in lower levels.

24. Type C Profile Type C * Larger meso-beta scale of Type C events suggest more widespread threat of flash flooding than in Type A or B events. Type B Type A

25. Distribution of Key Sounding Parameters • PWAT • WBZ height • WCD • LCL height

26. PWAT Distribution * PWAT values not excessive (1.5 -2.0”). * Mean value of 1.77” not all that unusual, only 117.3% of normal mid-summer value (1.51”).

27. WBZ Height Distribution * Relatively high values ranging from 10,077’ to 14,533’ AGL. * Non-normal distribution of data, though not surprising.

28. WCD Distribution * 15 of 17 cases above 3.0 km and a mean value of 3.5 km. * Supports prior research (Davis, 2004) suggesting deep WCD layers are a driving factor in many flash flood events. * Therefore it is hypothesized that WCD is a more important signal during potential flash flood days than PWAT or WBZ height.

29. LCL Height Distribution * All but 2 cases below 1,500’ AGL and 12 cases below 1,000’ AGL. * Low values in combination with high WCDs act to inhibit dry air entrainment and subsequent moist downdraft formation, thus limiting cell propagation. * Low LCL heights seemed to have the strongest signal during potential flash flood days of the 4 parameters analyzed.

30. Spatial Distribution of Events • Used entire 211 case dataset from Breitbach 2009 • Effort was done to determine spatial variability of events and to determine whether orography or other factors influenced location.

31. New York Events * Noticeable clustering of events along eastern slopes of Adirondacks highlighting orographical influence in this region. * Population bias evident in the Lake Placid area and to a lesser extent in the Saint Lawrence Valley near Potsdam and Canton. Also little to no reports in Adirondacks outside Lake Placid area. Lake Placid/Saranac Lake area

32. Vermont Events * Flash flood events more homogeneous in Vermont than in New York. * Some evidence of population bias in northeastern Vermont – low density.

33. Conclusions - NARR Analysis • NARR analysis indicated flash flood events characterized by: - relatively weak surface to mid-level troughing - modestly coupled mid to upper level jet - south to north oriented deep moisture axis - only modest instability

34. Conclusions - Sounding Composites • Composite mean sounding indicated events were characterized by: - light to modest flow throughout the column - distinct veering/WAA profile in the boundary layer - modest values of CAPE/N-CAPE • Subtle differences between composite mean and individual convective types - Type B - majority of cases -> classic training - Type A - greater instability and lighter winds - Type C - cooler though with more marked veering

35. Conclusions - Sounding Parameters • Sounding Composite Parameters indicated events were characterized by: - modestly high PWAT values near 1.75” - a deep WCD greater than 3.0 km and WBZ heights generally from 10-14 kft AGL - very low LCL heights below 1,500’ AGL

36. Limitations and Further Study • Study did not examine antecedent conditions such as prior 24-hour rainfall or soil moisture. • Sample size of sounding database. • Build a more comprehensive database of archived events.

37. References • Breitbach, M., 2009: Flash Flooding Climatology (1975-2009) for the WFO Burlington, VT County Warning Area. Poster, NWS Eastern Region Flash Flood Conference 2009. • Chappell, C. F., 1986: Quasi-Stationary Convective Events. Mesoscale Meteorology and Forecasting, P.S. Ray, Ed., Amer. Meteor. Soc., 289-310. • Chappell, C. F., 1993: Dissecting the Flash Flood Forecasting Problem. Post-Print Volume, Third Heavy Precipitation Workshop, NOAA Tech Memo. NWS ER-87, 293-297. • Cope, A. M., and L. R. Robinson, 2007: Composite Means and Anomalies of Meteorological Parameters for Summertime Flash Flooding in the National Weather Service Eastern Region. 22nd Conf. on Wea. Anal. And Fcstg., June 2007. • Davis, R. S., 2001: Flash Flood Forecast and Detection Methods. Severe Convective Storms, Meteor. Monogr., No. 50, Amer. Meteor. Soc., 481-525. • Davis, R. S., 2004: The Impact of Tropical Rainfall Rates on Flash Flood Detection. Preprints, 22nd Conference on Severe Local Storms, AMS, compact disk. • Federal Emergency Management Agency, cited 2011: Vermont State Disaster History. [Available online at http://www.fema.gov/news/disasters_state.fema?id=50.] • Hanson, G., 2004: 28 August 2004 Flash Flood in Addison County Vermont. 6th Northeast Regional Operational Workshop, November 2004. • Harnack, R., K. Apffel, M. Georgescu, and S. Baines, 2001: The Determination of Observed Atmospheric Differences between Heavy and Light Precipitation Events in New Jersey, USA. Int. J. Climatol., 21, 1529-1560. • Jessup, S. M., and A. T. DeGaetano, 2008: A Statistical Comparison of the Properties of Flash Flooding and Nonflooding Precipitation Events in Portions of New York and Pennsylvania. Wea. Forecasting, 23, 114-130. • LaPenta, K. D., and Coauthors, 1995: The Challenge of Forecasting Significant Rain and Flooding Throughout the Eastern Region of the National Weather Service. Part I: Characteristics and Events. Wea. Forecasting, 10, 78-90. • Maddox, R. A., C. F. Chappell, and L. R. Hoxit, 1979: Synoptic and Mesoscale Aspects of Flash Flood Events. Bull. Amer. Meteor. Soc., 60, 115-123. • Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343-360. • National Weather Service, cited 2011: Precipitable Water Plots. [Available online at http://www.crh.noaa.gov/unr/?n=pw] • Pontrelli, M. D., G. Bryan, and J. M. Fritsch, 1999: The Madison County, Virginia, Flash Flood of 27 June 1995. Wea. Forecasting, 14, 384-404.

38. Thank You!