A Reanalysis of Hurricane Hazel (1954) Scott Weese, Ron McTaggart-Cowan and John R. Gyakum Department of Atmospheric and Oceanic Sciences McGill University.

A Reanalysis of Hurricane Hazel (1954)

Scott Weese, Ron McTaggart-Cowan and John R. Gyakum

Department of Atmospheric and Oceanic Sciences

McGill University

Montreal, Quebec, Canada

June 4th, 2003

Talk Outline

1 . Background

2 . Synoptic Overview

3 . MC2 Simulation Results

4 . Conclusions

Purpose: to study the thermodynamic and dynamic properties of one of the most deadly and costly weather disasters to strike southern Ontario

• Hazel struck Ontario on 15-16 October 1954

• rains led to flooding in Humber, Credit and Don River valleys situated near Toronto, and in the Holland Marsh north of Toronto

• over 80 fatalities, and $175 million damage was done by Hurricane Hazel in Ontario

Track of Hurricane Hazel

Reference: Mason, A.H., M.K. Thomas and D.W. Boyd. 1955. The October 15-16, 1954 Storm, “Hurricane Hazel” in Ontario.

Synoptic Overview

• Considering the time between October 15th at 00z to October 16th at 12z as this is the period of the most intense rainfall in southern Ontario

• Using the NCEP reanalysis data (Kalnay et al. 1996) to generate fields of interest

Figs. (a)-(d): SLP (hPa) and Thickness (dam)00 UTC 15 to 12 UTC 16 October

Dynamic Tropopause Maps

• The dynamic tropopause is defined as the 1.5 PVU surface

• 1 PVU = 10-6m2s-1K kg-1

• Ertel’s Potential Vorticity (PV) (Ertel 1942):

1

EPV

Figs. (a)-(d): DT Winds and and 850 hPa

00 UTC 15 to 12 UTC 16 October

Coupling Index, Equivalent Potential Temperature and Precipitable Water

• Coupling Index (CI) is a measure of bulk atmospheric stability (Bosart and Lackmann 1995)

• Values of CI < 10 suggest convective instability

• Precipitable water is the column integrated water vapour

)(CI sfcmaxeDT

2

1

p

pdpw

g

1PWTR

Figs. (a)-(d): Coupling Index and Precip. Water (mm)

00 UTC 15 October to 12 UTC 16 October

MC2 Simulation

• Using MC2 (Mesoscale Compressible Community Model) to simulate Hazel

• NCEP Reanalysis data utilized for the initial and boundary conditions

• following the work of Palmen (1958) and Anthes (1990)

• test sensitivity of Hazel to improved vortex structure using model by Kurihara (1993), and increased spatial resolution (36 km)

Tracks of Hurricane Hazel

Average Depth of Precipitation

0

5

10

15

20

0-3

3.-6

6.-9

9.-12

12.-1

5

15.-1

8

18.-2

1

21.-2

4.

24.-2

7

Time Period (hours)

Amou

nt (m

m)

Sea Level Pressure for Hurricane Hazel

960

970

980

990

1000

1010

15/00z

15/06z

15/12z

15/18z

16/00z

October 1954 Date

SLP

(mb)

AnthesNCEPBogusKnox

SLP Comparison 15 UTC 15

October

top: Palmen’s analysislower left: MC2lower right: Anthes

SLP Comparison3 UTC 16October

top: Palmen’s analysislower left: MC2 lower right: Anthes

Knox’s Jump

SLP for 21 UTC 15 October to 03 UTC 16 October

Precipitation Comparison, 15 October

Conclusions

• Successfully reproduced transformation of Hazel with accompanying frontogenesis and heavy precipitation over U.S.

• Improved vortex structure and increased spatial resolution necessary for successful mesoscale modeling of Hazel