Extreme Rain and Wind Storms in the

Mid-Latitudes I

G. Tetzlaff

Universität Leipzig

Singapore, 21.04.2009

Contents Part I1. Introduction2. Mid-latitue weather systems2.1 The structure of mid-lat. Weather systems2.2 Wind : Kyrill 20072.3 Flood : Elbe 2002

Part II3. Damage4. Weather models5. Past events of (extreme) weather and estimates for users6. Wind maximisation (probable maximum gust)7. Rain maximisation PMP (probable maximum precipitation)8. Conclusions

1. Introduction

2. Mid-latitude weather systems

2.1 The structure of mid-latitude

weather Systems

Differential solar heating

Difference :

301 – 257 ~ 44 K

! at 1000hPa !

average gradient equ.-pole :

~0.3K/ 100 km

frontal zone gradient :

~5K/ 100 km

~150 m/s in ~300hPa

Two air masses :

Tropical

Polar

Frontal zone

in between,

with

~5-10 times the

average

zonal temperature

gradient

The Ocean (thin) and Atmospheric (dotted) contributions to the total northwards

heat flux based on the NCEP reanalysis (in PW) by (i) estimating the net

surface heat flux over the ocean (ii) the associated oceanic contribution,

correcting for heat storage associated with global warming and constraining the

ocean heat transport to be -0.1 PW at 68 S (iii) deducing the atmospheric

contribution as a residual. The total meridional heat flux is also plotted (thick)

(Trenberth and Caron 2001).

Average horizontal transports of a specific property ψ

are fully described

T = ρ u ψ

with

A = A + A‘ for u and ψ,

neglecting ρ‘ and avering with respect

to time and latitudonal circle

four types of transports remain

(two stationary ones with respect to time

and two instationay ones) :

1. stationary flow, constant over the whole latitudonal circle

(stationary cell with horizontal axis)

2. stationary flow, variable over lat. circle (stationary eddies with

vertical axis of rotation)

3. instationary flow, constant over whole lat. Circle (fluctuating

cell with horizontal axis)

4. Instationary flow, variable over lat. Circle (transient eddies,

with vertical rotational axis).

Under the assumptions these four types of transports are

complete.

In the mid-latitudes transports are managed by #4. mainly.

MMC : mean meridional Circulation (horizontal axis)

Eddies : vortices with vertical axis (Speth et al. 1974)

mid-latitudes

January

Mid-latitudes : typically eddies dominate the scene!

Transportable specific (per kg) energy

in the atmosphere :

1. potential energie P = g z

2. enthalpie (sensible) energy H = cp T

3. latent energy L‘ = L q

4. kinetic energy K = u2/2

(enthalpie from inner energy U= cvT)

Petterssen, Weather Analysis and Forecasting 1958

Surface isobars (solid)

and 1000-500 mb

thickness (dashed)

• thickness contours

concentrated between sfc

front and upper level frontal

zone

Basic driver is the

eq.-pole temperature

difference

(necessarily resulting in

differences of

latent, internal,

and potential energies

transient eddies =

mid-lat. cyclones

mid-lat. :

transient eddies

are the major

constituent of

all weather,

because of the

transport requirements.

typical propagation

velocity of

mid-lat. cyclone

10 m/s.

With the intense

wind and rain „zones“

(fronts) extending over

a few 100 km this results

in a typical

extreme weather duration

of 3 hours to

½ day.

whole cyclone :

a few 1000 km

intense rain :

a few 100 km

Types of Fronts

• Cold Fronts:

– slope ~1:50 to 1:100

• Warm Fronts:

– slope ~1:100 to 1:200

• Occluded fronts

Concentration of T-gradient

Wave length

Low pressure system close to Iceland, NOAA März 2004,

„organised“ clouds

mid-latitudes : eddies with vertical rotation axis

T (frontal zone

at 500 hPa) ~

15K/300km

mid-latitudes :

wind speed max.!

climate : what can happen where, how strong, how often, for how long?

Summary : dominance of the

mid-lat. cyclones with respect to

atm. motions (boundary layer

extra)

2.2 Wind : Kyrill 2007

25. Dezember 1999 Orkan Lothar : Fotoaufnahme: Sturmwurffläche am 26.12.1999 bei Ettlingen, Lkr. Karlsruhe

Maximalböen Karlsruhe 151 km/h, d.h. ~1/1000a

strong pressure gradient

Propagation velocity Kyrill

18.01. 01 UTC to 19.01. 01 UTC :

~2400 km/24h = 100 km/h with

~constant core pressure

>300 km/h

wind in 300 hPa ~ 100 m/s

are more violent processes possible?

winter storm with gale force winds

and damage; storm „Kyrill“

Kyrill

Max-gusts

18.01.2007,

Last 24h, in km/h;

wetteronline

Düsseldorf :

~1/100a-

event!

damage :

~5 –10 10^9 EUR

with 18m/s as the average wind speed there is a band width

of gusts plausible :

standard gust factor (building codes) 1.6 29m/s

thunderstorm (theoret. gust, with u=3.9 T (here 10K) 39m/s

observed gust factor (widely spread) 1.8 32m/s

city gust factor (high roughness) 2.3 41 m/s

airport Düsseldorf observed 2.4 40 m/s

combined influences

of convection +

surface roughness

Sturmtief Kyrill vom 18.01.2007

0

2

4

6

8

10

12

14

16

18:40 19:00 19:20 19:40 20:00 20:20 20:40 21:00 21:20 21:40 22:00 22:20

Uhr

Tem

pera

tur

/

Nie

ders

ch

lag

960

962

964

966

968

970

972

Lu

ftd

ruck

Temperatur Niederschlag Druck

temperature drop : ~9 C in ~10 minutes;

within that interval : ~10 mm of rain

measurements : Institut für Meteorologie, Universität Leipzig

There is no climate information available for

parameters like : K/s, phase shift between parameters, etc.

0

5

10

15

20

25

30

18:40 19:00 19:20 19:40 20:00 20:20 20:40 21:00 21:20 21:40 22:00 22:20

0

5

10

15

20

25

30

Temperatur Windspitzen

Kyrill 18.01.2007Measurements Institut für Meteorologie, Universität Leipzig

---Windgeschwindigkeit

gust factor : 25m/s/11m/s = 2.27;

max. gust : 25m/s = 90 km/h

location Date

376 km/h Pacific, Typhoon Angela, gust 01./02. Nov. 1995

372 km/h* Mt. Washington, 10-Minute-average 12. April 1934

357 km/h Caribik, Hurricane Wilma, gust 18./19. Oct. 2005

335 km/h* Zugspitze, winter storm, gust 12. June 1985

333 km/h Thule (Greenland), winter storm, gust 08. March 1972

324 km/h Dumont d‘Urville (Antarctica), June (1980-1992)

catabatic wind, gust

310 km/h Pacific offshore Japan, Typhoon Tip, gust 12. Oct. 1979

294 km/h* Atlantic (buoy 59N/11W), cyclone Gero, gust 12. Jan. 2005

263 km/h* Brocken (Germany), winter storm, gust November 1984

262 km/h Havanna (Cuba), Hurricane, gust 18. Oct. 1944

216 km/h* Granville (Frankreich), winter storm, gust 16. October 1987

184 km/h* List/Germany, winter storm Anatol, gust 03. Dec. 1999

151 km/h** Karlsruhe, winter storm Lothar, gust 26. Dec. 1999

144 km/h** Düsseldorf, winter storm Kyrill, gust 18. Jan. 2007

*mid-lat.

**mid-lat. Inland, lowlands

2.3 Flood : Elbe 2002

Flood in Germany

August 2002

Overall flood damage in Europe amounted to more than $ 20 10^9

Munich Re,

CRED, 2003/07

Schumann,

2003

track of the Vb-cyclone bringing moisture northwards,

and forcing winds upslope at central European

mountain ranges

air flow steered

against orography

(Erzgebirge)

140 mm areal average rains

340 mm peak of the mountain

The 36h forecast was more accurate than the 12h one!

River Elbe at Dresden

(Source: Engel et al., 2002; "Der Elbestrom", 1898)

1342: 824

1432: 830

1501: 857

1655: 8381275: 840

2002: 940

1799: 824

1784: 857

1862: 824

1890: 837

1845: 944 (877)

700

750

800

850

900

950

1000

1200 1300 1400 1500 1600 1700 1800 1900 2000 2100

Year

Leve

l in

cm

700

750

800

850

900

950

1000

Summer

1845, corr.

Winter

runoff coefficient : 0.44

areal rain 139 mm (2 days)

Extreme Rain and Wind Storms in the

Mid-Latitudes II

G. Tetzlaff

Universität Leipzig

Singapore, 22.04.2009

3. Damage

CRED 2007

Changes (linearised) :

dark red ~15%/a

purple ~ 4%/a

red ~-0.1%/a

CRED 2008

role of climate change?

Source: VROM (the Netherlands

Ministry of Housing, Spatial Planning

and the Environment)

Tue curve does represent

ONE out of many social

consenus!

4. Weather models

The set of equations to describe the relevant atmospheric

processes causing rain and wind storm

consists of 3 conservation equations :

mass, energy, momentum

and the equation of state.

The number of forces acting on an air parcel is 5 :

pressure gradient, Coriolis, inertia, friction, gravity

There are 4 forms of energy to be

considered here : potential, internal, kinetic, latent.

DKKV Workshop Severe Storms 26-28 March 2007

Limitation: Model grid box large

Grid box 25 km x 25 km

ECMWF 2006

The AVHRR data resolve 2 km² with four measurements per day the METEOSAT

resolves about 50 km² with 48 measurements per day.

Scaling : theory needs clarification on the temporal and spatial resolution!

Divergence div u and vertical motion w :

w ~ ∫ div u dz;

with dz ~ 10 km and div u ~ 10-5

w ~ 0.1 m/s in the „intense“ parts.

Terrain-induced vertical motion

Steeper terrain contributes to greater vertical motion. Terrain-induced vertical motion influences precipitation amount,can lead to gravity waves downstream

hVterrainwt )( hVcos

after Stull 1988

phase shift of w(z) small

5. Past events

of (extreme) weather

and estimates for

users

~2000 mm/day, La

Réunion, cone-shaped

mountain with ~3000m

height

NOAA, 2001

------ 2000mm/d

This is NOT PMP!

City of Hamburg, 2004

Practical application of extreme weather frequencies

Return period in yearsduration

Design values for precipitation in the building sector

Probabilities of occurrence of gusts, DWD 2005

6. Wind maximisation

(probable maximum gust)

Structure of a mid-lat. cyclone

absolute temperature in 850 hPa

vertical velocity in Pa/s in 500 hPa

Example : Anatol 3.12.1999

Mechanisms of cylonic formation

PVA

Vorticity in 300 hPa in 10-4 s-1

Geopotential (82000 bis 93000 m2/s2 –

Isolines at 1000 m2/s2) in 300 hPa

WAA)

in 850 hPa in K/h

Synthetic increase of the

horizontal temperature gradientIncrease T=5 K

„Added“ temperature in K:

Horizontal field in 700 hPa

Cross section at ca. 50 n.Br.

original temperature (Isolines at 2,5 K)

Reference Temperature difference 2,5 K

Temperature difference 5 K Temperature difference 7,5

K

Undine 6.-7.1.1991

Max winds (gusts) in m/s

0,0

10,0

20,0

30,0

40,0

50,0

60,0

70,0

Böen, w

elc

he in 1

/1000 d

es M

odellg

ebie

tes

übers

chritten w

urd

en, in

m/s

Und

ine

(199

1)

Dar

ia (1

990)

Loth

ar (1

999)

1991

1229

12

Wie

bke

(199

0)

Mar

tin (1

999)

1989

0211

06

Silk

e (1

998)

1962

1213

18

Ham

burg

er F

lut (

1962

)

1976

1129

18

1983

0116

12

Cap

ella (1

976)

1984

0113

00

1988

0208

00

Ana

tol (

1999

)

Referenzlauf

stärkster Lauf

16 winter storms: observed vs. „maximised“

gusts in m/s (exceeded in

1/1000 of the model area)

Frequency of exceedance of gust velocity in

events per year

1E-3

0,01

0,1

1

0 10 20 30 40 50 60 70

vB in m/s

Hä

ufig

ke

it d

er

Üb

ers

ch

reitu

ng

in A

nte

ilen a

m M

ode

llge

bie

t

R

2,5 K

5 K

7,5 K

Radtke 2006

7. The rain maximation

PMP (probable maximum

precipitation)

Daily sums of precipitation on the Erzgebirge in August

2002%PMP* %PMP*

11.08. 12.08. 13.08. 3 days 12.08. 3 days

Estimated

Near Res. --- 410 --- --- 103 --- Altenberg

Reservoir 31,5 239,5 24,3 295,3 60 54 Lehnmühle

Reservoir 9,8 280,6 23,2 313,6 70 57 Klingenb.

Reservoir 20,8 218,0 11,1 249,9 55 45 Malter

Reservoir 26,0 217,2 12,8 256,0 54 47 Gottleuba

•MGN - Maximierte Gebietsniederschlagshöhen or PMP (data DVWK, 1997) „traditional

estimate of PMP following the structure of WMO guide-lines.

River Main at Würzburg

(Source: SpektrumWasser 1, München 1998, p. 55, 77)

Italic: approximate values1342: 3350

1845: 2170

1909: 1800

1682: 2250

1595: 2000

1451: 2250

1442: 24501546: 2350

1633: 2000

1784: 2650

1882: 1670

1000

1500

2000

2500

3000

3500

4000

1200 1300 1400 1500 1600 1700 1800 1900 2000 2100

Year

Runoff

in m

³/s

1000

1500

2000

2500

3000

3500

4000

Summer

Winter

flood of July 1342

Increase of precipitation depth

with decrease of preciptation area,

Elbe flood August 2002

Station Zinnwald

Model, area

Mulde catchment area

Elbe catchment area,

DresdenElbe catchment

area, total

Model, point

0

100

200

300

400

500

0,010,11101001000100001000001000000

Precipitation area in km²

Pre

cip

itation d

epth

in m

m/d

ay

With 1342 river Main values : ~800 mm/day

We estimated the areal rains for the 1342-flood event.The basic

orographic setting is similar to the one of the Erzgebirge

only for a different range of mountains. The 1342 flood was heaviest

where the contributaries originating in the mountain range united.

Applying the „1342-rains“ to the Erzgebirge : almost doubling the rains would occur.

Even allowing for errors in the estimates it calls for closer inspection of PMP-estimates.

Quantification of topographic effects on predicted precipitation for typical Elbe-catchments

in the Erzgebirge

Validation of model results, confirmation of of orographic effects.

Radar-derived precipitation [mm] accumulated for 6h on May, 15th 2004 with a strong

upper-level flow from the north-northwest.

U

EinzugsgebietTalsperre Malter

Dresden

Orographic profile

distance in m

Catchment Malter

CAPE very small!

Height dependent temperature and

dew point temperature

Quantification of topographic effects on predicted precipitation for typical Elbe-catchments

in the Erzgebirge

Influence of orography shape on structure of induced lifting for

standard rain conditions. The effects of stability (Froude Number)

are distinct.

Cross-section of vertical velocity w [m/s]

as a result of LM simulations

U

U

bell-shaped

ridge

vs.

Erzgebirge

orography

Vertical profiles for w :

1. Standard case

2. Maximum found from sensitivity analysis using LM

-1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0

0

2000

4000

6000

8000

10000

12000

exponential decay

adapted from LM

z [m

]

w*

dRR/km aus MAXRR: 5.5K/km, 15m/s, rH~98%

0 0,5 1 1,5 2 2,5 3

1

2

3

4

5

6

7

8

9

10

11

12

z [km

]

dRR/km [mm/(h*km)]

no scaling of w

w * exponential, Hw=5km

w* adapted from LM

resulting liquid water

release per km of height

Model runs are used to estimate the „maximum“ profile of w(z)

Orograpgically induced precipitation

height

The relation between

precipitation rate and vertical velocity w is linear.

The precipitation rate almost doubles per 10 K

change of the dew point temperature (at saturation level).

In real cases the maximum precipitation

is not reached (in observations the

highest values amounted to about 70%).

Precipitation rates show high sensitivity.

Rain in orography are influenced by :

vertical wind velocity (wind speed and slope)

vertical specific humidity profile

vertical profile of Froude-Number/Brunt-Väisälä-Frequency

vertical wind shear

wind drift of precipitation

8. Conclusions

Simple conceptual models can give

orientation estimating PMG and PMP.

Most critical seems to be the lack of relevant information

(how well is time and site dependant air flow

known, …?).

Thank you!