Observations and modelling of severe windstorms

1
School of Earth and Environment ICAS
Observations and modelling of severe
windstorms during T-REX importance of the
upstream profile
Ralph Burton1, Simon Vosper2, Peter Sheridan2,
Stephen Mobbs1

• 1 Institute for Climate and Atmospheric
  Science, School of Earth and Environment,
• University of Leeds, U.K.
• Met Office, Exeter, U.K.

Leeds T-REX team Barbara Brooks, Ian Brooks,
Rosey Grant, James Groves, Martin Hill, Matt
Hobby, Felicity Perry, Victoria Smith, Will
Thurston
2
School of Earth and Environment ICAS
Introduction T-REX Comparing models and
observations Relating the winds in the valley to
the upstream flow Summary
3
T-REX March April 2006
The biggest field campaign ever mounted to
study rotors/gravity waves
ARL White Sands Missile Range Scripps Institute
of Oceanography Colorado Research
Associates Cooperative Research in Environmental
Science Desert Research Institute DLR Lawrence
Livermore National Laboratory UK Met
Office NASA NCAR Naval Research Laboratory NOAA
Arizona State University Colorado State
University Harvard University University of
Houston University of Innsbruck University of
Leeds University of New Hampshire North Carolina
State University Stanford University University
of Utah Yale University
4
From Hazardous Mountain Winds and their Visual
Indicators, 1988
Accident rate 40 higher in the 11
mountain states
Accident rate less than 3 per 100,000
Accident rate greater than 3 per 100,000
5
DEM of the U.S. showing regions of elevated
terrain
6
(No Transcript)
7
(No Transcript)
8
Will Thirsty Thurston
9
Inyo Register, March 2006
10
Photo Carlyle Calvin, UCAR
11
Schematic of wave-rotor system
12
T-REX IOP6 25th 26th March 2006
Kuettner 1959
A full temporal evolution of a trapped-lee-wave
rotor event was captured in this IOP. There was a
strong and well defined wave/rotor event with
wave clouds, roll clouds, cap cloud over the
Sierra, and a dust storm in Owens
Valley. (Mission Summary)
Photo A. Doernbrack
13
26th March 2006 639PM LT
Height (m)
w (m/s)
14
What we want to do, in a nutshell
Take an upstream profile, and attach a number N1
to it
15
Height (km)
10
5
T (K)
15
What we want to do, in a nutshell
Take an upstream profile at time T, and attach a
number N1 to it
15
Height (km)
10
5
T (K)
Look at the downstream winds at time T attach a
number N2 to them
Look at lots of cases (different T). Are N1 and
N2 related? - Predictability
16
For the downstream winds. We seek a statistical
parameter to describe the effect of severe winds.
From Mobbs et al. 2005 (Observations of
downslope winds and rotors in the Falkland
Islands, QJRMS, 131, 2839-2860
s2 su2 sv2 U average wind speed
Produce a time series of s U
17
Locations of the AWS DRI Leeds
Produce a time series of s U
18
For the upstream profiles.
It is thought that the stratification and the
shear of the upstream profile are important
parameters in determining the nature of the
downstream response.
From Hertenstein and Kuettner, Rotor types
associated with steep lee topography influence
of the wind profile, Tellus A, 57, 117-135
19
UM model simulations for IOP6
UM configuration
20
Principal Components Analysis (PCA) of the
upstream profiles Take time series of upstream
profiles UM output every 10 seconds Perform
PCA extracts two features EOFs (Empirical
Orthogonal Functions) Describe the dominant
structures in the profiles (inversions etc) PC
Scores (Principal Component Scores) A numerical
value a measure of how similar the profile is to
the dominant structures at each point in time

PCA is an objective method
Produce a time series of the principal component
scores
21
UM Principal components analysis of the upstream
profile between 3 and 5 km
dU/dz
N
These are the dominant structures in the N and
dU/dz profiles for all upstream profiles (sample
size 6840) Maximum in both shear and stability
at 4250m
22
Time series of principal components wind effect
parameter s U UM
23
Time series of principal components wind effect
parameter s U AWS
24
Correlations UM wind effect parameter s U
with upstream profile structure
UM winds
25
Correlations AWS wind effect parameter s U
with upstream profile structure
Insert pic here
AWS winds
26
School of Earth and Environment ICAS
Summary
The dominant structure in the UM profiles between
3km and 5km is a peak in shear and a peak in
stability at z 4250m There is a very strong (
r 0.95 r 0.87) correlation between the UM
winds in the valley and the upstream N and dU/dz
profiles for all of IOP6 (19 hours) The
correlation is not as strong for the observed
winds. But then the model is reacting to the
model profiles, which may differ slightly to the
observed profiles.
Is this behaviour unique? Or does it apply to
other IOPs? Ongoing.
27
Correlations between first PC score for N and
dU/dz