Stratospheric variability and stratosphere-troposphere connections (MOD 3 and KEY 6)

1
Stratospheric variability and stratosphere-troposp
here connections(MOD 3 and KEY 6)

• Mark P. Baldwin
• Northwest Research Associates, Bellevue, WA USA
• Cargese UTLS Summer School, 6 October 2005

2
Overview

• EOFs
• Variability in the stratosphere-troposphere
• The Arctic Oscillation (AO) and Northern Annular
  Mode (NAM)
• Downward coupling to the troposphere
• The stratosphere and weather prediction
• Explanations for downward coupling
• The quasi-biennial oscillationand hurricanes

3
EOFs
Empirical Orthogonal Functions Maximum Variance
Patterns 1) Pattern 2) time series
4
(No Transcript)
5
(No Transcript)
6
Composite surface maps for high and low AO
index. (From Thompson and Wallace, Science 2001)
gt0.9C
7
Northern Annular Mode (NAM)
10 hPa
1000 hPa (Arctic Oscillation)
Annular mode patterns are the leading EOF of
low-frequency Z variability. Annular mode
patterns are similar from Earths surface to
50km.
8
Annular Mode Terminology

• Arctic Oscillation (AO) At or near Earths
  surface.
• North Atlantic Oscillation (NAO) Very similar to
  AO, but more oriented to the Atlantic sector.
• Northern Annular Mode (NAM) Same as the AO, but
  can describe higher levels in the atmosphere.
• The surface NAM is the Arctic Oscillation.

9
What are annular modes?

• Not fundamental dynamical quantities
• Not the solution to any equation
• Why do the NAM and SAM look so similar and in the
  troposphere stratosphere?
• Why does the atmosphere tend to vary in
  ring-like structures?
• Why not multiple rings?
• How are annular modes related to jets?

10
(No Transcript)
11
Composite surface maps for high and low AO
index. (From Thompson and Wallace, Science 2001)
gt0.9C
12
Southern Hemisphere surface climate response to
ozone depletion

• Observations and model
• Springtime ozone loss appears to drive changes in
  surface climate from late spring to summer.

13
Simulated and observed geopotential height and
temperature changes
Model results from Gillett Thompson, Science
2003
14
Tropospheric changes
Model results from Gillett Thompson, 2003
15
NAM index for 19981999. 23 separate levels are
used. The lowest level is the AO index. From
Baldwin and Dunkerton, Science 2001
16
Composites of strong and weak vortex events

• Select strong events based only on the daily
  10-hPa NAM index.
• Form positive and negative composites.
• Examine average behavior for the composites.

17
Time Delay
Long timescale
Baldwin and Dunkerton, 2001
18
Surface pressure anomalies after stratospheric
events look like the Arctic Oscillation.
Baldwin and Dunkerton, 2001
19
Northern Annular Mode (NAM)
1000 hPa (Arctic Oscillation)
20
Storm tracks during weak and strong regimes
21
Baldwin and Dunkerton, 2001
22
Thompson et el., JAS, 2005
23
Predictability Beyond 10 Daysa role for the
stratosphere?

• Boundary Conditions (SSTs, Snow and Ice, Soil
  Moisture, etc.)
• Persistent Phenomena (MJO, QBO, ENSO, etc.)
• Persistent stratospheric anomaliescan they
  affect the troposphere?

24
Autocorrelation of daily surface AO index
(W. Norton, GRL, 2003)
Control Run
Normal Model
Model with damped stratosphere
Without stratospheric variability, the timescale
of the surface AO is shorter.
25
e-folding timescale time for autocorrelation to
drop to 1/e (0.38). In the troposphere the
longest timescale occurs during winter. Does
this effect depend on the stratosphere, or is it
simply an annual cycle?
Baldwin et al., Science, 2003
26
(No Transcript)
27
Statistical AO Forecasts

• Exploit the enhanced timescale of the AO during
  winter.
• Predict the average AO during a future time
  period, such as 1040 days.
• Use one or more linear predictors
• 1) The present value of the AO
• 2) The present values of the annular modes at
  all other levels, including the stratosphere.

28
Baldwin et al., 2003
29
Cross-validated AO Forecasts

• Remove one winter at a time forecast the
  remaining winter.
• 150-hPa NAM only during DJF skill is reduced
  from 20 to 18.
• AO predicts AO cross-validated skill is 12
• Adding the the AO as a second predictor does not
  improve skill.

30
(No Transcript)
31
Composite surface maps for high and low AO
index. (From Thompson and Wallace, Science 2001)
gt0.9C
32
Possible Dynamical Mechanisms

• Direct effect of stratospheric wave driving and
  stratospheric temperature anomalies
• Indirect effects involving waves
• Effect on baroclinic waves/life cycles
• Effect on planetary-scale waves
• Wave reflection in the stratosphere

33
From Thompson et al., JAS submitted
34
From Thompson et al., JAS submitted Gstratospher
ic wave driving Qstratospheric radiative
cooling Ffriction
35
Stratospheric influence on baroclinic lifecycles
and its connection to the Arctic Oscillation

• M. Wittman, L. Polvani, R. Scott, A. Charlton
• GRL, 2004
• Baroclinic lifecycle model (wave-6 sector model)
• Numerical experiments show that baroclinic
  lifecycles are sensitive to changes in
  stratospheric winds.
• The surface response to stratospheric changes is
  an AO-like pattern (the model is too simple to
  have an AO).
• The stratosphere has an effect on the surface by
  influencing individual synoptic systems.

36
From Alexander and Holton, 1997
37
The Equatorial Quasi-Biennial Oscillation
Baldwin et al., Reviews of Geophysics, 2001
38
500-hPa (mid-troposphere)
39
QBO email from Barbara Naujokat during first talk
40
500-hPa (mid-troposphere)
41

42
Do Stratospheric Winds Affect Hurricanes?
43
Tracks of intense hurricanes. Top 15 years when
the QBO winds were easterly. Bottom 15 years
when the QBO winds were westerly. 1950-1990
data. From Gray et al. 1992
44
(No Transcript)
45
Hurricane, as seen in column ozone
46
20N
10N
47
(No Transcript)
48
(No Transcript)
49
E-W wind component averaged over all longitudes
50
E-W wind component averaged over 20W to 100W
51
Wind shear averaged over all longitudes
52
Wind shear averaged over 20W to 100W
53
A stronger, colder vortex?

• With increasing GHGs, the modeled stratospheric
  vortex becomes stronger and colder, and there is
  a positive AO trend.
• Shindell et al., 1999
• With increasing GHGs, the model NAO index
  decreases significantly from 1990 to 2015. The
  polar stratosphere becomes warmer.
• Schnadt and Dameris, 2003

54
250-day running mean AM time series.
55
Polar vortex response to 4x CO2
January Temperature Change
January U Change

• The polar vortex gets weaker and warmer due to
  enhanced wave driving from the troposphere.
• These plots show the response in a 4xCO2
  integration - easier to distinguish from noise
  (from Jamie Kettleborough).

56
A simulation of the separate climate effects of
middle atmospheric and tropospheric CO2 doubling

• M. Sigmond, P.C. Siegmund, E. Manzini, H. Kelder
• J. Climate, 2004
• MA ECHAM4, CO2 doubling
• Increased residual circulation small warming in
  the Arctic lower stratosphere weaker winds in
  the Arctic middle atmosphere increase in
  tropospheric westerlies.
• The increase of the tropospheric NH midlatitude
  westerlies can be attributed mainly to the
  middle-atmospheric CO2 doubling, indicating a
  crucial importance of the middle-atmospheric CO2
  doubling for tropospheric climate change.

57
IPCC Altitude of the Model Top
Stratopause
Tropopause
Number of levels
(IPCC slides from Eugene Cordero)
58
Questions

• How important is the stratosphere to tropospheric
  climate change and variability?
• What impact does resolution have?
• How high and how many vertical levels
• How many levels near the tropopause?
• First step How well do these climate model
  simulate stratospheric structure and variability?

59
January Temperature (Global Avg 1900-2000)
60
1960-2000 Zonal Wind _at_ 60N and 50hPa
Zonal Wind (m/s)
61
DJF Temperature (50S 80S) _at_ 10hPa
62
IPCC Summary

• The IPCC AR4 simulations vary in their ability to
  simulate stratospheric variability and change
  during the 20th century.
• The evolution of the polar vortex is a good
  diagnostic for model variability and skill.
• There is no model consensus regarding changes in
  the 21st century stratospheric climate.

(From Eugene Cordero)
63
The Future of the Stratosphere?

• Increasing GHGs cool the stratosphere.
• Stratospheric NAM index trend would depend on
  relative cooling of the polar cap.
• Ozone loss cools the lower stratosphere (late
  winter/spring). Highly dependant on temperature.
• Most models show an enhanced wave driving, a
  stronger Brewer-Dobson circulation, and a warmer
  NH vortex in winter.

64
Summary

• Persistence and predictability of the AO depend
  on the long timescale of large circulation
  anomalies in the lowermost stratosphere.
• This relationship yields statistical prediction
  skill for the monthly-mean AO index.
• Changes to the circulation of the lower
  stratosphere affect surface climate.
• The future evolution of the stratosphere will
  depend on increasing GHGs, ozone changes, and
  changes to the troposphere.
• We do not know how the stratosphere will change
  in the coming decades and centuries, and we do
  not know how surface climate will be affected.
• We do not fully understand the dynamics of
  stratosphere-troposphere coupling.

mark_at_nwra.com www.nwra.com/baldwin