Transcript ppt - Harvard University

Trends in Mid-Latitude Cyclone Frequency And
Their Effect On Air Quality
Eric M. Leibensperger, Loretta J. Mickley, and Daniel J. Jacob
School of Engineering and Applied Sciences - Harvard University
GCAP Meeting – October 12, 2007
A Decrease in Mid-Latitude Cyclones: 2000-2050
1999-2001
2049-2051
[Wu et al., in press]
1999-2001 vs. 2049-2051: 17% decrease
1999-2005 vs. 2049-2055: 16% decrease
Past and Future Trends of Mid-Latitude Cyclones
Previous studies have found a decrease in the frequency
of observed cyclones and cyclones objectively tracked
from reanalysis data.
Additional studies have identified a decrease in cyclones
in future climate simulations.
High-latitudes
[McCabe, et al., 2001]
Mid-latitudes
[Lambert, et al., 2006]
Storm Tracking
[Zishka and Smith, 1981]
4x daily SLP fields (NCEP/NCAR or
ECMWF Reanalysis, GISS GCM 3
4x5 and 2x2.5) used with the GISS
Storm Tracker [Bauer and Del
Genio, 2007]
720 km
990 mb
1000 mb
1950-2005 Summer Cyclone Frequencies
NCEP/NCAR Reanalysis
GISS GCM 3– 2x2.5 – Specified SST
(data courtesy of Jean Lerner)
GISS GCM 3– 4x5 – Q-Flux Ocean
1980-2005 trend is comparable
between GISS GCM 4x5 and
Reanalysis
[Leibensperger et al., in prep.]
Relationship between Ozone and Mid-latitude Cyclones –
1980-1998
Correlation coefficient between # of
cyclones and # of O3 exceedances
The frequency of cyclones tracking
through the southern track is strongly
anti-correlated with the number of
ozone exceedances
[Leibensperger et al., in prep.]
Cyclone Frequency and Ozone in the Northeast
r=-0.29
r=-0.64
Removing linear trends in cyclone
frequency and exceedances
enhances the anti-correlation
[Leibensperger et al., in prep.]
The Air Quality Penalty Due to Decreasing Cyclone
Frequency
Without the decrease in the number of
cyclones, the number of exceedances
would have been lowered to 5 rather than
19 days.
Future decreases in cyclone frequency will
make air quality goals more difficult to
achieve.
19 – 5 = 14 days
 d (exceedances)   d (exceedances) 
 d (exceedances) 





dt
dt
dt

 
emissions 
cyclones
d (exceedances ) d (cyclones )
-1
 d (exceedances ) 

 (4.15)( 0.19)  0.79 days yr


dt
d (cyclones )
dt

 cyclones
-1
 d (exeedances) 

  1.05 days yr
dt


-1
 d (exceedances ) 
 1.84 days yr


dt

 emissions
[Leibensperger et al., in prep.]
Future Work
• Extend analysis from 1998 to
2006, publish results
COt = Carbon Monoxide Tracer
• Use tracer within GCM to
diagnose changes in transport
• Effects of climate change on air
pollution meteorology
- Transport (cyclones,
convection)
[Mickley et al., 2004]
CPD of CO tracer at the surface during low
cyclone years versus high cyclone years is
similar to Loretta’s findings for 2050
- Lightning
EXTRA SLIDES
Northeast at surface
Midwest at surface
Southeast at surface
Northeast at 400hPa
Southeast at 400hPa
Midwest at 400hPa
Less tracer makes it this
high
Less outflow from N.A. – but more to the north,
could be from increased frequency there
Most tracer sourced where lower track most
effective – net effect is less in upper trop.
Time Series of Cyclones – NCEP/NCAR Reanalysis 1
Tropical Storm Brenda
Hurricane Cleo
Time Series of Cyclones – GISS GCM 4x5
Time Series of Cyclones – GISS GCM 2x2.5
1. Can the GCM and reanalysis data reproduce expected cyclone statistics?
[Whittaker and Horn, 1984]
(cyclones/20 years)
60
40
20
GISS 4x5 and NCEP main storm tracks
are comparable to the observed main
track frequencies in position. NCEP
storm tracks well represented
quantitatively, but GISS 4x5 is too low.
[Whittaker and Horn, 1984]
Cyclones in the GISS GCM III at 2ºx2.5º
(cyclones/20 years)
The frequency of cyclones in the GISS
GCM 2x2.5 is improved, but the tracks
appear shifted to the southeast.
60
40
20
Main Summer Storm Tracks in the Models
Why mid-latitude cyclones?
Mean 4-10 Day Integrated Power of FFT
of Max 8-hr Avg Time Series
Cold
Warm
Lighter winds
In the northeastern United States, many variables affecting ozone production
(temperature, cloud cover, stagnation) and transport change on a synoptic
timescale thus making mid-latitude cyclones a major driver of surface ozone
concentrations