Weather and Climate Extremes in a Changing Climate. Regions of ...

Vecchi and Soden (2007) have assessed the
different components of the Emanuel and
Nolan (2004) scheme using outputs from the
IPCC AR4 models. Their results suggest that
a decrease in tropical cyclone frequency may
occur over some parts of the Atlantic basin
associated with a SW-NE oriented band of less
favorable conditions for tropical cyclogenesis
and intensification, including enhanced vertical
wind shear, reduced mid-tropospheric relative
humidity, and slight decrease in potential intensity.
The enhanced vertical shear feature (present
in about 14 of 18 models in the Caribbean
region) also extends into the main cyclogenesis
region of the Eastern Pacific basin. Physically,
this projection is related to the weakening of
the east-west oriented Walker Circulation in the
Pacific region, similar to that occurring during
El Niño events. During El Niño conditions in
the present-day climate, hurricane activity is
reduced, as occurred, for example, in the latter
part of the 2006 season. While this projection
may appear at odds with observational evidence
for an increase in Atlantic tropical cyclone
counts during the past century (Holland and
Webster, 2007; Vecchi and Knutson, 2008),
there is evidence that this has occurred in
conjunction with a regional decreasing trend
in storm occurrence and formation rates in
the western part of the Caribbean and Gulf of
Mexico (Vecchi and Knutson, 2008; Holland,
2007). Earlier, Knutson and Tuleya (2004) had
examined the vertical wind shear of the zonal
wind component for a key region of the tropical
Atlantic basin using nine different coupled
models from the CMIP2+ project. Their analysis
showed a slight preference for increased vertical
shear under high CO 2 conditions if all of the
models are considered, and a somewhat greater
preference for increased shear if only the six
models with the most realistic present-day
simulation of shear in the basin are considered.
Note that these studies are based on different
sets of models, and that a more idealized future
forcing scenario was used in the earlier Knutson
and Tuleya study.
Alternative approaches to the empirical analysis
of large-scale fields are the global and regional
climate simulations, in which the occurrence
of model tropical cyclones can be tracked.
Beginning with the early studies of Broccoli
and Manabe (1990), Haarsma et al. (1993), and
Weather and Climate Extremes in a Changing Climate
Regions of Focus: North America, Hawaii, Caribbean, and U.S. Pacific Islands
Bengtsson et al. (1996), a number of investigators
have shown that global models can
generate tropical cyclone-like disturbances in
roughly the correct geographical locations with
roughly the correct seasonal timing. The annual
occurrence rate of these systems can be quite
model dependent (Camargo et al., 2005), and is
apparently sensitive to various aspects of model
physics (e.g., Vitart et al., 2001).
The notion of using global models to simulate
the climate change response of tropical cyclone
counts is given some support by several studies
showing that such models can successfully
simulate certain aspects of interannual
to interdecadal variability of tropical-cyclone
occurrence seen in the real world (Vitart et al.,
1997; Camargo et al., 2005; Vitart and Anderson,
2001). A recent regional model dynamical
downscaling study (Knutson et al., 2007) with
an 18 km grid model and a more idealized
modeling approach (Emanuel et al., 2008)
both indicate that the increase in hurricane
activity in the Atlantic from 1980-2005 can be
reproduced in a model using specified SSTs and
large-scale historical atmospheric information
from reanalyses.
Since tropical cyclones are relatively rare
events and can exhibit large interannual to
interdecadal variability, large sample sizes (i.e.,
many seasons) are typically required to test the
significance of any changes in a model simulation
against the model’s “natural variability.”
The most recent future projection results obtained
from medium and high resolution (120
km-20 km) GCMs are summarized in Table
3.2. Among these models, those with higher
resolution indicate a consistent signal of fewer
tropical cyclones globally in a warmer climate,
while two lower resolution models find essentially
no change. There are, however, regional
variations in the sign of the changes, and these
vary substantially between models (Table
3.2). For the North Atlantic in particular, more
tropical cyclones are projected in some models,
despite a large reduction globally (Sugi et al.,
2002; Oouchi et al., 2006), while fewer Atlantic
tropical cyclones are projected by other models
(e.g., McDonald et al., 2005; Bengtsson et al.,
2007). It is not clear at present how the Sugi et
al. (2002) and Oouchi et al. (2006) results for
While there is
recent observational
evidence for an
increase in the
number of tropical
storms and
hurricanes in the
Atlantic, a confident
assessment of future
storm frequency
cannot be made at
this time.
111