Executive Summary
Introduction
This chapter assesses regional climate information from Atmosphere-Ocean General
Circulation Models (AOGCMs) and techniques used to enhance regional detail.
These techniques have been substantially improved since the IPCC WGI Second
Assessment Report (IPCC, 1996) (hereafter SAR) and have become more widely applied.
They fall into three categories: high and variable resolution Atmosphere General
Circulation Models (AGCMs); regional (or nested limited area) climate models
(RCMs); and empirical/statistical and statistical/dynamical methods. The techniques
exhibit different strengths and weaknesses and their use depends on the needs
of specific applications.
Simulations of present day climate
Coarse resolution AOGCMs simulate atmospheric general circulation features well
in general. At the regional scale the models display area-average biases that
are highly variable from region-to-region and among models, with sub-continental
area-averaged seasonal temperature biases typically within 4ºC and precipitation
biases mostly between -40 and +80% of observations. In most cases, these represent
an improvement compared to the AOGCM results evaluated in the SAR.
The development of high resolution/variable resolution AGCMs since the SAR
shows that the models’ dynamics and large-scale flow improve as resolution
increases. In some cases, however, systematic errors are worsened compared with
coarser resolution models although only very few results have been documented.
RCMs consistently improve the spatial detail of simulated climate compared
to General Circulation Models (GCMs). RCMs driven by observed boundary conditions
show area-averaged temperature biases (regional scales of 105 to 106 km2) generally
within 2ºC and precipitation biases within 50% of observations. Statistical
downscaling demonstrates similar performance, although greatly depending on
the methodological implementation and application.
Simulation of climate change for the late decades of the 21st century
Climate means
The following conclusions are based on seasonal mean patterns at sub-continental
scales emerging from current AOGCM simulations. Based on considerations of consistency
of changes from two IS92a-type emission scenarios and preliminary results from
two SRES emission scenarios, within the range of these four scenarios:
- It is very likely that: nearly all land areas will warm more rapidly than
the global average, particularly those at high latitudes in the cold season;
in Alaska, northern Canada, Greenland, northern Asia, and Tibet in winter
and central Asia and Tibet in summer the warming will exceed the global mean
warming in each model by more than 40% (1.3 to 6.9°C for the range of
models and scenarios considered). In contrast, the warming will be less than
the global mean in south and Southeast Asia in June-July-August (JJA), and
in southern South America in winter.
- It is likely that: precipitation will increase over northern mid-latitude
regions in winter and over northern high latitude regions and Antarctica in
both summer and winter. In December-January-February (DJF), rainfall will
increase in tropical Africa, show little change in Southeast Asia and decrease
in central America. There will be increase or little change in JJA over South
Asia. Precipitation will decrease over Australia in winter and over the Mediterranean
region in summer. Change of precipitation will be largest over the high northern
latitudes.
Results from regional studies indicate that at finer scales the changes can
be substantially different in magnitude or sign from the large area average
results. A relatively large spread exists between models, although attribution
is unclear.
Climate variability and extremes
The following conclusions are based on patterns emerging from a limited number
of studies with current AOGCMs, older GCMs and regionalisation studies.
- Daily to interannual variability of temperature will likely decrease in
winter and increase in summer in mid-latitude Northern Hemisphere land areas.
- Daily high temperature extremes will likely increase in frequency as a function
of the increase in mean temperature, but this increase is modified by changes
in daily variability of temperature. There is a corresponding decrease in
the frequency of daily low temperature extremes.
- There is a strong correlation between precipitation interannual variability
and mean precipitation. Future increases in mean precipitation will very likely
lead to increases in variability. Conversely, precipitation variability will
likely decrease only in areas of reduced mean precipitation.
- For regions where daily precipitation intensities have been analysed (e.g.,
Europe, North America, South Asia, Sahel, southern Africa, Australia and the
South Pacific) extreme precipitation intensity may increase.
- Increases in the occurrence of drought or dry spells are indicated in studies
for Europe, North America and Australia.
Tropical cyclones
Despite no clear trends in the observations, a series of theoretical and model-based
studies, including the use of a high resolution hurricane prediction model,
suggest:
- It is likely that peak wind intensities will increase by 5 to 10% and mean
and peak precipitation intensities by 20 to 30% in some regions;
- There is no direct evidence of changes in the frequency or areas of formation.
Recommendations
The material assessed identifies key priorities for future work:
GCMs:
- Continued improvement in GCMs, as their use is fundamental to deriving regional
climate information.
- GCM simulations with a greater range of forcing scenarios and an increased
ensemble size to assess the spread of regional predictions.
- More assessment of GCM regional attributes and climate change simulations.
- A much greater effort in the evaluation of variability (daily to interannual)
and extreme events.
RCMs:
- A more systematic and wide application of RCMs to adequately assess their
performance and to provide information for regional scenarios.
- Ensemble RCM simulations with a range of regional models driven by different
AOGCM simulations.
- A much greater effort in the evaluation of variability (daily to interannual)
and extreme events.
Empirical/statistical and statistical/dynamical methods:
- More regional observations to provide for more comprehensive statistical
downscaling functions.
- Much further work to identify the important climate change predictors for
statistical downscaling.
- Application of different techniques to a range of AOGCM simulations.
Tropical cyclones:
- A greater range of models and techniques for a comprehensive assessment
of the future behaviour of tropical cyclones.
Cross-cutting:
- Systematic comparisons of the relative strengths and weak-nesses of techniques
to derive regional climate information.
- The development of high-resolution observed climatologies, especially for
remote and physiographically complex regions.
- A systematic evaluation of uncertainties in regional climate information.
