Transcript Columbia_CS_masters - UCLA: Atmospheric and Oceanic

Oct 17, 2003 MISR
REGIONAL CLIMATE
CHANGE SIMULATIONS
Columbia University, April
3, 2006
One-way nested modeling has been applied to climate
change studies in the last few years. This technique
consists of using output from GCM simulations to
provide initial and driving lateral meteorological
boundary conditions for high-resolution regional climate
model simulations, generally with no feedback from the
regional climate model to the driving GCM. Hence, a
regional increase in resolution can be attained through
the use of nested regional climate models to account for
sub-GCM grid-scale forcings.
This is attractive because regional models can
introduce sensitivities at the landscape scale to the
simulation. Moreover, as climate change impacts on
ecosystems, air pollution, water resources etc. take
place on the comparable scales, they may provide more
relevant information to models of other earth system
components.
SIMULATING
PRESENT-DAY
CLIMATE
Experimental Design
model: MM5
boundary conditions: eta
model reanalysis
resolution: domain 1, 54
km, domain 2, 18 km,
domain 3, 6 km
time period: 1995 to
present.
One can think of this as a
reconstruction of weather
conditions over this time
period consistent with
three constraints: (1) our
best guess of the largescale conditions, (2) the
physics of the MM5
model, and (3) the
prescribed topography,
consistent with model
resolution.
VALIDATION
This shows the correlation of
daily-mean wind speed and
direction between model and
observations at 18 locations.
Both wind speed and direction
are reasonably well-correlated
at all locations, in spite of
many possible reasons for
disagreement,
including
systematic
and
random
measurement error.
This
gives confidence that the
timing and magnitude of
circulation
anomalies
are
correct in the simulation.
GLOBAL CLIMATE
CHANGE
SIMULATIONS
WATER VAPOR FEEDBACK
Increase in
temperature
Enhancement
of the
greenhouse
effect
Water vapor feedback is
thought to be a positive
feedback mechanism.
Water vapor feedback
might amplify the
climate’s equilibrium
response to increasing
greenhouse gases by as
much as a factor of two.
It acts globally.
Increase in
water vapor
in the
atmosphere
SURFACE ALBEDO FEEDBACK
Increase in
temperature
Increase in
incoming
sunshine
Surface albedo feedback
is thought to be a
positive feedback
mechanism. Its effect is
strongest in mid to high
latitudes, where there is
significant coverage of
snow and sea ice.
Decrease in
sea ice and
snow cover
Equilibrium response of a
climate model when
feedbacks are removed.
To understand how cloud feedback might work, you
have to understand some facts about clouds:
(1) Clouds absorb radiation in the infrared, and
therefore have a greenhouse effect on the climate. If
you put a cloud high in the atmosphere, it will have a
stronger greenhouse effect than if you put it low in the
atmosphere.
(2) Clouds reflect sunshine back to space. So more
clouds means less sunshine for earth. If you put a
cloud high in the atmosphere, it will reflect about the
same amount of sunshine as if you put a cloud low in
the atmosphere.
Decrease in
cloudiness?
Reduced
greenhouse
effect
Increased
sunshine
Increase in
temperature
CLOUD FEEDBACK
Enhanced
greenhouse
effect
Increase in
cloudiness?
Reduced
sunshine
Which effect is
stronger depends
on the geographical
and vertical
distribution
of the decrease
in cloudiness
Models predict both an
increase and decrease in
cloudiness, and both
positive and negative
cloud feedbacks.
Which effect is
stronger depends
on the geographical
and vertical
distribution
of the increase
in cloudiness
Transient vs Equilibrium climate response
Transient response refers to the evolution of
the climate system as it responds to external
forcing, such as an increase in greenhouse
gases.
Equilibrium response refers to the final
state of the climate system after it has
adjusted to the external forcing. The
magnitude of the equilibrium response
compared to the magnitude of the forcing is
referred to as the climate sensitivity.
Evolution of simulated global mean temperature when CO2 changes
This shows the warming in
a climate model when two
scenarios of CO2 increases
are imposed: one is a 1%
per year increase in CO2
leading to a CO2 doubling,
and the other is an increase
at the same rate leading to
a CO2 quadrupling.
It
shows that the warming
continues
for
several
centuries even when CO2
levels
are
stabilized,
leading
to
significant
differences
between
transient and equilibrium
climate
responses
to
external forcing.
The difference between the transient and equilibrium responses of a
climate model to increasing greenhouse gases varies a great deal
geographically.
The colors show 21st century warming taking place in response to a plausible
scenario of radiative forcing. The values are averaged over all the ~20 simulations
used in the most recent UN Intergovernmental Panel on Climate Change Report.
The warming is calculated by subtracting temperatures at the end of the 20th
century (1961-1990) from temperatures at the end of the 21st century (2071-2100).
The thin blue lines show the range in warming across all the models.
The thick green lines show the ratio of the mean change in
temperature to the standard deviation of the temperature change.
HYDROLOGIC CYCLE
Increase in
greenhouse gases
means more
longwave
radiation reaches
the surface
Increase in
temperatures
favors loss of
surface heat through
evaporation rather than
sensible heat
INTENSIFICATION
Increase in
evaporation
(fairly
uniform
globally)
Increase in
precipitation
(not
uniform)
Colors show the simulated 21st century percent change in precipitation
averaged over the simulations of the UN Intergovernmental Panel on
Climate Change 3rd assessment report.
The red lines show the range in the percent increase in precipitation.
The thick green lines show the ratio of the mean change in
precipitation to the standard deviation of the precipitation change.
LESSONS FROM
EXAMPLES OF
REGIONAL
CLIMATE CHANGE
PREDICTION
Leung and Ghan (1999)
imposed
2XCO2
conditions from CCM3
on a regional model
based on MM5 covering
the Pacific Northwest.
They found the regional
model has a different
temperature sensitivity
than the parent model,
suggesting feedbacks
operating on scales
smaller
than
those
resolved by the parent
model are influencing
sensitivity.
They suggest this stems from the fact that in the parent model snow
feedback processes are inadequately resolved in this region of intense
topography.
The regional simulation shows a large reduction in
snow cover at elevations near the present-day snow
line (1000-2000 m). This is due to reduced snowfall
and increased snow melt.
The Union of Concerned Scientists recently published an
assessment of climate change in California.
They based their assessment on the results from two global climate
models, one with a relatively low sensitivity to CO2 doubling (PCM),
and the other with a relatively high sensitivity (HADCM3).
They looked at outcomes in California for two scenarios. One is
“business as usual” scenario, that envisages fossil fuel emissions
increasing at approximately the same rate as present for the
remainder of the 21st century. The other is a lower emissions
scenario, where emissions continue to increase but at a lower rate,
stabilizing around 2050, then declining to levels below the present
level by 2100.
The global models’ resolutions are on the order of 200 km. Regional
details have been supplied using a statistical downscaling
technique.
Diminishing Sierra Snowpack
% Remaining, Relative to 1961-1990
This shows how the more
sensitive global model
projects snowpack to
change in the Sierras.
The change in snowpack is
significant because it
comprises approximately
half the total water storage
capacity of California, the
other half being contained
mainly in human-made
reservoirs.
Source: A Luers/Union of Concerned Scientists
Most precipitation over the
Sierras falls in wintertime,
where it is stored in the
snow pack. The snowpack
comprises approximately
half the total water storage
capacity of California, the
other half being contained
mainly in human-made
reservoirs.
As the snow
melts, water
flows to
reservoirs,
where it makes
its way through
aqueducts to
agricultural and
urban areas.
This shows
aqueducts for
water resource
re-distribution
in California
Percent of Water Year Runoff
The Sierra snow pack has been steadily shrinking over the past century…
70%
70%
65%
65%
60%
60%
55%
55%
50%
50%
45%
45%
40%
40%
35%
35%
30%
30%
25%
25%
20%
20%
15%
15%
10%
10%
1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Water Year (October 1 - September 30)
Sacramento River Runoff (1906-2001)
April to July as a Percent of Total Runoff
Source: California Protection Agency, Environmental Protection Indicators for California, 2001
The large-scale conditions in this regional simulation originated from transient
climate-change simulations using the coupled ocean–atmosphere global
climate model ECHAM4/OPYC (300-km grid). The regional details are supplied
by a 50 km regional model (HIRAM). (Change = 2071-2100 minus 1961-1990)
Relative percentage change in
precipitation for July–September
in the Intergovernmental Panel
on Climate Change's A2 scenario
with respect to the present day
The relative change in the mean
five-day precipitation for July–
September that exceeds the 99th
percentiles in scenario A2 with
respect to the control
Christensen & Christensen, Nature 2003
How credible are these results in light of this divergence in the
global models?
CONCLUSIONS
Like their global counterparts, predictions of
regional climate change are most credible when
they involve a plausible set of physical
mechanisms. So far, the regional climate
predictions that can be taken the most seriously
involve temperature-cryosphere feedbacks.
Regional climate predictions are subject to
misinterpretation because their high resolution
gives a false impression of fidelity. This is
particularly true in the case of the predictions of
local changes in the hydrologic cycle. There is
huge divergence in hydrologic cycle
intensification, and these signals are poorly
understood in global models.