Transcript Slide 1

Lecture 33: Climate Changes: Past & Future (Ch 16)
“Climate change” – change in any statistical property of the atmosphere
This chapter will clarify our thinking:
• climate change relative to what “normal”?
• is “climate warming” happening?
• if so, is it man-induced, or not?
Iceberg 100km east of
Dunedin, South Island,
New Zealand
Dunedin 16 November 2006… closest sighting off New Zealand for 75 years
according to NIWA (National Institute of Water and Atmospheric Research, NZ)
oC
Sec. 16-2 Fig. 1
1880
1980
Logical to consider earth’s climate to be a function of these external factors (the text
calls them “boundary conditions”:
• intensity of sunlight (solar output, sun-earth geometry)
• arrangement of continents and oceans
• composition of the atmosphere
Some believe earth does not have a unique climate for fixed values of the
above (“earth’s climate intransitive”). Perhaps it would be surprising if it did –
that would have to mean these factors control earth’s processes down to a
surprising level of detail, eg. evolution of plants …
• present climate is unusual - earth mostly has
been considerably warmer than now
• most of its 4.5B yr life, earth free of
permanent ice
• warmth (perhaps 5 to 15oC warmer global
mean!) punctuated by seven ice ages
• all human existence has been spent in the last
of these (last 15 MY)
T
cold
warm
Long-Term Changes
The long term (100's million years) paleoclimate record is characterized by
relatively few, isolated glacial outbreaks - the great Ice Ages.
• can’t be explained by variations in solar output
• can’t be explained by orbital changes (which are too fast)
• best guess today associated with "Plate Tectonics" and its influence on the
atmospheric greenhouse effect.
www.globalchange.umich.edu/
• within “our” ice age, numerous climate oscillations (“glacial/interglacial cycles”)
Fig. 16-3
oC
• planet is now in a warm
interglacial
From Antarctic (Vostok) ice core
record…
From Antarctic (Vostok) ice core record… “Both methane and carbon dioxide
correlate with temperature - i.e., an increase in temperature is associated with
an increase in the abundance of both these two gases. It is unclear whether the
gas abundance changes are a consequence of the temperature changes or vice
versa.”
http://www.ucar.edu/
Maximum extent of ice, last glaciation
Fig. 16-4
• smaller changes to
Antarctic ice sheet
• but timing of major
cooling/warming cycles
in step over past 150KY
• a 5 degree change in
global mean temp
suffices to cause massive
change!
Laurentide
Ice Sheet
• millenial scale oscillations suggest a climate “flip-flop” (two climate states for
given boundary conditions)
• on this timescale “climate” entails state of oceans too
Fig. 16-7
So is earth’s climate warming?
(Yes or no, depending on the time scale
on which we view the record)
So-called “hockey stick”
Fig. 16-6
“We must now ask if this recent rise in temperature is just part of the natural
variability in climate, or if it marks the onset of human-induced warming from
emission of “greenhouse gases” into the atmosphere. At present, it is impossible to
conclude one way or the other with any certainty.” (p484)
24 November 2005 (New Scientist)
“The longest ice-core record of climate history ever obtained … shows that levels of
greenhouse gases really do march in lockstep with changes in temperature.
The frozen record of the Earth's atmosphere is 3270 metres long and covers the last
650,000 years. It was obtained from the tiny air bubbles trapped in a deep ice core
from Antarctica.
The bubbles record how the planet’s atmosphere changed over six ice ages and the
warmer periods in between. But during all that time, the atmosphere has never had
anywhere near the levels of greenhouse gases seen today. Today's level of 380 parts
per million of carbon dioxide is 27% above its previous peaks of about 300 ppm,
according to the team led by Thomas Stocker of the University of Bern.”
Since our climate models, albeit imperfect, do anticipate warming due
to rising CO2, it is not illogical to suggest the recent warming is a
response – though we cannot prove this is so.
What are the paleo-climatological lines of evidence?
A vast array of techniques has been applied to extend the instrumental record.
Typically, it is held there is a correlation between the observed quantity, and some
climate statistic. There will be some form of “calibration” of the relationship from
a known record. Proxy climate indicators include
• tree growth rings (going back several KY): correlation with temp & precip
• oxygen isotopic content of sediments of marine organisms
• pollen (whose dating connects vegetation types with time)
• ice cores (to 650KYBP)
Ice cores
• bubbles in the ice give a direct sample of past air chemistry
Ice cores
• snow that falls during period of warmer
climate has higher ratio of 18O to 16O… the
connection with temperature is indirect, but
experts don’t doubt its validity
“paleoclimatologists have conducted a number of tests to
calibrate this "paleoclimate thermometer" in the ice.
Figure 1. Ice-core oxygen isotopic measurements from
Greenland (right hand side) and from Antarctica (left hand
side). The isotope measurements can be interpreted to yield
the global sea surface temperatures to ~160,000 years ago
(colder temperatures to the left). The two traces are
consistent with each other and depict the most recent glacial
period, ending ~15,000 years ago.
A decrease of one part per million (ppm) in the 18O
measurement is equivalent to a reduction in temperature of
approximately 1.5oC at the time that the water evaporated
from the oceans.”
www.globalchange.umich.edu/
Factors involved in climatic change
• varying solar output? … inconclusive: correlations have been found on some
timescales
• changes in earth’s orbit (Milankovitch cycles)… “widely accepted as driving
glacial/interglacial cycles”. Their periods (100, 41, 11) KY are short compared to
the very long term record.
• changing continent/ocean distribution
• atmospheric composition
• tropospheric aerosols
• stratospheric aerosols
• CO2, methane,…
“it should not be surprising that there are a host of processes and conditions that are
not known well enough for us to establish with certainty the exact outcomes” (of the
buildup of CO2) p494
Global Climate Modelling**
History
 possibility that climate could be affected by changing
concentrations of greenhouse gases first put forward by Arrhenhius
(1896; “On the influence of carbonic acid in the air upon the
temperature of the ground”. Philos. Mag., Vol. 41, 237–276)
 mid C20th attempts were made to estimate the equilibrium
temperature rise due to doubling of atmos. CO2, based largely on
radiative equilibrium calculations
 1967 importance of convective processes in regulating the surface
temperature of the earth was taken into account by Manabe and
Wetherald (J. Atmos. Sci. 24, 241-259).
**Mitchell (2004, “Can we believe predictions of climate change?”
Quart. J. Royal. Meteorol. Soc., Vol. 130, pp. 2341–2360)
History (ctd…)
Zonally averaged atmospheric temperature changes due to doubling atmospheric
CO2. Contours are every oC, stippled (grey) where negative and cross-hatched where
greater than +4 oC (from Manabe and Wetherald, 1975, “The effects of doubling the
CO2 concentration on the climate of a general circulation model.” J. Atmos. Sci., Vol.
32, 3–15).
Climatic warming under CO2 doubling:
(Canadian Climate Centre model)
Current estimates using
atmos. models coupled to a
simple ocean give a range
of 2 to 6oC for global mean
temperature response to
CO2 doubling
(latitudedependent)
Fig. 16-13a
How does forecasting climate differ from forecasting weather?
 many more processes, acting on longer timescales, need to be included, eg.
 ocean temperature (& salinity) changes
 ocean circulations influencing CO2 budget
 sun-earth geometry changes
 locations of continents?
 ice sheets and ice packs
 vegetation responses interacting with CO2, temperature and humidity
 natural aerosols
 anthropogenic gases and particles
 climate simulation computes the equilibrium climate for certain fixed “external”
conditions (eg. perhaps fixed ocean temps; fixed CO2; fixed sun-earth geometry).
Thus initial conditions are irrelevant (one integrates for long enough to “forget” the
initial condition).
 It may be possible to neglect or simplify some “rapid” processes, and even to
neglect a spatial dimension, eg. zonally-averaged models
Uncertainties in Climate Modelling using GCM’s
The main uncertainties arise with processes for which we do not have a reliable
underlying theory (including cloud formation and dissipation), and processes
which are not resolved on the model grid (including transfer of heat, moisture and
momentum from the surface, convection and cloud processes)
There remain model parameters which cannot be measured or do not correspond
to any measurable quantity, eg. some cloud parametrizations define a relativehumidity threshold above which cloud is allowed to form. Even if there is a single
threshold in the real world, it is unlikely that using it would give the correct cloud
amount... Small errors in cloud amounts and microphysical properties can produce
large errors in the radiative budget, and hence large drifts in surface temperature.
“Positive feedbacks” (see Sec. 6-3) are those which reinforce (or act additively
with) the original disturbance, eg. the ice albedo feedback.
“Negative feedbacks” oppose the root disturbance. Thus if global warming
increases global cloud coverage, increased solar reflection is a negative feedback,
but increased absorption of upwelling longwave radiation is a positive feedback.
Overall cloud feedback is a complex sum of several feedbacks: GCM’s disagree on
overall sign!
There are complex feedbacks whose parametrization needs to be refined, e.g.
dimethyl sulphide (DMS) gas, released by decay of ocean biota, forms sulphate
aerosols that act as CCN: will warmer ocean temperatures mean greater ocean
productivity and consequently greater biotic decay rate, causing higher
atmospheric concentrations of CCN and changes to cloud amount and type?
Global Climate Modelling
Four criteria to judge ‘is a climate model reliable for predicting climate
change?’
 physical basis
 simulation of present climate
 simulation of historical climate (period of instrumental records, or equilibrium
simulation of much more distant climates, eg. Last Glacial Maximum, 21kBP, ie.
21,000 years ago)
 numerical weather prediction
Only a few fully coupled simulations have been published to date, but these all
show global-scale cooling broadly consistent with the paleoclimatic
reconstructions... there is still little or no confidence in the regional detail
predicted by models... most of the range in climate sensitivity across various
GCM’s is associated with differences in cloud feedback
The second phase of the Paleoclimate Modeling Intercomparison Project is coordinating
simulations and data syntheses for the Last Glacial Maximum (LGM; 21000 yr before
present; 21 ka) and mid-Holocene (6000 yr before present; 6 ka) to contribute to the
assessment of the ability of current climate models to simulate climate change. Here the
“Community Climate System Model version 3” shows global cooling of 4.5°C
compared to pre-industrial (PI) conditions with amplification of this cooling at high
latitudes and over the continental ice sheets present at LGM.
Change in mean
annual surface
temperature (°C)…
LGM minus PreIndustrial
Otto-Bliesner et al. (2006; J. Climate, Vol. 19)
The forcings changed for the LGM are reduced atmospheric greenhouse gases, a 2–3km ice sheet over North America and northern Europe, lowered sea level resulting in
new land areas, and small Milankovitch anomalies in solar radiation. The reduced LGM
levels of atmospheric CO2 are 66% of preindustrial levels and 55% of present levels.
Change in mean
annual surface
temperature (°C)…
LGM minus PreIndustrial
Otto-Bliesner et al. (2006; J. Climate, Vol. 19)
U. Vic Coupled Ocean-Atmosphere Climate Model (“intermediate complexity”)
Ocean
 3-dimensional
 3.6o zonal x 1.8o meridional
 19 vertical levels (z=50 m near surface, z =500 m near ocean bottom)
 dynamic-thermodynamic sea ice (ie. winddriven motion; melting)
 inorganic carbon cycle
Atmosphere
 2-dimensional, ie. vertically well-mixed (single layer), horiz. resolution same as
ocean
 horizontal diffusion of energy and moisture, present-day climatological surface
winds determine sea-air transfer coefficient for heat, CO2, etc.
 Precipitation when RH > 85% returns instantly to ocean via one of 33 rivers
U. Vic Coupled Ocean-Atmosphere Climate Model (“intermediate complexity”)
Land
 dynamic vegetation (responds to climate, incl. CO2)
 terrestrial carbon cycle
 ice & snow albedo feedback
 specified lapse rate used to reduce surface temperature over topography
Performance
 does not need air-sea flux adjustments to keep present climate stable
 when forced by historical CO2 emissions reproduces historic CO2 trends
 soon to be used for 120,000 year simulation to examine glacial-interglacial
transitions