Transcript Document

Glacier modeling: Current Status and Needed Improvements
Sarah Raper
Centre for Air Transport and the Environment
Manchester Metropolitan University
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Schematic diagram illustrating the cascade of uncertainties
in modelling the contribution to SLR from glaciers and ice caps
glacier area size
distribution
climate and
climate change
climate
ice dynamics
downscale
climate change
hypsometry
mass balance
change in
volume
change in
sea level
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Location of glaciers and icecaps
glaciers in Greenland and
Antarctica are problematic
Estimates often exclude these for the following reasons:
• inventory data for these regions is poor or non-existent.
• it is not clear where the ice sheets end and the glaciers begin
in for instance the Antarctic Peninsula.
• ice-sheet retreat under global warming will result in ice masses
becoming separated from the ice-sheets, thus increasing the
ice volume in the glacier and ice cap category.
From Braithwaite and Raper, 2002
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Area and Volume size distribution
Volume estimates in sea level equivalent
Outside Greenland and Antarctica:
0.15m
Ohmura (2004)
0.24 (0.26)m Raper and Braithwaite (2005)
0.37 m
Dyurgerov and Meier (2005)
Greenland and Antarctica (excluding ice-sheets):
0.34 m
Dyurgerov and Meier (2005)
Need to make the addition of the largest glaciers to the
glacier inventory a priority.
adapted from Raper and Braithwaite, 2005
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Uncertainty in ice volume from the previous
slide is about  45%, leading to uncertainty in
glacier and icecap contribution to 21st C SLR
Sensitivity of modelled glacier and ice cap SLR contribution to
±10% uncertainty in various reference period input s based on
Raper and Braithwaite (2006).
Reference (1961-1990) 20thC
SLR
Volum e* ±10%
~ 0%
21stC
SLR
±4%
Mass balance
* for a single estimate of reference period area
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Observationally based global estimate of mass balance
period
1960-1992
mm SLE a-1
+ 0.35  0.26
Consensus estimate based on Ohmura (2004), Cogley (2005),
Dyurgerov and Meier (2005)
The above includes glaciers and ice caps in Greenland and Antarctica.
Uncertainty in global mass balance from the previous
slide is about  75%, leading to a major uncertainty in
glacier and icecap contribution to 20th C SLR
Sensitivity of modelled glacier and ice cap SLR
contribution to ±10% uncertainty in various
reference period inputs based on Raper and
Braithwaite (2006).
Reference (1961-1990)
Volume* ±10%
Mass balance ±10%
Mass balance gradients ±10%
* single estimate of area
20thC
SLR
~ 0%
21stC
SLR
±4%
±9%
±1%
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Mass Balance Modeling
concerned with the interaction of the climate with the glacier surface
energy balance models
reviewed in Hock (2005)
degree-day models
(temperature-index models)
reviewed in Hock (2003)
•physically based
•require detailed input data
•more suitable for high
resolution in space and time.
•empirically based
•main inputs temperature
and precipitation are readily
available in gridded form
from AOGCMs.
Challenge for SLR is to interpolate mass balance model results to give
estimates of time evolution of mass balance over all glaciers and icecaps.
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Mass balance sensitivity is the change in mass balance
for a 1oC warming
ELA
• Fit mass balance model to glacier
• Increase temperature by 1oC
• Compared here are changes for a
maritime glacier (warm-wet) and
a continental glacier (dry-cold)
supplied by Braithwaite
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Mass balance sensitivity can be estimated from precipitation
• most glacier and ice cap
projections for SLR
are based on
mass balance sensitivity
(Gregory and Oerlemans
1998)
• it is necessary to scale
down the glacier area as
the volume decreases
(van de Wal and
Wild, 2001)
• methodological problem:
using this method all the
ice would eventually melt
for any warming
From Braithwaite and Raper, 2002
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Mass balance sensitivity is the change in mass balance
for a 1oC warming
• to get the mass balance
sensitivity we need to
integrate over the area altitude
distribution (hypsometry)
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
An alternative is to extrapolate the mass balance
gradients to all glacier regions
• mass balance model fitted for grid
points in seven regions with good data
• regress the resulting balance
gradients on monthly
temperature and precipitation
climatology
From Raper and Braithwaite, 2006
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Uncertainty in mass balance gradients leads to uncertainty in
glacier and icecap contribution to SLR in 21st C
Sensitivity of modelled glacier and ice cap SLR
contribution to ±10% uncertainty in various
reference period inputs based on Raper and
Braithwaite (2006).
20thC 21stC
SLR
SLR
Volume* ±10%
~ 0% ±4%
Mass balance ±10%
±9%
±1%
Mass balance gradients ±10% ±2% ±6%
* single estimate of area
Reference (1961-1990)
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Ice dynamic modelling is concerned with the response of the
ice geometry and hence changes in the ice exposure to climate.
Schematic of
glacier and ice-cap shrinkage
glacier colder
ice-cap
warmer
Hierarchy of ice dynamic models:
3-D models as in Schneeberger et al. (2003)
2-D models as in Oerlemans et al. (1998)
Simple geometric model as in Raper and Braithwaite (2006)
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Using coupled mass balance and ice flow models, the response of
several glaciers in different climate settings under global warming
is compared
0.02 K/a
Vavilov Ice Cap
We can
use these
results to
verify
simpler
models
applied to
all glaciers
0.02 K/a
Pioneer Ice Cap
From Oerlemans et al, 1998
From Bassford et al. 2006
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris
Summary
Table 3: Sensitivity of modelled glacier and ice cap SLR contribution to ±10% uncertainty in
various reference period inputs based on Raper and Braithwaite (2006).
Reference (1961-1990)
Volume* ±10%
Mass balance ±10%
Mass balance gradients ±10%
20thC
SLR
~ 0%
±9%
±2%
21stC SLR
±4%
±1%
±6%
* single area estimate
• uncertainties in climate forcing at glacier surface
• simplified geometric models with implicit ice dynamics
need further testing and verifying through parallel runs
with more complex models
• none of the models discussed above account for the effect
of basal sliding and ice calving
• assumed that all melt finds its way directly into the oceans.
Sarah Raper, Understanding Sea-level Rise and Variability, 6-9 June, 2006 Paris