The lag-1 autocorrelation of the annual time series shows the degree of predictability from one year...
-
Upload
edwin-wiggins -
Category
Documents
-
view
215 -
download
0
Transcript of The lag-1 autocorrelation of the annual time series shows the degree of predictability from one year...
The lag-1 autocorrelation of the annual time series shows the degree of predictability from one year to the next. It shows the
importance of the long-term climate modes, and is also important for quantifying the uncertainty in time-series quantities. Red = AE,
Blue = UE, black = control, black-dash = UE10x(fixedCH 4), thin colored lines are Monte Carlo realizations with control statistics, to
indicate confidence. The three methane runs show decreased predictability in southern ocean.
The Impact of Methane Clathrate Emissions on the Earth System
Warming may release methane from large Arctic reservoirs
Location of emission has some effect on climate
Methane emissions may change variability
Methane emissions change interannual autocorrelation
We acknowledge the contribution of our collaborators, who have primary responsibility for the ocean model (S.Elliott & M.Maltrud, LANL) and sediment model (M.Reagan & G. Moridis, LBNL). We also acknowledge J.Stolaroff (LLNL) for including us in a review for technical mitigation measures for methane.
This research used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science (BER) of the U.S. Dept. of Energy under Contract No. DE-AC02-05CH11231
Philip Cameron-Smith1, Subarna Bhattacharyya1, Daniel Bergmann1, Matthew Reagan2, Scott Elliott3 and George Moridis2
1 Lawrence Livermore National Laboratory, 2 Lawrence Berkeley National Laboratory, 3Los Alamos National Laboratory
Future plans• Couple our atmospheric methane model to ocean and
permafrost methane codes.• Determine emission amplification factors, ie the amount of
extra methane released due to the warming from the original methane emission (cross-amplification between reservoirs will also occur).
• Assess the likelihood of runaway warming (aka, the clathrate gun).
• Participate in future methane related model intercomparisons (we participated in the Atmospheric Chemistry-Climate Model Intercomparison Project (ACCMIP), which helped confirm and debug our model chemical behavior, and resulted in several papers).
Acknowledgements
Clathrates are methane locked in a water ice structure.ESAS = East Siberian Arctic Shelf (methane under submerged
permafrost)
SPARC Meeting, Queenstown, 2014
Expected methane sources due to warming
Onset of Clathrate emissions expected to be abrupt
Fraction of methane that passes through ocean is uncertain, but could be large
Stolaroff, et al., 2012
Reagan, et al., 2011
This is the sediment response to a temperature ramp of 5K for 100 years.
This emission rate is ~20% of global methane emissions.Most of the emission comes from 300-400m depth.
Methane is consumed in the ocean by methanotrophs.But, bubble plumes may inject methane to upper ocean levels which
will allow faster release to the atmosphere.
50 m depth
150 m depth
Sea-floor
% m
etha
ne re
leas
ed to
atm
osph
ere
(no
met
hano
trop
hs)
Months since start of simulation
Methane conc. increases non-uniformly
Note that methane suppresses the specie that destroys it, so a 3x increase in total methane emissions (10x clathrates) produces 6-8x
methane concentration increase over our control (present-day).
We added a fast chemical mechanism to CESM that is designed to handle methane . We simulated over 600 years (after spinup) under present-day conditions with 1x and10x expected clathrate emissions (AE), and the same amount of extra emission spread uniformly over
the globe (UE). We used an active ocean to see the coupled chemistry-climate response.
Temperature increase is modestly dependent onEmission location
LEFT: Zonal mean surface temperature increase over control (AE=Red, UE=Blue, UE10x(fixedCH4)=Green, 2xCO2=Cyan, 4xCO2CMIP=Magenta). RIGHT: same data, but expressed as %difference for AE over UE for different regions (GL=globe,
SP=south pole, LL=low latitudes, NP=north pole). The error bars are the uncertainty in the mean. The UE simulation is slightly warmer
than the AE scenario for almost all latitudes except the Arctic. The modest difference in the Arctic seems to be a compensation
between increased warming from extra methane in AE, with greater poleward heat transport in UE.
Ozone increases most in polluted regions
Annual-mean surface ozone increase due to clathrate emission (vmr)
Methane is an important precursor for chemical smog, and the 1x and 10x clathrate emissions increase ozone concentrations by
~5ppb and ~35ppb in polluted regions, respectively. The ozone increase is likely to be even greater during pollution events.
We developed a chemistry-climate version of CESM There appears to be difference the interannual variability between AE, UE, and control
The increased methane variability in AE over UE (from the effect of synoptic weather patterns on the clathrate plumes) seems to affect the PDFs of the annual temperature. The effect on the Arctic isn’t
as noticeable because it is already a highly variable region.Note that the mean, std. dev., and skewness all seem to be
affected.
Interannual variability in zonal-mean temperature changes
Variability decreases for El Nino and Sea-ice
The spatial pattern of the change in the standard deviation of the annual mean surface temperatures looks like it is mostly a function
of the radiative forcing in the scenario.
The year-to-year auto correlation in southern ocean seems to be reduced
Elliott, et al., 2010, 2011
10x scenario
s
10x scena
rio
10x scena
rio
Arctic
of a
nnua
l tem
pera
ture
s
Annual temperature (K)
10x scenario
s
Global
There seem to be changes roughly proportional to the level of global radiative forcing at several latitudes (indicated by arrows). In many
cases the decreases are clearly related to sea-ice. In the zonal means, the difference between AE and UE is hard to distinguish
from the uncertainty (1 sigma ~ 0.04). Note: the various 4xCO 2 runs are very short, so their variability is much more uncertain.
Lag-
1 au
toco
rrela
tion
coef
ficie
nt
Lattitude
10x scenari
os
10x scenari
os
AE10x UE10x UE10x(fixedCH4)
4xCO2CMIP2xCO2
Ratio
of s
tand
ard
devia
tions
of
zona
l-mea
n su
rface
tem
pera
ture
s to
con
trol
Latitude
AE10x
UE10x
UE10x(fixe
d)
The dots are the variability in the means of each time window of the specified length within the 400 year simulations (with different
starting years when possible). The circles are the standard error formula without any autocorrelation correction. The difference is
due to auto-correlation.
Auto-correlation important for uncertainty
Red Dots: Arctic EmissionBlack Dots: Control
Length of time-window (years)
Unce
rtain
ty in
sur
face
te
mpe
ratu
re (K
)
https://www.google.ca/search?q=sparc+poster+pjc+2014+methane+hydrate&oq=sparc+poster+pjc+2014+methane+hydrate&aqs=chrome..69i57.17768j0j8&sourceid=chrome&es_sm=93&ie=UTF-8#q=The+Impact+of+Methane+Clathrate+Emissions++Los+Alamos+National+Laboratory
June 2014
More than 900 organinizations in over 200 countries
Climate Action Network calls for phasing out all fossil fuel emissions and phasing in a 100% renewable energy future with sustainable energy access for all, as early as possible, but not later than 2050.
CAN believes that global average surface temperature warming should be limited to less than 1.5°CAbove pre-industrial levels by 2100.
Some present and projected future impacts, such as those on food security, sea level rise or ocean acidification, are occurring with more intensity than previously anticipated. These impacts will be disruptive for all countries; especially for the global poor and vulnerable peoples.