Discussion with NAS ITAP Committee - PowerPoint PPT Presentation

About This Presentation
Title:

Discussion with NAS ITAP Committee

Description:

Discussion with NAS ITAP Committee Arlene M. Fiore (arlene.fiore_at_noaa.gov) characterizing uncertainties in transport / photochemistry models – PowerPoint PPT presentation

Number of Views:67
Avg rating:3.0/5.0
Slides: 18
Provided by: aff89
Category:

less

Transcript and Presenter's Notes

Title: Discussion with NAS ITAP Committee


1
Discussion with NAS ITAP Committee
Arlene M. Fiore (arlene.fiore_at_noaa.gov)
characterizing uncertainties in transport /
photochemistry models
Acknowledgments D. Reidmiller, I. Bey, F.
Dentener, M. Evans, I. Held, D. Jaffe, T.
Keating, M. Schultz, R. Stouffer the TF HTAP
modeling team
August 26, 2008
2
Characterizing uncertainty IPCC AR-4 suggests
approx. equal contributions from scenarios and
models
Figure SPM.5, Summary for Policymakers
Estimates with CTMs generally include the same
uncertainties as in climate models, plus
chemistry (deposition, etc), plus ??
3
Possible considerations for the
panelCharacterizing uncertainty in CTMs
Characterizing uncertainty Need for new
approaches?
Evaluation with observations Screen out less
useful models?
Role of individual models vs. ensembles?
What information is most useful (in the policy
process)?
4
Approaches in sister communitiesGrading
models, highlighting the value of multi-model mean
Reichler and Kim, BAMS, 2008 IPCC AR-4 models
(vs. earlier intercomparisons)
Model ensemble mean
best
Index of agreement with observations
worst
5
A measure of uncertainty in SR relationships (sfc
o3) Model spread can be gt factor of 2
To generate more useful information
  • Identify processes responsible for this spread
  • (emissions/transport/chemistry)
  • Evaluate which models most accurately
  • represent those processes

begin by comparing with surface obs
6
Simulated vs. observed monthly mean surface O3
CENTRAL EUROPE
EASTERN USA
OBS (EMEP)
MODELS
MODEL ENS. MEAN
Surface Ozone (ppb)
OBS (CASTNet)
Month of 2001
Does the EUS bias influence the SR estimates in
summer?
7
EUS Bias not correlated with EU response to NA...
...can we identify observational constraints on
the model spread in SR relationships?
Work in progress
8
Utility of idealized tracer simulations to
quantify model differences due to transport (vs.
emissions and chemistry) for ozone and aerosol?
A first step Contributions to inter-model spread
in annual mean Arctic CO concentrations
Surface
250 hPa
Source region
Source region
Adapted from Figure 8 of Shindell et al.,
ACPD Role of processes isolated using both
idealized tracers full chemistry simulations
Role for other (less complex) models to estimate
uncertainty in representations of specific
processes in CTMs
9
HTAP Event Simulations Moving towards
process-based evaluation
I. Bey, M. Evans, K. Law, R. Park, E. Real, S.
Turquety
1. chemical signatures of air masses, 2.
chemical evolution in background vs. polluted
plume ensembles, 3. export efficiencies, 4.
injection heights on biomass burning plumes
Preliminary results from the French model
MOCAGE, courtesy of N. Bousserez and J.-L-
Attié, Laboratoire daérologie Toulouse, France
observations
model
clean lower trop.
biomass burning influenced air masses
middle-upper troposphere
polluted lower trop.
from Isabelle Beys presentation at DC June 2008
HTAP meeting
10
Wide model range in AQ-relevant metrics MAM
average MDA8 surface O3 over the USA (HTAP models)
Work in progress
11
VOY
WST
LAV
ROM
GRC
SND
6 CASTNet sites have been analyzed to date
potential to extend to regional average
Varying degrees of model skill in capturing
observed max daily 8-hr avg (MDA8) O3
c/o David Reidmiller, U Washington, work in
progress
12
Model Evaluation Spring (MAM)
Multi-model mean
Multi-model mean
c/o David Reidmiller, U Washington, work in
progress
13
Model Evaluation Spring (MAM)
c/o David Reidmiller, U Washington, work in
progress
14
Foreign Contribution through O3 distribution Sum
of influence from ALL 3 foreign source regions
c/o David Reidmiller, U Washington, work in
progress
15
10-model mean response of MDA8 ozone to 20
reductions of foreign emissions MAM average
Ensemble stddev in O3 decrease from -20 (EA
EU SA)
Ensemble mean O3 decrease from -20 (EA EU SA)
Work in progress
16
Possible considerations for the panel
Characterizing uncertainty in CTMs Need for new
approaches?
HTAP (SR TP1x ES simulations) transport,
emissions, chemistry Adjoint / Sensitivity
techniques ( 1 model vs. ensemble) Parameter
perturbation approaches? e.g.,
climateprediction.net, QUMP Simpler models
Evaluation with observations Screen out less
useful models?
Pros/cons of summary statistics to rapidly
communicate model evaluation Which obs best
differentiate the most useful models for our
application?
Role of individual models vs. ensembles?
What information is most useful (in the policy
process)?
17
Distinction between sensitivity and
contribution
SENSITIVITY ( TF HTAP approach)
CONTRIBUTION
What is the response of surface ozone to a 20
reduction in EA anthrop. emissions?
What ozone concentrations would exist if EA
anthrop. emissions were turned off?
Response to (20 decrease in emissions 5) lt
Response to zeroing out
Non-linearity well known e.g., Liu et al., 1987,
Lin et al., 1988, NRC, 1991
Distinction driven mainly by non-linear O3
response to NOx
? Melding of the two approaches in the ITAP
literature
Care is needed in defining question being asked
(and appropriate approach)
Write a Comment
User Comments (0)
About PowerShow.com