TY - JOUR

T1 - Uncertainties in models of tropospheric ozone based on Monte Carlo analysis

T2 - Tropospheric ozone burdens, atmospheric lifetimes and surface distributions

AU - Derwent, R. G.

AU - Parrish, David

AU - Galbally, Ian

AU - Stevenson, D. S.

AU - Doherty, R. M.

AU - Naik, Vaishali

AU - Young, Paul John

N1 - This is the author’s version of a work that was accepted for publication in Atmospheric Environment. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Atmospheric Environment, 180, 2018 DOI: 10.1016/j.atmosenv.2018.02.047

PY - 2018/5

Y1 - 2018/5

N2 - Recognising that global tropospheric ozone models have many uncertain input parameters, an attempt has been made to employ Monte Carlo sampling to quantify the uncertainties in model output that arise from global tropospheric ozone precursor emissions and from ozone production and destruction in a global Lagrangian chemistry-transport model. Ninetyeight quasi-randomly Monte Carlo sampled model runs were completed and the uncertainties were quantified in tropospheric burdens and lifetimes of ozone, carbon monoxide and methane, together with the surface distribution and seasonal cycle in ozone. The results have shown a satisfactory degree of convergence and provide a first estimate of the likely uncertainties in tropospheric ozone model outputs. There are likely to be diminishing returns in carrying out many more Monte Carlo runs in order to refine further these outputs. Uncertainties due to model formulation were separately addressed using the results from 14 Atmospheric Chemistry Coupled Climate Model Intercomparison Project (ACCMIP) chemistry-climate models. The 95% confidence ranges surrounding the ACCMIP model burdens and lifetimes for ozone, carbon monoxide and methane were somewhat smaller than for the Monte Carlo estimates. This reflected the situation where the ACCMIP models used harmonised emissions data and differed only in their meteorological data and model formulations whereas a conscious effort was made to describe the uncertainties in the ozone precursor emissions and in the kinetic and photochemical data in the Monte Carlo runs. Attention was focussed on the model predictions of the ozone seasonal cycles at three marine boundary layer stations: Mace Head, Ireland, Trinidad Head, California and Cape Grim, Tasmania. Despite comprehensively addressing the uncertainties due to global emissions and ozone sources and sinks, none of the Monte Carlo runs were able to generate seasonal cycles that matched the observations at all three MBL stations. Although the observed seasonal cycles were found to fall within the confidence limits of the ACCMIP members, this was because the model seasonal cycles spanned extremely wide ranges and there was no single ACCMIP member that performed best for each station. Further work is required to examine the parameterisation of convective mixing in the models to see if this erodes the isolation of the marine boundary layer from the free troposphere and thus hides the models' real ability to reproduce ozone seasonal cycles over marine stations.

AB - Recognising that global tropospheric ozone models have many uncertain input parameters, an attempt has been made to employ Monte Carlo sampling to quantify the uncertainties in model output that arise from global tropospheric ozone precursor emissions and from ozone production and destruction in a global Lagrangian chemistry-transport model. Ninetyeight quasi-randomly Monte Carlo sampled model runs were completed and the uncertainties were quantified in tropospheric burdens and lifetimes of ozone, carbon monoxide and methane, together with the surface distribution and seasonal cycle in ozone. The results have shown a satisfactory degree of convergence and provide a first estimate of the likely uncertainties in tropospheric ozone model outputs. There are likely to be diminishing returns in carrying out many more Monte Carlo runs in order to refine further these outputs. Uncertainties due to model formulation were separately addressed using the results from 14 Atmospheric Chemistry Coupled Climate Model Intercomparison Project (ACCMIP) chemistry-climate models. The 95% confidence ranges surrounding the ACCMIP model burdens and lifetimes for ozone, carbon monoxide and methane were somewhat smaller than for the Monte Carlo estimates. This reflected the situation where the ACCMIP models used harmonised emissions data and differed only in their meteorological data and model formulations whereas a conscious effort was made to describe the uncertainties in the ozone precursor emissions and in the kinetic and photochemical data in the Monte Carlo runs. Attention was focussed on the model predictions of the ozone seasonal cycles at three marine boundary layer stations: Mace Head, Ireland, Trinidad Head, California and Cape Grim, Tasmania. Despite comprehensively addressing the uncertainties due to global emissions and ozone sources and sinks, none of the Monte Carlo runs were able to generate seasonal cycles that matched the observations at all three MBL stations. Although the observed seasonal cycles were found to fall within the confidence limits of the ACCMIP members, this was because the model seasonal cycles spanned extremely wide ranges and there was no single ACCMIP member that performed best for each station. Further work is required to examine the parameterisation of convective mixing in the models to see if this erodes the isolation of the marine boundary layer from the free troposphere and thus hides the models' real ability to reproduce ozone seasonal cycles over marine stations.

KW - Global tropospheric models

KW - Tropospheric burdens

KW - Atmospheric lifetimes

KW - Surface ozone

KW - Seasonal cycles

U2 - 10.1016/j.atmosenv.2018.02.047

DO - 10.1016/j.atmosenv.2018.02.047

M3 - Journal article

VL - 180

SP - 93

EP - 102

JO - Atmospheric Environment

JF - Atmospheric Environment

SN - 1352-2310

ER -