Overlaps of AQ and climate policy global modelling

Overlaps of AQ and climate policy – global modelling perspectives David Stevenson Institute of Atmospheric and Environmental Science School of Geo. Sciences The University of Edinburgh Thanks to: Ruth Doherty (Univ. Edinburgh) Dick Derwent (rdscientific) Mike Sanderson, Colin Johnson, Bill Collins (Met Office) Frank Dentener, Peter Bergamaschi, Frank Raes (JRC Ispra) Markus Amann, Janusz Cofala, Reinhard Mechler (IIASA) NERC and the Environment Agency for funding

Material mainly from 2 current publications: The impact of air pollutant and methane emission controls on tropospheric ozone and radiative forcing: CTM calculations for the period 1990 -2030 Dentener et al (2004) Atmos. Chem. Phys. Disc. (currently open for discussion on the web) Impacts of climate change and variability on tropospheric ozone and its precursors Stevenson et al (2005) Faraday Discussions (upcoming discussion meeting at Leeds in April)

Rationale • Regional-global scale AQ legislation has implications for climate forcing – quantify these for current and possible future policies (use 2 very different models to try and reduce model uncertainty) • Climate change will influence AQ – use coupled climate-chemistry model to identify potentially important interactions

Modelling Approach • • Global chemistry-climate model: STOCHEMHad. AM 3 (also some results from TM 3+others) Three transient runs: 1990 → 2030, following different emissions/climate scenarios: 1. Current Legislation (CLE) Assumes full implementation of all current legislation 2. Maximum Feasible Reductions (MFR) Assumes full implementation of all available current emission reduction technology 3. CLE + climate change For 1 and 2, climate is unforced, and doesn’t change. For 3, climate is forced by the is 92 a scenario, and shows a global surface warming of ~1 K between 1990 and 2030.

STOCHEM-Had. AM 3 • • • Global Lagrangian chemistry-climate model Meteorology: Had. AM 3 + prescribed SSTs GCM grid: 3. 75° x 2. 5° x 19 levels CTM: 50, 000 air parcels, 1 hour timestep CTM output: 5° x 9 levels Detailed tropospheric chemistry • Interactive lightning NOx, C 5 H 8 from veg. • − CH 4 -CO-NOx-hydrocarbons (70 species) − includes S chemistry • these respond to changing climate ~3 years/day on 36 processors (SGI Altix)

Global NOx emissions SRES A 2 CLE MFR Figure 1. Projected development of IIASA anthropogenic NOx emissions by SRES world region (Tg NO 2 yr-1).

Global CO emissions SRES A 2 CLE MFR Figure 2 Projected development of IIASA anthropogenic CO emissions by SRES world region (Tg CO yr-1).

Global CH 4 emissions SRES A 2 CLE MFR Figure 3: Projected development of IIASA anthropogenic CH 4 emissions by SRES region (Tg CH 4 yr-1).

1990 2000 2030 CLE 2030 MFR Regional NOx emissions Figure 4. Regional emissions separated for sources categories in 1990, 2000, 2030 -CLE and 2030 -MFR for NOx [Tg NO 2 yr-1]

Surface O 3 (ppbv) 1990 s

CLE +2 to 4 ppbv over N. Atlantic/Pacific >+10 ppbv India A large fraction is due to ship NOx Change in surface O 3, CLE 2020 s-1990 s BAU

CLE Surface Annual Mean O 3 2020 s-1990 s TM 3 (top) and STOCHEM (bottom) Figure 13. Decadal averaged ozone volume mixing ratio differences [ppbv] comparing the 2020 s and 1990 s for (a) TM 3 CLE and STOCHEM CLE.

Surface ΔO 3 2030 CLE– 2000 (NB July) 18 Models from IPCC-ACCENT intercomparison

Up to -10 ppbv over continents Change in surface O 3, MFR 2020 s-1990 s MRF BAU

MFR Surface Annual Mean O 3 2020 s-1990 s TM 3 (top) and STOCHEM (bottom) Figure 13(b) Decadal averaged ozone volume mixing ratio differences [ppbv] comparing the 2020 s and 1990 s for TM 3 MFR and STOCHEM MFR

Surface ΔO 3 2030 MFR– 2000 (NB July) 18 Models from IPCC-ACCENT intercomparison

CH 4, CH 4 & OH trajectories 1990 -2030 CLEcc

If the world opts for MFR over CLE, net reduction in radiative forcing of 0. 2 -0. 3 W m-2 for the period 2000 -2030 Methane controls are the most effective for RF

Part 1 Summary • Co-benefits for both AQ and climate from some emissions controls • Methane offers the best opportunity (also CO and NMVOCs) • NOx controls (alone) benefit AQ, but probably worsen climate forcing (via OH and CH 4) (Similarly for SO 2) • AQ policies influence climate – this study gives a quantitative assessment • Use of many models shows results are quite consistent

ΔO 3 from climate change Warmer temperatures & higher humidities increase O 3 destruction over the oceans But also a role from increases in isoprene emissions from vegetation & changes in lightning NOx 2020 s CLEcc 2020 s CLE

Zonal mean ΔT (2020 s-1990 s)

Zonal mean H 2 O increase 2020 s 1990 s

Zonal mean change in convective updraught flux 2020 s-1990 s

C 5 H 8 change 2020 s (climate change – fixed climate)

Lightning NOx change 2020 s (climate change – fixed climate) Had. CM 3 Amazon drying More lightning in N mid-lats Less, but higher, tropical convection No overall trend in Lightning NOx emissions

Zonal mean PAN decrease 2020 s (climate change – fixed climate) Colder LS Increased PAN thermal decomposition, due to increased T

Zonal mean NOx change 2020 s (climate change – fixed climate) Less tropical convection and lightning Increased N mid-lat convection and lightning Increased PAN decomposition

Zonal mean O 3 budget changes 2020 s (climate change – fixed climate)

Zonal mean O 3 decrease 2020 s (climate change – fixed climate)

Zonal mean OH change 2020 s (climate change – fixed climate) Complex function: F(H 2 O, NOx, O 3, T, …)

Influence of climate change on O 3 – 4 IPCC ACCENT models

Part 2 Summary • Climate change will introduce feedbacks that • modify air quality These include: – More O 3 destruction from H 2 O – More stratospheric input of ozone – More isoprene emissions from vegetation – Changes in lightning NOx – Increases in sulphate from OH and H 2 O 2 – Wetland CH 4 emissions (not studied here) – Changes in stomatal uptake? (``) • These are quite poorly constrained – different models show quite a wide range of response: large uncertainties