www.bsc.es Evaluating the global contribution from MACC when modelling an ozone episode over Spain M.T. Pay 1, V. Valverde 1, J.M. Baldasano 1,2 1 Earth Sciences Department, Barcelona Supercomputing Center, Barcelona, Spain 2 Environmental Modeling Laboratory, Technical University of Catalonia, Barcelona, Spain MACC-III Policy User Workshop, Wien, Austria, March 3-4, 2015
The problem of O 3 in Europe & Spain O 3 precursors emissions have declined (27% for NO x and 28% NMVOC) (EEA, 2013a) High O 3 (rural background), concentrations are still a problem: exceedances TV (21%), LTO (85%), IT (28%) and AT (3%) in summer 2012 (EEA, 2013b). O 3 complex pollutant: secondary (NO x /VOC), long-range transported, stratospheric, etc. O 3 is not longer a local air quality issue, it is a hemispheric and global problem Regions: Medit. area > Central & E EU > NW EU > N EU Period EU (2008/50/EC) WHO (2005) NAAQS (2008) μg m -3 μg m -3 μg m -3 O 3 1 h 180 IT - 235 8 h 120 (25d/3y), 120 LTO 100 147 (3y) 1 h 240 AT - EEA, 2013b LTO: Long-Term Objective, TV: Target Value, IT: Information Threshold, AT: Alert Threshold
Introduction: Chemical Boundary Conditions An important aspect in the regional/area applications of AQF systems is the way in which models are initialized: initial and chemical boundary conditions. The use of climatological averages is one of the common practices, but implementing boundary conditions obtained from global models is currently a significant challenge: Which are the required parameters from the global model (key pollutants, meteo parameters)? Are the temporal and spatial resolutions of the global model consistent with our application? Which vertical interpolation techniques from the global to regional/city scale? Is the forecast time of the global model enough to produce regional/city forecasts? Are there consistency between chemical mechanisms and aerosol sizes? Global products Download links MACC project forecast and analysis http://join.iek.fz-juelich.de/macc/access MOZART-4 model forecast http://www.acd.ucar.edu/acresp/forecast/ MATCH-MPIC model forecast http://cwf.mpic.de/~cwf/cgibin/match/std_figs/forecasts_std_figs.pl 3
CALIOPE Air Quality Forecasting System METEOROLOGICAL FORECAST EMISSION FORECAST HERMES-DIS 12 km x 12 km IC/BC nesting EMEP inventory Meteo. fields HERMES-BOUP 4 km x 4 km NCEP GLOBAL SIMULATIONS WRF-ARW 12 km x 12 km WRF-ARW 4 km x 4 km AIR QUALITY FORECAST Meteo. fields Emis. fields DESERT DUST FORECAST IC/BC nesting NCAR MOZART4 1.9 º x 2.5º / MACC global 1.125º x 1.125º CMAQ 12 km x 12 km CMAQ 4 km x 4 km BSC-DREAM8b 50 km x 50 km AIR QUALITY PRODUCTS
Objectives and method Objective: to evaluate the O 3 global contribution from global MACC during a summer episode in Spain Action 1: Evaluate and intercompare forecasted O 3 rates from two global model: MACC and Mozart-4 a. Validation of vertical profiles over the European domain (ozone sondes). Action 2: Evaluate the MACC chemical boundary contribution for O 3 within the CALIOPE system affecting Spain a. Implement an interpolation method from global to regional for boundary conditions. b. Track O 3 global contribution from MACC: source apportionment algorithms. c. Quantify the relative contribution from different scales (global, regional, local) to O 3 in Spain.
Temp & wind MSL P & prep Study episode: 21-31 July 2012 21-24 th = ITL 25-29 th = NWad 30-31 st = ITL ITL: Iberian Thermal Low NWad: NW advection ITL NWAd L H H L The largest O 3 episode in 2012 (24-28 th July): Total number of exceedances: 33% of the IT, 32% of the AT, and 12% of the LTO Circulation Type Classification 2012 (Freq, patterns > 22%) NWad ITL Seasonal freq (%): DJF/MAM/ JJA/SON 2.5/37.5/ 37.5/22.5 15.2/20.3/ 43.0/21.5 Mean/Max persist. Transitions 4/10 days 3/8 days ITL/ W-NWad NWad /W- NWad (Valverde et al., 2014)
Evaluation global data: MACC and MOZART-4 MACC MOZART4-GEOS5 Definition Monitoring Atmospheric Composition and Climate (MACC) Model of Ozone And Related Tracers Reference Inness et al. (2013) Emmons et al. (2010) Operational institution 7 FP of the European Union National Center for Atmospheric Research (NCAR) Chemical transport model /chemical mech. Meteorological data Emission data Vertical resolution (top)/ first layer thickness (m) Model for Ozone And Related chemical Tracers (MOZART-3, Kinnison et al., 2007; Stein et al., 2009)/ Medium-Range Weather Forecasts (ECMWF) Integrated Forecast System (IFS) Anthropogenic emission from MACCity (Granier et al., 2011) from Atmopheric Chemistry and Climate Model Intercomparison Project (ACCMIP) emission. Shipping emission from RETRO project. Biomass burning emission from Global Fire Emissions Database. Biogenic emission from MEGANv2 60 hybrid sigma-pressure levels (0.1hPa)/~10 m and 14 levels covering low troposphere (< 2km) Model for Ozone And Related chemical Tracers( MOZART-4 )/CB05 GEOS5 Argonne National Laboratory 56 σ-layers (2 hpa)/~63 m and 14 levels covering low troposphere (< 2km) Horizontal resolution 1.125ºx1.125º 1.9 º x 2.5º Temporal resolution 8 time steps 6 time steps Higher spatial and temporal resolution
Evaluation global data: MACC and MOZART-4 (forecast) Vertical profiles at WOUDC stations Ozone sondes MACC global forecasts show higher performance than MOZART-4 MACC global MOZART-4 Altitude range 0-2km >2-8km 0-2km >2-8km Mean(obs/mod, ppb) 53.59 / 54.91 97.34 / 73.74 53.59 / 57.92 97.34 / 71.52 SD (obs/mod, ppb) 3.02 / 5.45 13.28 / 3.48 3.02 / 4.66 13.28 / 10.62 Bias(ppb) 1.31-23.6 4.12-25.82 NMB (%) 2.45-24.25 7.66-26.52 MGE (ppb) 2.7 23.6 7.37 25.82 NMGE (%) 5.04 24.25 13.7 26.52 RMSE(ppb) 4.43 25.42 7.8 26.3 r 0.58 0.94-0.6 0.92
Interpolation method from global MACC to regional CALIOPE 15 layers 60 layers (top 50 hpa) (top 0.1 hpa) 35/60 15/15 Zoom < 500 m a.s.l Nearest-neighbor interpolation : Interpolation between the two nearest layers Proxy: the High of the mean layers (global and regional) (Preasure and T). Layer 4 Layer 3 Layer 2 Layer 1 CMAQ MACC (CB05, AERO5) O3 vmr_o3 NO vmr_no NO2 vmr_no2 CO vmr_co PAN vmr_pan OH vmr_oh FORM vmr_ch2o ISOP vmr_isop SO2 vmr_so2 ETHA vmr_c2h6 ETH N/A ASO4I vmr_sulfaer (50%) ASO4J vmr_sulfaer (50%) HNO3 vmr_hno3 H2O2 N/A
Track O 3 global contribution from MACC: source apportionment algorithms Apply Source Apportionment in CALIOPE (CMAQ-ISAM): 1. Check the performance of the MACC boundary condition 2. Assess the origin of O 3 exceedances in Spain from different scales (global, regional, country) Tool: Source Apportionment CMAQ-ISAM CMAQ-ISAM: Integrated Source Apportionment Method. Evaluated by means brute force zero-out scenarios (Kwok et al., 2013, 2014) Hybrid approach for O 3 : using the PH 2 O 2 /PHNO 3 indicator ratio. Tracers for NO x (9) and VOC families (19) are based on CB05 species contributing to O 3 formation. BCON: MACC global (global O 3 transport out from EU12) EURO: countries surrounding Spain* ESP: Spain* OTHR: Ocean ICON: initial conditions ISAM-EU12 (Europe, 12 km x 12 km) Spain* (Spanish IP + Balearic Islands)
Region contribution to O 3 in Spain (EU12): Mean concentration 21-31 July 2012 TOTAL O 3 O 3 from Spain O 3 from Europe O 3 from shipping O 3 from GLOBAL MACC
Region contribution to O 3 in Spain hourly contribution at Niembro station Under ITL, BCON dominates. Under the NWAd: EURO shows the highest contribution: 40 µg/m 3 mean, 60 µm/m 3 during the peak Persistent shipping contribution (10-20 µm/m 3 ) Low ESP contribution (< 20 µg/m 3 ) LTO O 3 Iberian Thermal Low North Western Advection Iberian Thermal Low
Region contribution to O 3 in Spain at rural background sites LTO LTO EURO controls EURO controls LTO LTO Portugal controls ESP controls LTO LTO ESP controls Portugal + ESP + Shipping controls
Summary of SA: region contribution to O 3 in Spain Regions contribution to O 3 in Spain (RB stations) 13% OCEAN Shipping 20% ESP 10% SNAP7 + 6% OTHR + 2% SNAP1 + 2% SNAP34 64% non-esp 20% EURO + 44% global MACC 1% ICON
Conclusions and future work MACC global forecasts is a valuable product to feed regional forecast simulations: Enough time at sufficient spatial resolution to produce a real-time forecast. Large number of chemical components. High performance in terms of vertical profiles in Europe. ISAM source apportionment is useful as a diagnostic tool: Check chemical boundary conditions. Quantify the contribution at different scales and sources. O 3 overestimation in the eastern domain of the CALIOPE system are linked to how MACC boundary conditions are implemented in the boundaries. Next steps: Interpolation methods from global to regional: test Weighted average instead Nearest-neighbor interpolation (current) Increase the vertical resolution of the regional model (first layer from 40 to ~20 m). Assessment of the origin of O 3 exceedances in Spain in a typical summer episode: Not only national (~20%), but continental (~20%), hemispheric and global (~44%). NWad brings O 3 and precursors from N EU countries and international shipping inland Spain up to 60 µg/m 3 (~40-50 µg/m 3 ) in N Spain; meanwhile central and costal areas are more affected by recirculation of local precursors.
Thank you for your attention References: EEA, 2013a. Air quality in Europe 2013 report. European Environmental Agency, EEA Report, No 9/2013. ISSN 1725-9177. 112 pp. EEA, 2013b. Air pollution by ozone across Europe during summer 2012. Overview of exceedances of EC ozone threshold values for April-September 2012. European Environmental Agency, EEA Report, No 3/2013. ISSN 1725-2237. 52 pp. Kwok, R., Baker, K., Napelenok, S., Tonnesen, G., 2014. Photochemical grid model implementation of VOC, NOx, and O 3 source apportionment. Geoscientific Model Development Discussions 7, 5791-5829. Kwok, R., Napelenok, S., Baker, K., 2013. Implementation and evaluation of PM2.5 source contribution analysis in a photochemical model. Atmospheric Environment 80, 398-407. Valverde, V., Pay, M.T., Baldasano, J.M., 2014. Circulation-type classification derived on a climatic basis to study air quality dynamics over the Iberian Peninsula. International Journal of Climatology, doi: 10.1002/joc.4179 Webs: Air Quality Forecasts Europe / Spain: http://www.bsc.es/caliope Funding: 16