1. Effect of increasing Asian Emissions and meteorological variability on the composition of the UT/LS J. E. Williams 1, P. F. J. van Velthoven 1, T. J. Schuck 2 and C. A. M. Brenninmeijer 2 [1] Chemistry & Climate, KNMI, The Netherlands [2] Atms. Chem. Div., Max Planck Institute for Chemistry, Germany

2. 2 Motivation ● Rapid economic growth in India, South East Asia and China has introduced the largest perturbation in the Anthropogenic emission component over the last decade which will influence tropical tropospheric composition. ● Strong convective activity should rapidly distribute the effects into the free troposphere and upper troposphere. ● What is the effect on UT composition. ● Can we examine the seasonality in such a perturbation using CARIBIC measurements and attribute a fraction due these increasing Asian Anthropogenic emissions??

3. 3 CARIBIC Flights into the REAS region ____ 2008 CARIBIC flights (India) ____ 2006-2008 CARIBIC flights (China/Philippines) REAS – Regional Emission inventory for Asia 16 flights above Middle East/Pakistan/India during 2008 Flights occur from April -> Dec covering the Indian Monsoon period Flight routes cross India and Pakistan, with some in the outflow regions

4. 4 CTM model simulations (TM5) ● 3D global CTM TM5-chem-v3.0 (Huijnen et al., GMD, 2010) ● 3° x 2° with 34 vertical levels (1km resolution in the UTLS) ● Driven by the ERA-interim meterological re-analysis ● Modified CBM4 chemistry : NOx-HOx-CH4-CO-VOC-SOx ● Anthro : RETRO/REAS hybrid accounting for increasing emissions in the REAS region (Ohara et al., ACP., 2007; Schultz et al, 2008) ● Biomass burning: GFEDv2 monthly global fire emission inventory. ● Biogenic : ORCHIDEE 12-year annual average ● Lightning: related to convective precipitation (Meijer et al, 2001 scaled to 5 TgN yr-1) Three runs: Continually increasing Asian emissions (BASE) Asian anthro. emissions fixed at yr 2000 values (FXANTH) Fixed WV, T, P, transport, precip etc for yr 2000 (FXWVTP) Stratospheric O 3 is constrained using the Multi-Sensor re-analysis data (van der A et al., 2010) 90hPa outside tropics/45hPa in tropics

5. 5 Surface Differences in REAS region CO and O 3 : FXANTH BASE (FXANTH-BASE/BASE)*100 Biomass Burning

6. 6 Seasonal Cycle in Indian Emissions REAS-HYBRID Integrated between 0-36°N, 60-90°E Seasonal cycle identical in both simulations ; Inter-annual variability caused by BB Minimum during January and Maximum during June ~10% contributions from anth. growth towards total NO x and CO emissions ~ 7% contributions towards NMVOC Growth rate for NO x (+10%) higher than for CO (+3%) Full emission scenario Fixed 2000 Asian emission

7. 7 FXANTH: Indian Summer Monsoon O 3 and CO: 2008 Zonal Means (60-90°E) Difference = (FXANTH-BASE/BASE) Strong convective uplift of CO from boundary layer to UT during Indian Monsoon Source region LS inc. convective uplift Less Inflow

8. 8 UT wind direction during CARIBIC days (2008)

9. 9 Seasonal cycle due to convection during Monsoon CH4 mixing ratios along CARIBIC flight track during 2008 (Schuck et al, ACP, 2010.) Seasonal and latitudinal cycle evident. CO can act as a proxy for strength of convective uplift from boundary layer Seasonal cycle in Chinese CO emissions in REAS inventory peak during summer when transported eastward NOAA flask Surface trend wind

10. 10 Correlation: REAS-HYBRID emission inventory vs CARIBIC CO Surface emission integrated between 60-90°N High Correlation between CO surface emissions and UT CO between 25-30°N. Low correlation towards the tropics 20-15°N more affected by transport west Caribic measurements ( ___ ), Anthro + Biogenic + BB (----)

11. 11 TM5 comparisons against CARIBIC flights I Measurements/model output binned on 5° bins. Forward/backward flight averaged together. TM5 captures shape of the observed latitudinal variation in UTLS CO over India (except for Sept.) Strength of Convective uplift and easterly transport is variable from year to year (using fixed surface emissions). Implies CO emission maybe too high in REAS from China. Middle-East O 3 maximum captured well by TM5 although less steep gradient (except for July). Middle-East O3 maximum Online interpolation of model fields (Brunner et al, ACP, 2003) Over- estimated Better Agreement Tagged CO Experiment needed !!

12. 12 Influence of Stratospheric O 3 on Lat. profile Tagged stratospheric O 3 tracer used to diagnose the seasonal variability of the influence. For the tropics (30°N-30°S) nudging towards the Multi-Sensor re-analysis monthly mean values is performed at pressure < 45hPa. Shows that increases due to downwelling of stratospheric O 3 is NOT the source of the over-estimation of the Middle-East O 3 maximum. Cruise altitude of ~250hPa

13. 13 Differences in CO & O 3 at flight level for FXANTH (250hPa) Diffs = (FXANTH-BASE/BASE) Reductions in [O 3 ] above China are transported West over India Asian Anthro. Emission growth since 2000 results in ~2-7% CO and O 3 in tropical UTLS over region 7

14. 14 TM5 Comparisons against CARIBIC flights II :seasonal cycle Online interpolation of model fields CARIBIC OBS BASE FXANTH FXWVTP Measurements/model output binned on 5° bins. Forward/backward flight averaged together. TM5 captures seasonal variability in CO well 25- 30°N. Fails to capture that observed for O 3 in tropics along flight track. ● Missing CO April-May across all latititudes FXWVTP CO comparisons show that with 2000 meteorology the convective cycle during the Monsoon period is stronger between 25-30°N and similar 15-20°N Stratospheric contribution to total model [O 3 ] ~5% Perturbations in UT O 3 prod. due to increasing asian emissions < 1%. Most changes in boundary layer then lofted upwards.

15. 15 Differences in NO x at flight level for FXANTH (250hPa) NOy = [HNO 3 ] + [PAN] + [NO x ] TM5 overstimates UTLS NOy 15-25°N Weak correlation during Sept Largest decreases in NO x occur above China (~6-7%). Modest differences over India (~1-3%).

16. 16 Conclusions ● TM5 is able to capture both the latitudinal and seasonal gradients in UT CO over India (2008). For O 3 results the quality is more variable. ● The middle East O 3 maximum is missed in July but captured well during August and September. ● Comparisons of latititudinally binned CO emissions from REAS correlate well with UT CO (30-25°N). For 20-15°N transport into the region from China seems dominant. ● Capturing the UTLS composition around 15°N with TM5 is difficult: either too strong chinese emissions or transport into the region ? ● The growth in Asian anthropogenic emissions since 2000 increases UT O 3 and CO between ~2-7% over India during the summer Monsoon indicating a significant contribution to UTLS. With continally increasing Asian emissions this effect is likely to become more pronounced influencing UT chemistry. Ongoing: Simulations using REAS inventory estimates for 1980 Future: Tagging of Chinese Anthropogenic emissions to quantify inflow