Daily Tropospheric Ozone Residual from OMI-MLS M. Schoeberl, J. Ziemke, P. K. Bhartia, B. Bojkov, A. Douglass, B. Duncan & GMI Team & Ozonesonde groups NASA/GSFC L. Froidevaux, N. Livesey, J. Waters JPL P. Levelt KNMI, Netherlands

Tropospheric Ozone Residual (TOR) MLS Stratospheric O3 Column Map made from MLS measurement points shown How hard can it be to subtract two numbers?

Increasing the Resolution of the Stratospheric Column with Forward trajectories from six previous days of MLS V1.5 data are used to increase the stratospheric column resolution OMI L2G with measurement time stamps are used to correct for asynoptic affects Forward trajectories are grouped in time for target day (1/10th day resolution) Tropopause is time interpolated from 6 hour DAS fields ----------- Product has same zonal mean as asynoptic TOR Fold TOR anomalies are better handled June 24 2005 MLS Ascending node measurements Forward Trajectory Points

Time Synchronized Product Tropospheric Column Not Time Synchronized Time Synchronized

Column Ozone vs Average Column Mixing Ratio Sept-Aug‘05 Trop. Column Ozone and Average Mixing Ratio DU ppbv (HTOR) The tropospheric equivalent mixing ratio tends to reduce the topographic effects.

Seasonal Tropospheric O3 Mixing Ratio Dec. - Feb. March - May June - Aug. Sept. - Nov.

Validation Sonde Comparisons TES Comparisons Ziemke product Comparisons Most comparisons are made using the surface-200hPa product from sondes and TES - this avoids issues with different definitions of the tropopause. Dates: October 2004 to December 31, 2006 Data is interpolated to location of sonde or TES measurement. Sonde data base >6000 comparisons from Envisat, Shadoz, Canadian, IONS,etc.

2.4 DU offset (sondes larger) Tropical Sondes 690 sondes 2.4 DU offset (sondes larger)

3.4 DU offset (sondes larger) Mid-latitude Sondes June -July- August 688 sondes 3.4 DU offset (sondes larger) Wider pdf

2.9 DU offset (sondes larger) Mid-latitude Sondes March-April-May 462 sondes 2.9 DU offset (sondes larger) Even wider pdf

TES Comparison Tropics 4.1 DU offset (TES larger) For sondes the ST DEV was 5.4 DU

TES Comparison at Mid-Latitudes 3.4 DU offset (TES larger) For sondes the ST DEV was 8.7 DU

Validation Summary

For sondes the ST DEV was 5.4 DU Z06 Comparison Tropics 2.3 DU offset For sondes the ST DEV was 5.4 DU

For tropics the ST DEV was 4.2 DU Z06 Comparison April -7.7 DU offset For tropics the ST DEV was 4.2 DU

Comparison to GMI Combo Jan 6, 2005

Seasonal Variations (2005) Extra-tropical (tropical) region is defined as region where the tropopause is below (above) 12 km. The mass is normalized to the annual averaged area occupied by the respective regions.

Zonal Mean Comparison with GMI Overall good agreement Arrows indicate some differences: GMI SH and NH mid-latitude maxima starts sooner OMI-MLS values too low at high latitudes Winter tropical values are lower than GMI

PDF Comparison with GMI No longer surf -200 hPa column

Summary Forward Trajectory transport of MLS data is used to create a high horizontal resolution stratospheric ozone column which is then used to produce an improved tropospheric ozone residual (TOR). Fold events appear to dominate TOR day-to-day extra-tropics, but using better time synchronization, fold event amplitudes are greatly reduced. Validation with sondes shows good tropical comparisons, but more variability in TOR at mid-latitudes than sondes TES comparisons show similar results Both TES and Sondes show a systematic offset with OMI-MLS being a few DU low Comparisons of total tropospheric ozone amount to GMI Combo model show overall good agreement Differences in extra-tropics may be due to more vigorous circulation in GMI Differences in tropics due to different emissions and convection

Questions Why is there more variability in the TOR than in the GMI or sondes or TES? Clue: variability is associated with folds - or strong vertical changes Clue: both enhancements and depletions are larger Clue: independent of tropopause definition Clue: TOR is systematically low - this bias is larger at mid-latitudes One explanation that fits all these data is that the ozone amount at 215 hPa is too large or too variable - stratospheric ozone variability is miss-assigned to the troposphere Consistent with data presented by Froidevaux et al. [2006], Livesey et al. [2007] More variation in TOR Less variation in TOR X - > X - >

Now I will make some rambling semi-coherent remarks… V2.2 vs V1.5 GEOS Chem vs MLS Livesey et al. [2007]