MIMOSA
MIMOSA-CHIM
General
MIMOSA-CHIM, combines the PV advection scheme described here, and the chemistry scheme of the three-dimensional Chemical Transport Model (CTM) REPROBUS (Reactive Processes Ruling the Ozone Budget in the Stratosphere).
Chemical scheme description:
The chemical fields of MIMOSA-CHIM are initialized by the REPROBUS output fields interpolated on isentropic levels and are advected along with PV with a time step of one hour. For several months-long simulations, the diabatic transport of air across isentropic surfaces, and the diabatic evolution of PV have to be taken into account. Diabatic mass fluxes are computed from the heating rates calculated using the radiation scheme of the SLIMCAT model taken from MIDRAD. Climatological water vapour, CO2, and interactive ozone fields (taken from the model itself) are used for the calculation of heating rates.
The JPL version of MIMOSA-CHIM has a horizontal resolution is 1º in latitude and longitude. Twenty one vertical levels are used (from 350 K and 950 K with best resolution of 5 K at levels near 430 K, yielding a vertical resolution of about 2 km). The model domain in this chemistry version is not global and extends to 10 degrees of latitude across the equator, i.e., the latitude range of the simulation for the Northern Hemisphere is 10°S – 90°N. To take into account the influence of air masses originating from regions outside of the model domain, PV fields of ECMWF and chemical fields of REPROBUS are used for the forcing at the boundary of the model. Tests have shown that the effect of forcing fields on the ozone loss inside the vortex remains small, either using REPROBUS data or climatological fields.
The REPROBUS chemical scheme includes 55 chemical species and about 160 gas phase, heterogeneous, and photolytic reactions. The updated photochemical data of JPL-2002 recommendations and a detailed scheme of liquid and solid PSC formation and growth including the existence of NAT and water-ice particles are used. A denitrification scheme accounts for the sedimentation of HNO3-containing particles where the NAT particles are assumed to be in equilibrium with gas-phase nitric acid.
Case Study
Around March 13, 2005, the PV-only forecast version of MIMOSA running in real-time at JPL triggered a so-called “polar filament alert” at Mauna Loa Observatory (MLO), Hawaii. Such an alert is triggered each time an air mass with characteristics of the polar vortex is forecasted to pass over Table Mountain Facility (California), or Mauna Loa, the locations of the two JPL lidars. Though alerts are not uncommon for Table Mountain, they are extremely rare for Mauna Loa, because of the nearly-tropical location of the Hawaiian site. This alert, initially triggered from the 72-hour forecast, was confirmed on every day’s forecast until the event actually occurred, on March 16-19, 2005. The southward stretching of the filament coincided in time with the onset of the so-called northern hemisphere spring “Final Warming”.
Ozone mixing ratio output from MIMOSA-CHIM on 16 March at 0000 UT on the 435 K isentropic surface are shown in Figure 1 together with the corresponding Aura-Microwave Limb Sounder (MLS) version 1.5 ozone mixing ratios measured along the instrument’s suborbital track. A time coincidence of +/- 1 hour was chosen here, which corresponds to the very first MLS suborbital track of 16 March. The contour color scale (MIMOSA-CHIM ozone) and the symbols color scale (MLS ozone) are identical. Best agreement is reached when symbols cannot be distinguished from the contours. Despite a few areas of disagreement, MIMOSA-CHIM and MLS ozone agree remarkably well. In particular, the polar filament produced by MIMOSA-CHIM stretching from Alaska to Hawaii is well captured by MLS. Quantitatively, mixing ratio values of about 2 ppmv in the center of the filament near Hawaii are found for both MLS and the model. Values measured by MLS farther north of Hawaii (i.e., as the suborbital track intersects the filament near 37ºN) are slightly below that of the model.
Watch the animated GIF (10 MB) showing the development of the filament and corresponding Aura-MLS (v1.5) measurements by clicking here.
Lidar measurements were obtained on the nights of March 16, 17 and 18, 2005. The ozone mixing ratio profiles measured by lidar on March 16 every hour between 0600 UT and 0900 UT are plotted in Figure 2a (below-left).
Each profile corresponds to a 1-hour integration measurement centered at the time indicated on the figure. On these profiles, the filament is characterized by a layer of enhanced ozone reaching about 1.6 ppmv, as compared to typical climatological values of 1.0 ppmv (as measured, for example, on March 9 and March 22). In figure 2a, the local ozone peak is reached at 435 K (18.5 km), but this level actually varied from 425 K to 450 K (17 to 20 km) over the few days between the beginning and the end of the event (not shown). Even over the course of one night (figure 2a), the lidar detected the filament at varying isentropic levels (between 425 K and 435 K). The model ozone profiles above MLO on March 16 between 0600 and 0900 UT are shown in Figure 2b (above right). The ozone filament centered near 435 K was actually approaching MLO during the first hours of March 16, reached MLO at around 0600 UT, and remained above MLO until after 1200 UT. As the filament approaches Hawaii, it can be seen on both the model and lidar profiles as an emerging ozone peak throughout the night.
Figure 3 presents daily global maps of chlorine monoxide (ClO), at 435 K and 1800 UT (day time ClO), between March 13 and March 16 from MIMOSA-CHIM. They show the evolution of ClO in the air parcels within the filament during its development and “ejection” out of the vortex. From the maps it is clear that the air mass in the filament that left the polar vortex on March 13 was chemically active containing significant amount of ClO. Also the map on March 16, when the filament was already in tropical latitudes, shows that the chemically active air mass is transported all the way to the tropics, and chlorine deactivated as the filament passes over MLO.
A cluster of nine forward and nine backward isentropic trajectories, starting at each grid point within +/-1º of longitude and latitude from the location of MLO, and at the mean time of the March 16 lidar measurement, were calculated. The four-times daily ECMWF isentropic wind fields were used for trajectory calculation. The trajectories were calculated over 15 days in each direction with a time step of 1 hour, and outputs every 6 hours to match the MIMOSA-CHIM output times, allowing the study the chemical history of the air parcels over an entire month. The isentropic trajectories are a good approximation of the actual three-dimensional trajectories, as the parcels’ potential temperature changes do not exceed +/- 5 K over the course of 15 days. Inspection of the trajectories at multiple levels of 5 K intervals showed that, when considering neighboring levels, there was very little change in the position of the parcels and in the evolution of their composition as modeled by MIMOSA-CHIM.
The top panel of figure 4 shows the evolution of passive ozone and chemically processed ozone, as modeled by MIMOSA-CHIM along the four trajectories best centered to the “core” of the filament during its MLO overpass. The bottom panel shows the corresponding evolution of ClO and temperature.
From March 1 to March 12-13 the air parcels remain inside the polar vortex. Temperature and ClO are anti-correlated. On March 1, reactive ozone was about 2.5 ppmv and passive ozone about 4.8 ppmv at 435 K. This difference is due to chemical ozone depletion involving halogen-dominated chemistry inside the polar vortex. The winter 2004/2005 experienced large chlorine activation and large ozone loss. Low activated ClO during March 4 and March 5 (bottom panel) is due to a temporary warming of the air parcels as they traveled through a region of steep horizontal temperature gradient. Between March 9 and March 13, there is no heterogeneous production of ClO, and ClO is deactivated for ozone depletion. Until March 11, the air parcels were well inside the polar vortex, but on March 12, 1200 UT a significant fraction of the vortex air mass separated from the core. Within 24 hours the filament developed. The ClO maps indicate that the air mass in the filament on March 13 was chemically well activated and separated from the rest of the vortex. During this period reactive ozone along trajectories decreased from 2.5 ppmv to about 2 ppmv. During the forty eight hours from March 13 to March 15, the air parcels along trajectories went through a change in their ozone and ClO content. During the parcels’ southward travel, the filament remains connected to the main vortex. Early in this period, chlorine monoxide concentration drops to 0.45 ppb on March 13. From March 13 to 15 the ClO concentration drops to 0.2 ppbv and this chlorine deactivation is inside the filament. Passive ozone also decreases to about 3.5 to 4 ppmv. From March 15 on, the difference in concentration between passive and active ozone decreases sharply. The filament is in the mid-latitudes and subtropics. Very fast diffusion and mixing along with halogen processing takes place, reducing ClO to less than 0.05 ppbv. Around March 29, the ozone and passive ozone concentrations reach values typical of the tropical background.
The detailed above history of the modeled chemistry inside the filament suggests that the air mass was still polar ozone-depleted when passing over Hawaii.
References
Tripathi, O. P., T. Leblanc, I. S. McDermid, F. Lefèvre, M. Marchand, and A. Hauchecorne (2006), Forecast, measurement, and modeling of an unprecedented polar ozone filament event over Mauna Loa Observatory, Hawaii, J. Geophys. Res., 111, D20308, doi:10.1029/2006JD007177.
Leblanc, T., O. P. Tripathi, I. S. McDermid, L. Froidevaux, N. J. Livesey, W. G. Read, and J. W. Waters (2006), Simultaneous lidar and EOS MLS measurements, and modeling, of a rare polar ozone filament event over Mauna Loa Observatory, Hawaii, Geophys. Res. Lett., 33, L16801, doi:10.1029/2006GL026257.
Lefèvre, F., G.P. Brasseur, I. Folkins, A.K. Smith, and P. Simon (1994), Chemistry of the 1991/1992 stratospheric winter: Three dimensional model simulation, J. Geophys. Res., 99, 8183-8195.
Hauchecorne, A., S. Godin, M. Marchand, B. Heese, and C. Souprayen (2002), Quantification of the transport of chemical constituents from the polar vortex to midlatitudes in the lower stratosphere using the high-resolution advection model MIMOSA and effective diffusivity, J. Geophys. Res., 107 (D20), 8289, doi: 10.1029/2001JD000491.
Chipperfield, M.P. (1999), Multiannual simulations with a three-dimensional chemical transport model, J. Geophys. Res., 104(D1), doi: 10.1029/98JD02597, issn: 0148-0227.
Shine, K. P. (1987), The middle atmosphere in the absence of dynamical heat fluxes, Q. J. R. Meteorol. Soc., 113, 603–633.
Marchand M, S. Godin, A. Hauchecorne, F. Lefèvre, S. Bekki, M.P. Chipperfield, Influence of polar ozone loss on Northern mid-latitude regions estimated by a high resolution chemistry transport model during 1999-2000, J. Geophys. Res., 108, 10.1029/2001JD000906, 2003.