Ground-based millimeter-wave measurements of mid-stratospheric Arctic ozone have been performed at Ny-Ålesund, Spitsbergen (79°N, 11°E). The ozone measurements during March 1995 show periods with very high day-to-day variability (Fig. 1,left). This variability is almost absent during March 1997 (Fig. 1, right). The objective of this work is to show that the observed ozone variability can be explained well with usual gas-phase photochemistry. It is well known, that photochemical life times of ozone at altitudes around 30 km are on the order of some days. The observed ozone mixing ratio is therefore mainly determined by the photochemical loss and production during the last couple of days, depending on the solar zenith angles (SZA) the air parcel experienced. The atmospheric dynamics thus controls the photochemical loss and production of ozone. To test if the observed ozone concentrations can be explained by this ‘dynamically controlled photochemistry’, a photochemical box model was run along backward trajectories, initialized with standard ozone and trace gas mixing ratios.
The Radiometer for Atmospheric Measurements (RAM) is operated continuously since winter 1994/95 at Ny-Ålesund. It detects an ozone line at 142 GHz. One to five ozone profiles per hour are retrieved between 12 and 55 km with a resolution of about 10 km. Around 30 km the precision is about 0.1 ppm with an accuracy of about 0.2 ppm. For more details on the RAM and measurements of ozone during the Winter 1996/97 see also [5]. Bremen's Atmospheric Photochemical model (BRAPHO) was developed to simulate photochemical processes in the stratosphere. It includes a highly accurate calculation of photolysis rates using the model PhotoGT [1] which is based on the radiative transfer model GOMETRAN [7]. Especially the calculation of photolysis rates at high SZAs (pseudo-spherical approximation including refraction) and the high spectral resolution treatment of the O2 Schumann-Runge bands [6] enables BRAPHO to simulate the photochemistry, especially at twilight, with high accuracy. BRAPHO uses bi- and trimolecular rate coefficients taken from [3]. Heterogeneous reaction rates on liquid and frozen aerosols are calculated [2]. However, the model runs presented here do not include heterogeneous reactions. To solve the set of differential equations and to handle the databases the ASAD package (Carver et al., 1997) is used. For more detailed information about BRAPHO see also [4].
For mid of February to end of March 1995 and mid of February to end of April 1997, the photochemical model BRAPHO was run along 15-day backward trajectories at the 1000 K isentropic level ( 33 km). The ozone mixing ratio was initialized with 5.5 ppm, all other trace gas mixing ratios were taken from a 2D-model output. All trajectories (1995 and 1997 ) were initialized identically. As an example, the trajectories for March 10 and March 14, 1995 are shown in Fig. 2: The trajectory for March 10 moved to lower latitudes, resulting in lower SZAs. This led to higher O2 photolysis and thus more ozone production compared to March 14. The ozone loss processes on the other hand are almost identical for both cases. In the same way, the trajectory model calculations have been performed for every day during the considered periods in 1995 and 1997. The modelled ozone mixing ratios at the trajectory end-points are shown in Fig. 1: The model reproduces the observed variability very well. Including the interannual differences. As an additional test for the model, trajectories crossing the measurement site 2 or 3 times were investigated. An example is shown in Fig. 3: The trajectory crossed the measurement site also 5 and 11 days earlier. The modelled ozone along the trajectory is in very good agreement with the observed ozone mixing ratio.
The observed ozone variability, including interannual differences, is reproduced well by the photochemical box-trajectory model BRAPHO. Mid-stratospheric ozone is mainly governed by ‘dynamically controlled photochemistry’. Thus whenever trend analyses of mid-stratospheric ozone are made, the dynamics have to be taken carefully into account.
We thank Glenn Carver for providing the ASAD chemistry integration scheme. The trajectories were obtained from the NASA Goddard Space Flight Center. BMS thanks Inga Fløisand for many helpful discussions. Part of this work was supported by the European Commison, the German Federal Ministery for Education and Research and the Alfred-Wegener-Insititute for Polar Research.