On the temporal variance of stratospheric trace gas concentrations

Abstract

The relative variances in stratospheric observations of longlived trace gases CH4, N~O, CF2C12, CFC13 show large differences. These differences axe greatly reduced when the local mean standard deviation is normalized to the local vertical gradient. This ratio, called 'equivalent displacement height' exhibits ~ characteristic vertical profile which is very similar for all the longlived trace gases, as well as for O 3. With the help of this ratio, it is demonstrated that the variances are essentially due to natural causes, i.e., transport. Using the mixing length hypothesis a theoretical expression for the equivalent displacement height is derived. From it the meridional slope of the mixing surfaces and the mixing length can be calculated as function of the altitude. Additional information is obtained on the variance'of that slope. The uncertainty of the measured mean trace gas profiles is defined.Much of our faith in current models of stratospheric chemistry rests upon the fact that the model predicted trace gas distributions 'agree' with measured profiles. Agreement is not well deFmed; it usually means that the model predicted vertical profile follows approximately the shape outlined by the measurements and falls within the range of the data points. Such agreement for a sufficient number of trace gases is said to validate a model. Clearly one would prefer a more quantitative intercomparlson of model prediction with experimental observation. In the past, this was hampered by a lack of experimental data. The observations were sparse, made in different locations and seasons. Consequently the data showed a fair amount of scatter. It was difficult to decide how well the measurements represented the mean stratospheric trace gas distribution and therefore impossible to assign a meaningful uncertainty to the observations.In the meantime, the data base has somewhat improved. For a few trace gases, namely

Stratospheric Transport

“…3. This deduction remains true whether variations in mixing ratio are from temporal or spatial (vertical or horizontal) sampling, in agreement with observational findings (e.g., Ehhalt 1983;Fahey et al 1990). From a theoretical viewpoint, the tracer-tracer relationship can never be precisely compact in the presence of mixing unless the relationship is linear but, in practice, scatter is often weak, unless (as we shall see) data are included from outside the surf zone.…”

Climatologies of subtropical mixing derived from 3D models

“…In order to avoid these problems, Ehhalt et al (1983) normalized the local standard deviation with the local vertical gradient. This so-called "equivalent displacement height" is used to examine the temporal variance of stratospheric tracers.…”

Tidal signatures in temperature data from CRISTA 1 mission

“…Both the zonal mean signatures and the variations in the tidal amplitude with longitude are presented. The consequences of a Lagrangian interpretation of the data are then outlined and the approach used to convert the temperature differences to what are termed here equivalent vertical displacements (a concept similar to notion of "equivalent displacement height" developed by Ehhalt et al [1983]). The observations are compared to updated results from the global scale wave model .…”