Mixing layer height derived from radiosoundings and ground-based lidar - comparison and assessment

The inversion on top of the atmospheric boundary layer is a strong barrier for the transport of heat, momentum and matter from or to the earth's surface. Regarding aerosols and gaseous constituents like water vapour which originate from the surface, the concentration of those parameters within the boundary layer strongly depends on the height of the layer, the mixing height. During daytime the mixing height over land increases and reaches a maximum value in situations with constant synoptic conditions.

In many applications, e.g. the comparison of model output with observations the mixing height is taken from radiosoundings. Often this value is - due to lack of other measurements - also taken as the height of the fully developed convective boundary layer. Since the mixing height is strongly varying both in time and space an observation along a single line like a radiosonde track represents only an estimate of the mixing height.

Quasi-continuous measurements of the backscatter signal with a ground-based lidar from several field campains offer the opportunity to estimate the error associated with a mixing-height determination from radiosoundings. A method to determine the mixing height from the backscattered signal is presented.

Data from several field campaigns are used, namely the Nauru99 campaign in the tropical western Pacific in June/July 1999, three campaigns at the ARM-site in Oklahoma (USA) in September/October 1999, September/October 2000 and November/December 2000 and three campaigns in the frame of the German EVA-GRIPS (Regional Evaporation at Grid/Pixel Scale) project near Lindenberg (Brandenburg, Germany) in September 2002, April 2003 and May/June 2003 (LITFASS-2003).

From these campaigns measurements simultaneous to approximatly 50 radiosoundings exist and allow a statistical analysis of the results. The comparison of radiosonde mixing heights with lidar mixing heights over 10 min time intervalls reveal a good agreement, the better the shorter the distance between radiosonde launch point and lidar location. Lidar mixing heights averaged over 1 h, which are more representative for an area, may deviate up to 300 m from radiosonde mixing heights. The standard deviation within the averaging interval fairly represents the variability of the mixing layer height. The maximum mixing height which is given by an afternoon radiosounding is also compared with lidar measurements.