Comparison of vertical boundary-layer profiles from AMDAR to a wind profiler

Wind profilers have been proven to be a valuable tool at various airports to identify wind shear and to yield input for the prediction of wake vortex development. So far, only a few airports are equipped with such profilers, usually airports with frequent adverse winds. Some previous studies however suggest that measurements collected by one single profiler on the airport compound do not sufficiently represent the flow situation in the active approach and departure corridors. A possible solution would be setting up multiple profilers around the airport. Using aircraft measurements could be an alternative to this.

The AMDAR (Aircraft Meteorological DAta Relay) program uses commercial aircraft to collect upper-air measurements as well as vertical profiles of the atmosphere. Especially the European AMDAR program predominantly provides vertical profiles of the troposphere. In the planetary boundary layer (PBL), such profiles typically have a vertical resolution slightly lower or similar to that of wind profilers. Usually one profile per hour (or per three hours, respectively) is collected at the major European airports.

Since AMDAR data are collected inside the approach and departure corridors, it seems desirable to complement wind profiler measurements with AMDAR data. For this purpose, we investigate the suitability of AMDAR data to identify PBL structures such as inversions and shear layers.

Vertical profiles measured by AMDAR are compared to wind profiler soundings at a major European airport. The differences between aircraft and wind profiler measurements of both temperature and horizontal wind vector are analyzed. The differences show most characteristics known from previous studies comparing AMDAR to other upper-air data, such as radiosondes. Here, the differences are found additionally to have a significant annual cycle.

From profiles collected by both sources, temperature inversion and wind shear layers are detected using a simple algorithm. The spatial and temporal coincidence of the detected layers is examined and the degree of mutual probability of detection (POD) is calculated. The base heights of the detected layers are reasonably correlated, while the detected thickness depends much upon on the source. Both POD and correlation are also found to depend upon the season.