Tracing Trajectories of Air Parcels Transported through Spatially Resolved Horizontal Neutral Wind Fields Observed in the Thermosphere above Alaska

Abstract:

Transport by fluid flow is a very complex problem. Any type of velocity gradient introduces distortion of the original air masses which, over time, can become extraordinarily severe. In Earth’s thermosphere, it is however widely presumed that viscosity hinders both horizontal and vertical wind shears, and hence rapidly attenuates any gradients that might occur over distances shorter than synoptic scales. As a result, particle trajectories predicted by current models are often relatively simple, so that transport effects only slowly disperse and mix air masses. This means that regions of perturbed chemical composition, formed for example by intense aurora, would be expected to remain intact for many hours or even days. However, our observations show that this simple picture does not hold in practice; wind fields in the thermosphere have much more local-scale structure than predicated by models, at least in the auroral zone. These local small scale structures complicate air parcel trajectories enormously, relative to typical expectations. In Alaska, three Scanning Doppler Imaging Fabry-Perot interferometers are currently in operation. A single SDI instrument can simultaneously observe the thermospheric wind’s line-of-sight component in 115 (typically) independent look directions. From these data it is possible to reconstruct time-resolved two-dimensional maps of the horizontal vector wind field, and use these to infer forward and backward air parcel trajectories over time. Tracing parcel trajectories through a given geographic location maps where they will go from there (forward tracing in time) and where they come from previously (history of parcels or tracing back in time). Results show that transport of thermospheric neutral species in the presence of the local scale wind gradients that are actually observed is far more complicated than what current models typically predict.