Exploiting the AIRWAVEv2 Total Column Water Vapor dataset: atmospheric lee waves detection over the Mediterranean Basin

Lee waves are internal gravity waves which are produced when a stably stratified flow is forced over an obstacle. They are nowadays a well-recognized cause of hazard to aviation, especially under dry condition, when they are not clearly visible. They may also impact the global circulation and the climatological momentum balance, because of their interaction with high mountains, causing non-linear perturbations which may, in turn, interact with the larger-scale flow. The AIRWAVE (v1 and v2) datasets contain 20-years of day-night Total Column of Water Vapour (TCWV) over water surfaces at the ATSR native 1x1 km2 spatial resolution grid, allowing the investigation of small scale features as the lee waves. We focused our analysis on the Mediterranean region, which is the largest semi-enclosed basin on the Earth. The peculiarities of this area, characterized by complex orography and rough coastlines, may indeed lead to the development of such phenomena. For this purpose, the Mediterranean area was divided in three regions. To identify the regions characterized by lee waves we considered the AIRWAVEv2 TCWV products within an "optimized" latitude/longitude grid. For each grid point, the standard deviation was then computed and used as a proxy for the detection. The classification has been based on the number of grid points having a TCWV standard deviation exceeding a defined threshold. The atmospheric conditions which lead to the lee waves development were identified using the ECMWF wind fields at the finest resolution 0.125°x0.125° (10 metre U wind component, 10 metre V wind component every 6 hours). During the work it was clear that day time ATSR overpasses may sometimes be affected by wrong cloud mask classification. To overcome this problem, a cloud masking classification based only on ATSR infra-red channels was applied, and new TCWV values have been obtained from the re-analysis of these data. Here we show some examples of lee waves detections, comparing the AIRWAVEv2 TCWV fields with measurements from other sensors, e.g. from Synthetic Aperture Radar (SAR) observations, and with atmospheric model output from Weather Research and Forecasting (WRF) model. The application of the AIRWAVE algorithm to the current sensors, e.g. to Sentinel 3 SLSTR, would enable the extension of this study to a longer time period.