Toward automatic convective products for aeronautical meteorology

In Austria recent projects focused on automatic aeronautical products for seamless convective prediction such as the Analysis of Available Airspace for Air Traffic Management (AAA4ATM) and the detection of convection for AUTOMETAR using sensor fusion (FUSEMET-APP). Weather radar data are used extensively to facilitate such forecasts and observation for Air Traffic Management.

Within AAA4ATM computer vision algorithms for detection of objects, feature extraction, 2D/3D tracking and extrapolation as well as down- and up-scaling strategies in meteorological image processing and blending to NWP model are used. Newly developed forecaster intervention tools allow manual modification of the automatic forecasts to improve the results. Verification results show the high impact of feature extraction such as the detection and orientation of line of echoes in relation to the flight paths of aircrafts for different altitudes.

A Bayesian approach is used for sensor fusion of different remote sensing data in the FUSEMET-APP project where especially the intermediate state of convection without lightning (TCU or winter CB) are more difficult to detect. Additional error handling on radar data will be addressed (e.g. blocking, bright band or wind farm influences close to airports).