Convection detection for automatic METAR reports using radar, satellite and lightning data

The German national weather service (DWD) operates manned weather stations at all 17 international airports. The half hourly flight weather observations (METAR) done by human observers are subject for full automation. The “AutoMETAR” project started in 2014 at DWD. METARs provide information about the weather conditions at and in the vicinity of airports, e. g. wind, visibility, present weather, cloud cover and height, temperature and pressure. As showers and thunderstorms can be a security risk to civil and general aviation, it is necessary to encode convective clouds and present weather into the METAR. In order to detect towering cumulus (TCU), cumulonimbus (CB), showers (SH) and thunderstorms (TS) automatically, the sub-project “AutoKON” (automatic convection detection) was launched. Its aim is to use especially radar, satellite and lightning data. Remote sensing data have the advantage of an areal detection in contrast to local in-situ measurements. Multiple data types are used to receive the highest benefit, robust products and to avoid product outages. The work will outline the developed AutoKON structure. It is planned that the algorithms analyze the data in two steps. The detection algorithms for each data type as first part are followed by a data fusion to combine all available data in the second part. The first part is separated into three algorithms: the cloud detection for TCU and CB based on satellite information; the cell detection algorithm using high resolution radar data to detect stronger showers; and, as weaker showers are more difficult to detect by cell detection algorithms, a convective / stratiform separation algorithm is planned. In the second part of AutoKON the pre-analyzed radar and satellite data will be combined and completed with lightning data. This data fusion will determine the convective present weather and clouds that are provided for the following automatic generation of METARs.