Performance of the surface observation network during MAP

Abstract

The comprehensive data collection effort in the framework of the Mesoscale Alpine Programme (MAP) offers the unprecedented possibility to carry out a careful evaluation of the performance of one of the densest surface observational networks of the world. Quality control of meteorological data today is usually seen as an integrated step of prognostic model initialization. If mesoscale models should be validated independently, however, a model independent quality checking procedure becomes important, especially when operating over complex terrain. For that reason DAQUAMAP, a project sponsored by the EUMETNET programme MAP-NWS, was conducted to make a high quality data set available to the scientific community testing high resolution numerical weather forecast models as well as performing diagnostic studies. The applied method of quality control consists of an automatic spatial consistency check of primary atmospheric variables. It allows to recognize gross errors and biases of individual station data and to derive station characteristics as well. The latter is especially important when validating model results with single station data. Due to a separate treatment of GTS and combined GTS and non-GTS station data a distinction of the performance of both networks could be achieved. Also a comparison between a similar exercise done with the ALPEX data set in 1982 and the MAP data set has been carried out. This allows to assess the effect of automatization in the meteorological observation networks which has taken place during the last 20 years.