Accueil > Our research topics > Adaptive Observation > Impact of marine observations on weather predictions

Impact of marine observations on weather predictions

by Alexis Doerenbecher and Anne-Lise Dhomps

published in the Météo-France 2016 Research Report (ISSN : 2116-4541)

 

EUMETNET (EUropean METeorological NETwork) has funded studies to quantify the importance of sea surface observations (vessels and buoys) on the quality of weather forecasts. CNRM   (French National Meteorological Research Center) participated and implemented the linear approach FSOi (Forecast Sensitivity Observation impact) based on the sensitivity of forecasts to observations in ARPÈGE (one of Météo-France’s models) over a period of 21 months (July 2013 to April 2015), focusing on the 24-hours forecast error on Europe and Eastern Atlantic.

Most of moored buoys are not considered in the study because the focus was on the impact of mobile stations. Pressure and wind observations are assimilated and diagnosed with the FSOi. Various aspects of the impact have been studied : its dependence on latitude, on weather regimes and its relation to the density of observations. The distribution of the impact between buoys (BUOY), vessels (VOS) with automatic measurements or human measurements as well as contributions by parameter are examined.

Most drifting buoys do not measure wind but their surface pressure measurements are considered to have better quality than those from VOS. Hence, the impact of the buoys largely dominates the impact of VOS. But ships and buoys are complementary in their Coastal / High-Seas distribution. Indeed, the data from VOS are very numerous near the coasts along shipping routes. Buoys are very well distributed throughout the North Atlantic basin, except in the South of the domain and in the Mediterranean. Thus, on a large Gascogne coastal zone, a unit observation from a vessel contributes 10 times less than a buoy observation, but the highest density of the VOS produced a cumulative impact which reaches up to 50% of the impact of buoys.

 

Geographical distribution of observations from VOS (a) and buoys (b) for the whole sample. The black rectangles show the delimitation of the "High-Seas" and "Coastal" zones used in the diagrams (c, d and e) of relative contributions in number (c) in impact per unit observation (d) or accumulated impact (e) from Buoy and VOS measurements.