アイテム詳細

要旨

Re-analysis data obtained from data assimilation are increasingly used for diagnostic studies of the general circulation of the atmosphere, for the validation of modelling experiments and for estimating energy and water fluxes between the Earth surface and the atmosphere. Since these fluxes are not specifically observed, but determined by the data assimilation system, they are not only influenced by the utilized observations but also by model physics and dynamics and by the assimilation method. In order to better understand the relative importance of humidity observations for the determination of the hydrological cycle this paper describes an assimilation experiment using the ERA40 re-analysis system where all humidity data have been excluded from the observational data base. The somewhat surprising result is that the model, driven by the time evolution of wind, temperature and surface pressure, is able to almost completely reconstitute the large scale hydrological cycle of the control assimilation without the use of any humidity data. In addition analysis of the individual weather systems in the extra-tropics and tropics using an objective feature tracking analysis indicates that the humidity data have very little impact on these systems. A discussion of this result and possible consequences for the way moisture information is assimilated as well as the potential consequences for the design of observing systems for climate monitoring is included. It is further suggested, with support from a simple assimilation study with another model, that model physics and dynamics play a decisive role for the hydrological cycle stressing the need to better understand these aspects of model parameterization.