ALBERT

Description: For simple analysis of stochastic climate models the ocean is often forced with a statistical atmosphere model. The atmosphere model mimics the observed statistics of the atmospheric forcing, e.g. the heat fluxes and the wind stress. This study serves as the beginning of the development of such a global statistical atmosphere model. The starting point of the development is a Monte-Carlo-like model written by Dietmar Dommenget, which is coupled to an one-dimensional model of the upper ocean. An important question in this context is how good the existing model simulates the atmospheric forcing. For that purpose the probability distribution functions of the net heat flux and the wind stress (respectively the surface fricion velocity u* ) derived from observations and a coupled run of the GCM ECHAM5 with the above mentioned ocean model are examined. The results are compared to the output of the statistical atmosphere model. The spatial and temporal patterns of the statistical moments mean, standard deviation, skewness and kurtosis are considered particularly. The investigation of the moments shows considerable differences between the model data and the observations. Especially the wind speed and thereby the friction velocity of the observations differs from that calculated by ECHAM5. The distributions of the friction velocity simulated by the statistical model deviate from both the observations and the ECHAM5 model data. By performing sensitivity studies it is shown that the deviations between the probability distribution functions have a non-negligible influence on the evolution of the sea surface temperature. The results of this analysis lead to possible modifcations of the statistical atmosphere model. Two different atmosphere models including these modifcations are presented. Another approach for the development of a statistical model, the usage of spatial correlation patterns is elucidated. Because of the enormous number of EOF modes needed to reach 90 percent of explained variance, even in the coupled EOF analysis between u* and the netflux, it is refrained from developing a statistical atmosphere based on these EOF modes. Finally an atmosphere model based on the bulk formulas is formulated as a result of a cross spectral analysis, which indicates an underrepresentation of the low-frequency variability in the SST time series caused by the direct forcing of the ocean with the surface fluxes. The SST time series simulated by this model exhibit a much higher coherency with the SST of the ECHAM5 model run.

Description: The goal of the here presented study was to elucidate the effect of temperature on the production and lass mechanisms of dissolved organic matter in a microalgal culture. For this purpose, an axenic diatom culture was grown in North Sea water containing a natural bacterial community and exposed to three different temperatures (17°C, 21°C and 26°C). During the following 25 days the development of algal growth and bacterial activities as weil as changes in the concentration of DOC, carbohydrates and TEP were closely followed. The here obtained results show a clear dependence of the autotrophic and heterotrophic processes on temperature. At the highest temperature, algae grew about two times faster than those incubated at the lowest temperature. The overall produced amount of algal biomass did not differ between the three temperature incubations as indicated by the concentration of POC. Contrary to this, algae growing at the lowest temperature produced considerably less Chi a than those at the highest temperature. lt is suggested that cells growing at their lower temperature limit cannot efficiently use large amounts of light energy captured by photosynthesis, because their intracellular processes are considerably slowed down. Thus, by producing less Chi a they absorbed less light energy at a time, but used it efficiently to build up biomass over a langer period of time. All cultures reached a situation of nutrient depletion in the course of the experiment. The ratio of C: N rase up to values of ~20:1. Thus, the extracellular release of carbon-rich dissolved organic compounds was highly probable. Astonishingly, no effect of temperature on the pools of DOC and carbohydrates could be observed. Interestingly, the production of TEP was influenced by temperature. An increase by 9°C led to a more than two-fold increase in the amount of TEP produced. Based on this finding it is suggested that the production of TEP exhibits a temperature-sensitivity that is on the same scale as heterotrophic processes. Thus, the hypothesis that the production of TEP could be reduced at elevated temperatures due to a faster uptake of DOM by bacterial processes, has to be rejected. A decrease in POC and TEP at the highest temperature was observed in the second half of the experiment, whereas bacterial activities increased. High activities of bacterial a-Dglucosidase and ß-D-glucosidase suggest that a considerable amount of carbon-rich DOM was exuded in the course of the experiment. In summary, the data of DOC cannot be explained to this point of time. An increase in temperature clearly affected the dynamics in the autotrophic and the heterotrophic processes. Since even the pool of TEP was positively influenced, it seems likely that temperature has an effect on the pool of DOM.