Description: Simulations with a free-running coupled climate model show that heat release associated with Southern Ocean deep convection variability can drive centennial-scale Antarctic temperature variations of up to 2.0 °C. The mechanism involves three steps: Preconditioning: heat accumulates at depth in the Southern Ocean; Convection onset: wind and/or sea-ice changes tip the buoyantly unstable system into the convective state; Antarctic warming: fast sea-ice–albedo feedbacks (on annual–decadal timescales) and slow Southern Ocean frontal and sea-surface temperature adjustments to convective heat release (on multidecadal–century timescales) drive an increase in atmospheric heat and moisture transport toward Antarctica. We discuss the potential of this mechanism to help drive and amplify climate variability as observed in Antarctic ice-core records.

Description: Characterizing the variability across timescales is important for understanding the underlying dynamics of the Earth system. It remains challenging to do so from palaeoclimate archives since they are more often than not irregular, and traditional methods for producing timescale-dependent estimates of variability, such as the classical periodogram and the multitaper spectrum, generally require regular time sampling. We have compared those traditional methods using interpolation with interpolation-free methods, namely the Lomb–Scargle periodogram and the first-order Haar structure function. The ability of those methods to produce timescale-dependent estimates of variability when applied to irregular data was evaluated in a comparative framework, using surrogate palaeo-proxy data generated with realistic sampling. The metric we chose to compare them is the scaling exponent, i.e. the linear slope in log-transformed coordinates, since it summarizes the behaviour of the variability across timescales. We found that, for scaling estimates in irregular time series, the interpolation-free methods are to be preferred over the methods requiring interpolation as they allow for the utilization of the information from shorter timescales which are particularly affected by the irregularity. In addition, our results suggest that the Haar structure function is the safer choice of interpolation-free method since the Lomb–Scargle periodogram is unreliable when the underlying process generating the time series is not stationary. Given that we cannot know a priori what kind of scaling behaviour is contained in a palaeoclimate time series, and that it is also possible that this changes as a function of timescale, it is a desirable characteristic for the method to handle both stationary and non-stationary cases alike.

Description: El Nino has strong impacts on climate and ecology both regional and global. In this study the response to global warming in the twenty-first century is analyzed. Surface temperature data from the output of 21 models of the Coupled Model lntercomparison Project (CMIPS) are used. As a measure of the strength of El Nino the Nino 3 and Nino 4 indexes are used. To examine the amplitude of ENSO the standard deviation is analyzed, as well as the skewness and the frequency spectrum. The multi model mean can show the observations reasonably but at the same time the variability of each model is slightly overestimated. As expected the sea surface temperature in the tropical Pacific rises as a response to global warming. Changes in variability of ENSO cannot be detected. Some models simulate an increase, others a decrease and some show no change in variability at all. There are no changes in the frequency of El Nino. There are changes in the structure of El Nino in some models, but there are no further examinations made in this study. Nearly all models show a significant change concerning global impacts of El Nino in high latitudes but the models show no consistence as to whether temperatures will rise or fall, so no clear conclusion can be drawn.

Description: The tropical pacific is governed by the El Nino Southern Oscillation (ENSO). Understanding this coupled ocean-atmosphere oscillation is crucial due to its global teleconnections on temperature and pressure. The present study examines the natural variability of ENSO, using a 4500-yr unforced control run from the Kiel Climate Model, version 1.4.0. Compared with observations, the preindustrial run displays a cold bias of serveral degrees in the equatorial pacific. The model also exhibits realistic spatial patterns of variability in key variables like sea surface temperature (SST), sea level pressure (SLP), zonal 10m Windspeed (WO) and the heat content of the upper 300m of the ocean (WO). As seen in previous studys of the KCM, deviations between model and observations are prominent in anomalous SSTs in the western equatorial pacific. Although overestimated by ~ 17%, the amplitude of ENSO exhibits a realistic range of variability. The present study reveals that the characteristics of ENSO in this control run can mainly be described by the equatorial SST- and SLP-indices like Nino 3.4 Index or Equatorial Southern Oscillation Index (EQSOI). The common used Southern Oscillation Index (SOI) seems to be inapplicable for this purpose. In the analysed parameters of the nm, ENSO exibits a realistic periodicity on a time scale of 1,7-5 years with a maximum of variation around 3 years. There is also high variability on multidekadal time scales without any significant peaks. In this sense it can be explained by a damped AR1-Process.

Description: The vertical distributions of early developmental stages of oceanic fishes were investigated across the tropical and equatorial Atlantic, from oligotrophic waters close to the Brazilian coast to more productive waters close to the Mauritanian Upwelling Region. Stratification of the water column was observed throughout the study region. Fishes were caught with a MOCNESS-1 net with mouth area of 1 m2 at 11 stations. Each station was sampled both during the day and at night within a single 24-h period. The investigation covered both larvae and transforming stages from the surface to 800 m depth. Distribution patterns were analysed, and weighted mean depths for the larvae and transforming stages of each species were calculated for day and night conditions. Forty-seven different species were found. The highest number of species occurred in the three stations south of Cape Verde Islands, characterized by a mixture of South Atlantic Central Water (SACW) and Eastern North Atlantic Central Water (ENACW). There was a marked drop in species richness in the three stations closer to the African upwelling, dominated by ENACW. The highest abundances occurred in the families Myctophidae, Sternoptychidae, Gonostomatidae and Phosichthyidae. Day and night vertical distributions of larvae and transforming stages showed contrasting patterns, both in the depths of the main concentration layers in the water column, and in the diel migration patterns (where these were observed). Larvae generally showed a preference for the upper mixed layer (ca. 0–50 m) and upper thermocline (ca. 50–100 m), except for sternoptychids, which were also abundant in the lower thermocline layer (100–200 m) and even extended into the mesopelagic zone (down to 500 m). Transforming stages showed a more widespread distribution, with main concentrations in the mesopelagic zone (200–800 m). Larvae showed peak concentrations in the more illuminated and zooplankton-rich upper mixed layers during the day and a wider distribution through the upper 100 m during the night. For most species, transforming stages were concentrated in the mesopelagic layers both day and night, although in some species (Diaphus cf. vanhoeffeni and Vinciguerria nimbaria), the transforming stages displayed vertical migration into the upper 100 m at night, in a manner similar to their adult stages.

Description: Two time slice experiments (1959-1990 (20C) and 2069-2100 (21C) respectively) of the ECHAM5 model were performed and the results served as the basis for the analysis of storms in the Mediterranean Sea in this thesis. The goal is to investigate potential changes of the simulated storms under climate change as well as to elaborate differences between the identified "tropcial like" cyclones and true tropical systems. The analysis of the genesis positions and tracks of all simulated storms reveals two main genesis regions, one in the northwestern and a second in the southeastern Mediterranean Sea. Most of the systems move eastwards from where they developed. Both for genesis positions and tracks, no change is observed under climate change. In a frequency distribution of the maximum intensities of the storms, a statistically significant decrease of high maximum intensities is found for all the storms. A criterion found in the literature helping identifying true tropical cyclones is varied and applied to the data set. A decrease of 30% to 50% in the number of the "tropical like" storms can be found between 20C and 21C. No statistically significant changes in the frequency distribution of the maximum intensity is observed. Problems using the criterion are identified giving indications for how the criterion should be modified for "tropical like" cyclones in the Mediterranean Sea. It should hence take into account that the "tropical like" cyclones only develop a warm core in certain vertical levels and during certain time steps. A possible approach is considering the increase in temperature anomaly between certain height levels. This study reveals that the storms identified by the 554-criterion (minimum value in relative vorticity in 850 hPa (in 10-5 s-1), minimum decrease of relative vorticity between 850 hPa and 250 hPa (in 10-5 s-1) as well as the minimum number of time steps the first two named criteria have to be consecutively fulfilled) show an increase in temperature anomaly between 700 hPa and 400 hPa 2.5-times magnified compared to those storms not showing any decrease in relative vorticity between 850 hPa and 250 hPa The simulated genesis positions of the "tropical like" cyclones show a decrease in the number of systems in the central Mediterranean Sea in 21C compared to 20C. The analysis of the atmospheric wind shear and stability conditions don't give any clear explanation for that. A more detailed analysis of the atmospheric conditions for any particular season has to be performed to find more distinct correlations. In this study, besides differences in the warm core, further differences between "tropical like" cyclones and true tropical systems are found in the maximum attained wind speeds, the ,, radius of maximum winds" and the genesis date of the systems.

Description: The time series of the global mean temperature in the last century shows different periods of steeply increasing temperature and periods where the temperature stagnates. A possible reason for these differences might be internal variability of the climate system. Here I apply singular spectrum analysis (SSA) to surface climate data in order to find long-periodic oscillations. Therefore different regions in the northern and tropical Atlantic and Pacific are analysed, where variations on interannual to multidecadal timescales are already known (except for the tropical Atlantic). The Atlantic Multidecadal Oscillation (AMO), the Pacific Decadal Oscillation (PDO) and the El Nino / Southern Oscillation (ENSO) can be proofed by using the SSA. Comparing these oscillations with the time series of the global mean temperature, evidence that these oscillations can cause the different phases of the temperature profile can be found. Thus, this might be the reason for the current global warming hiatus.

Description: A major issue with current CMIP5 climate models is that all of them have a pronounced SST bias of several degrees centigrade in the eastern equatorial Atlantic resulting in a poorly reproduced climate in the tropical Atlantic region. Therefore, a look at this region is particularly interesting in order to determine if an increased model resolution would have a significant effect on the reproduction of the tropical Atlantic climate and subsequently the climate projections for this region but also globally in times of increasing human influence on the system. Understanding the role of resolution is a crucial factor in simulating and understanding the earth's climate system. A higher model resolution allows to simulate smaller-scale processes so that it seems reasonable to suggest a higher resolution would yield a better reproduction of the earth's climate system. However, model resolution is constrained by the computing capabilities of the available hardware and thus affecting computation time and costs. In this study the Kiel Climate Model (KCM), a coupled ocean-atmosphere general circulation model, was used to investigate the above question. Only the atmospheric horizontal and vertical resolution of the model have been increased while the oceanic component was left unchanged. This new setup was then used to render a global warming run with 1 % carbon dioxide increase per year. This was done both for the low- and high-resolution setups as well as a control run with no carbon dioxide rise respectively for each setup. The results were then compared to find differences between both configurations. The findings suggest that just increasing the atmospheric resolution can result in drastic changes of the model output. The comparison of the two KCM versions in this study shows a much more realistic reproduction of the mean state in the tropical Atlantic and furthermore a much stronger response of the tropical Atlantic climate to increasing atmospheric carbon dioxide when using a higher horizontal and vertical atmospheric resolution. The high resolution output shows distinct changes in the general atmospheric and oceanic circulation under global warming like a southward shift of the ITCZ, reduced seasonal climate variability in the equatorial Atlantic, a persisting zonal circulation change similar to the pacific El Ninio with subsequent changes in precipitation and a general intensification of the hydrological cycle. Many of these changes are not even present at all in the low-resolution output. From these findings we can estimate that we should expect many more surprises to come as climate models continue their journey towards higher and higher resolution and other improvements. The results of this study also serve as a strong reminder that we are still far from understanding all aspects of our planet's climate while humankind has already begun changing it.

Description: The Atlantic Meridional Overturning Circulation (AMOC) influences North Atlantic climate and is responsible for relatively warm temperatures in northern Europe compared to other places at same latitudes (Cunningham et al., 2007). Therefore the exact mechanisms and reactions to external impacts and fluctuations of different parameters are very important parts of current research for the reason that a certain wind stress field could possibly give information about the future strength of the AMOC. Within the scope of this Bachelor thesis ten model runs of the Kiel Climate Model (KCM) are driven with global wind forcing by ERA40 and NCEP wind stress datasets to observe the influence of wind stress on AMOC. It shows that the Overturning Circulation has a decreasing trend during the observed period from 1958-2001, while at the same time wind stress is increasing. This opposing trend allows the assumption that other processes like heat fluxes or density driven transports superpose the influence of the wind stress and that the decadal trend of the AMOC is hardly influenced by windstress (Cunningham et al., 2007). Furthermore a negative correlation between AMOC and wind stress, meaning that an increase of AMOC would lead to a decrease in wind stress, can be excluded (Eden et al., 2001). It rather shows, that wind stress is at least partly responsible for interannual variabilities. This influence has its maximum impact with a time delay ("lag") of three years after an event in wind stress. The highest positive correlations are found in the North Atlantic region in a belt from the US east coast to the British Islands. Here an increase of the windstress curl would lead to maximum changes of AMOC transport strength with a time delay of three years. In this thesis wind driven water mass transport is described by the Ekman transport, which makes up about 10% of the total Overturning transport. The variability of Ekman transport and zonal wind fluctuations are quite strong in the northern Atlantic and can significantly influence the AMOC on interannual timescales.