ALBERT

Description: The biogeochemistry of an observed cold filament off Peru and the representation of filaments in submesoscale-permitting (1/45°) coupled biogeochemical model simulations are the focus of this thesis. Furthermore, we use the simulations to investigate the effect of submesoscale frontal processes on the distribution of biogeochemical tracers in the shallow oxygen minimum zone off Peru. The observed filament contains relatively cold, fresh and nutrient-rich waters originating in the coastal upwelling. Enhanced nitrate concentrations and offshore velocities of up to 0.5 m/s within the filament suggest an offshore transport of nutrients. Despite low chlorophyll α concentrations in the core of the filament, depth integrated primary production is 40% higher than at the upwelling front and 25% higher than offshore. The highly variable relationship of surface chlorophyll α and depth-integrated primary production highlights the inherent uncertainty of primary production estimates based on ocean-color measured by satellites. The observations are used to assess the results of two different biogeochemical model simulations (PISCES / BioEBUS). Both simulations exhibit filaments that are similar in lateral scale, horizontal and vertical structure and offshore extent to those observed, but differences exist in the biogeochemistry: While the PISCES simulation exhibits nitrate concentrations within filaments comparable to observations, filaments are largely depleted of nitrate in the BioEBUS simulation. This difference can be related to a higher pyhtoplankton growth rate and faster nitrate uptake in BioEBUS. The importance of sufficiently slow phytoplankton growth for maintaining realistic concentrations of upwelled nutrients offshore is therefore stressed. Furthermore, the simulations suggest that submesoscale frontal processes increase subduction and offshore export of nitrate which leads to reduced primary production. An increase in oxygen that resembles the pattern of the decrease in nitrate suggests a ventilation of the shallow oxygen minimum zone off Peru by vertical and horizontal eddy-fluxes.

Description: The quality of eddy ux-gradient parametrizations in models with coarse resolution depends on whether the generated diffusivity is similar to that of reference solutions produced by eddy resolving models. It is therefore essential to accurately describe the transport properties of eddy resolving ocean models. This thesis is a survey of the lateral transport of passive tracers induced by mesoscale eddies in the velocity field of a 1/12° numerical model of the North Atlantic. Statistical tools are used to relate particle trajectories of Lagrangian floats to the effective eddy diffusivity: Both Taylor's theory of turbulent dispersion, which forms the foundation for the analysis, and further refinements thereof, are discussed and used for computations. The underlying theories rely on several restrictive assumptions about the statistics of the flow field, and one objective of this thesis is to make the reader aware of how difficult an interpretation of the results can be.

Description: Mesoscale variability of velocities is an important part of the global ocean circulation, as it contains more kinetic energy than the mean flow over most of the ocean. Understanding its generation, dissipation and modulation processes therefore is crucial to better understand ocean circulation in general. In this thesis, a global 1/12◦ ocean model (ORCA12) is used to study the distribution of mean surface Eddy Kinetic Energy (EKE), its seasonal cycle and possible driving mechanisms, averaged over 26 years (1981-2007). For the calculation of EKE, the deviations from yearly mean horizontal velocities u, v are found to be best suitable. The model is then evaluated using EKE derived from satellite altimetry (AVISO). The total EKE from the model, including geostrophic parts, realistically reproduces the observed geostrophic mean EKE and its seasonal cycle. Seasonal cycles of surface EKE in the subtropical gyres, including most of the Western Boundary Currents (WBCs), peak in the summer months in both hemispheres. The mean EKE and amplitudes of the annual cycle are generally larger in the Pacific, compared to the Atlantic. The seasonal variations of EKE in the WBCs are driven by dissipation processes at the sea surface, namely the wind stress and thermal interactions with the atmosphere in winter. Only in the core regions of the currents other processes play a role as the surface EKE there peaks in winter/spring, not consistent with the dissipation hypothesis. The balance of dissipation and generation terms in the strong, chaotic WBCs, however, varies from year to year. In the subtropical gyres’ interior, dissipation is not solely responsible for the annual cycle. Instead, the vertical shear of near-surface horizontal velocities is found to peak in summer, in phase with the EKE. This seasonal cycle of the shear can be observed down to ∼ 150m depth, depending on the region. Inspections of profiles of horizontal velocity and EKE reveal the vertical shear to be associated with the velocity differences between the Mixed Layer and the interior ocean, possibly leading to instabilities which locally generate surface intensified EKE, largest in summer. Therefore, the seasonal cycle of near-surface vertical shear of horizontal velocities seems to be responsible for the seasonal variations of surface EKE, although the general source of EKE in the subtropical gyres remains unclear.

Description: In this study, the stratospheric winter circulation in the coupled middle atmosphere ocean model MAECHAM5/MPI-OM, is analysed. Due to the dynamical and thermodynamical interaction with the ocean, the simulated atmospheric circulation is affected by the internal variability of the ocean. Differences of the stratospheric winter circulation between MAECHAM5/MPI-OM and former MAECHAM5 simulations may be attributed to the interactive ocean. This work is divided into three parts: first the climatology of the model is examined, then major Sudden Stratospheric Warmings (SSWs), and at last the relationship between these warmings and tropospheric blockings. To examine how the model reproduces the stratospheric winter circulation, the climatology of the zonal mean zonal wind and of planetary waves of zonal wavenumber 1 to 3, are carefully analysed and compared with those obtained from ERA-40 observations. While the zonal mean zonal wind is in good agreement with observations, amplitudes and phases of zonal wavenumber 2 are not well represented. Major SSWs are analysed because of the strong impact that such phenomena can have throughout the atmosphere, influencing the weather at the surface for several weeks after the onset of the warming. To identify major SSWs, a new algorithm based on the 10 hPa zonal mean zonal wind at 60°N, is developed. This is done because a comparison between a recent study by Charlton and Polvani (2007) and the Freie Universität Berlin climatology of mid-winter major SSWs, reveals a significant disagreement. The new algorithm is applied to two databases: one obtained from the model simulation and one from the ERA-40 assimilation. ERA-40 data are used for validation of MAECHAM5/MPI-OM. Comparison of the obtained frequencies of major SSWs shows that in the model a slightly higher number of events occurs. While in ERA-40 the average frequency is of 0.60 events per year, in the model it is of 0.70 events per year. The seasonal distributions show also that the highest number of major SSWs occurs in January and in February respectively for ERA-40 and model data, which is improved for MAECHAM5/MPI-OM compared to former MAECHAM5 simulations. In this work, unlike previous model studies, the state of the polar vortex is also examined during the pre-warming phase of major SSWs, by analysing the planetary wave activity, to determine the behaviour of waves with different zonal wavenumbers. Only planetary waves of zonal wavenumber 1 and 2 appear to have a key role in the development of major SSWs, with wavenumber-1 events being more frequent than wavenumber-2 events and a ratio of 57:13 similar as observed. Because of the influence of tropospheric blockings on major SSWs via alteration of planetary waves, a correlation analysis is performed to determine if the model represents this relationship well. It appears that Pacific blockings are correlated with wavenumber-2 major SSWs although a larger number of wavenumber-2 events would be necessary to make such assertion. No significant correlation is instead obtained for wavenumber-1 major warmings.

Description: Near-inertial oscillations are an important feature of the climate system. The output of a high-resolution ocean model of the North Atlantic was used to investigate interannual variability of wind power input (WPI) to near-inertial currents with respect to the North Atlantic Oscillation (NAO). The model is forced with NCEP/NCAR reanalysis wind stress for JFM of the years 1989 (strong positive NAO-phase) and 2010 (negative NAO). Atmospheric parameters are tightly related to the NAO. The storm track in 1989 is intensified and channels storms into the subpolar North Atlantic, while it is more fanned out in 2010, allowing single storms to travel into the Mediterranean Sea. Similar patterns emerge from the distribution of near-inertial wind stress magnitude (NIWSM), i.e. the part of the wind stress spectrum that is most efficient in generating near-inertial energy (NIE). Seasonally averaged NIWSM, however, is not anchored to the storm track but is shifted to the south. This behaviour is due to the latitude-dependent inertial frequency, which decouples synoptic variability from the near-inertial frequency band. WPI for the two considered years is consistent with the different distributions of storms: While 1989 produced a total rate of WPI of 6.48 x 109 W (= GW) and a secondary centre of weakly increased WPI in the eastern subpolar North Atlantic corresponding to the intensified storm track in this region, total WPI in 2010 amounted to 9.64 GW and was associated with a strongly enhanced secondary centre of WPI in the subtropics. Although anomalies both in the storm track and mean NIWSM are more pronounced in the subpolar ocean basin, WPI prefers the subtropics. It is proposed that a mixture of atmospheric and oceanic processes is responsible for this asymmetry, chief among them the variation of the Coriolis frequency with latitude. Patterns of WPI, mixed layer NIE, and NIE in the deep ocean are similar to each other. NIE decreases drastically with depth. Mean NIWSM is a promising atmospheric proxy to WPI. Linear statistical models of WPI built from this quantity allow the estimation of total WPI for each winter (JFM) from 1980 to 2013 in low, mid-, and high-latitudes. Total WPI as well as mid-latitude WPI is only weakly correlated with the NAO. Low- and high-latitude WPI on the other hand is strongly correlated with the NAO, with the magnitude of correlation coefficients exceeding values of 0.8, suggesting that the relationship of WPI to the NAO lies in shifting the patterns of WPI. During negative NAO phases, WPI is pulled towards the subtropics (and thus intensified), whereas it shifts towards the polar ocean during positive NAO phases, both in accordance with changes in the configuration of the storm track tail. Since the response of WPI to comparable mean NIWSM is weakening with latitude, total WPI is more strongly influenced by lowlatitude WPI. It is concluded that the relationship between WPI in the North Atlantic and the NAO is of a twofold nature: While total WPI is only weakly and inversely related to the NAO, the distribution of WPI is strongly depending on it.

Description: Understanding the dynamical influence of the 11-year solar cycle on Earth’s climate andits tropospheric effects is important in order to improve decadal climate prediction. Moststudies which investigate the downward transfer of the solar cycle signal focus on theNorthern Hemisphere (NH). In contrast, the dynamical influence of the 11-year solarcycle on the Southern Hemisphere (SH) is analyzed in this thesis with an exceptionallylong (999 years) high-top model simulation of CESM1(WACCM)under pre-industrialconditions but time-varying solar cycle forcing. The constructed solar forcing only includessolar variabilities on timescales up to the 11-year solar cycle and hence neglects long-termsecular changes of the solar irradiance such as the Gleissberg cycle.The strong relative increase in UV radiation during solar maximum (SolMax) induces awarming at the tropical upper stratosphere which leads to a strengthening of the merid-ional temperature gradient in the upper stratosphere from July to September in the SH.As a result of the thermal wind balance, a westerly wind anomaly appears in the mid-latitudinal upper stratosphere during SolMax which propagates downward and polewardfrom July to October. The interactions of planetary waves with the zonal mean flow leadto a dynamical solar response indicated by the onset of the Eliassen–Palm flux (E–Pflux) divergence difference in July which coincides with the beginning of the downwardpropagation.The zonal mean westerly wind anomaly is expected to strengthen the polar vortex whichwould result in a positive annular mode pattern as seen in theNH. However, the horizontalstructure of geopotential height anomalies exhibits a polar vortex shift towards the Pacificfrom August to October in the SH during SolMax instead of a strengthening. Althoughthese geopotential height differences are found by other studies, these studies only examinethe downward transfer in zonal mean analysis. However, in this study the regional sectoranalyses reveal that the strongest westerly wind anomaly inthe Indian ocean occurs inJuly whereas in the Atlantic sector the strongest anomaly is found in August. Both regionsshow the typical poleward propagation of the westerly wind anomaly which is missing inthe Pacific region. The Pacific ocean sector is the only sectorwhich reveals statisticallysignificant tropospheric impacts on the zonal wind from August to October which areconnected to the stratosphere by a downward transfer. This highlights the importance toinvestigate regional differences and suggests that the zonal mean approach is insufficientto examine the top-down mechanism in the SH.

Description: The inflow of fresh or brackish groundwater into the sea is referred to as Submarine Groundwater Discharge (SGD). The SGD is enforced by a terrestrial component which mainly depends on freshwater extraction and recharge by meteoric water and on aquifer permeability. And a marine component that is mainly controlled by the spatial distribution of outflows and water depth (hydraulic gradients between land and sea). This studyis motivated by the importance of freshwater in arid regions and, in particular, by the continuous challenges posed by the exploration and exploitation of fresh water sources inthe Sultanate of Oman. Moreover, there is a lack of studies on SGD phenomena along the 1000 km coastline in the South of Oman. The objective of this study is to develop a method to detect SGD spots in the offshore region, autonomously, and understanding the hydrodynamics of the discharge seepage for future backtracking, quantification and coastaland groundwater management. The study area Salalah, Dhofar Governorate, South of Oman is known to have a high natural groundwater recharge during the monsoon seasonand a karstic coastal seafloor, which results in a high potential of submarine groundwater discharge spots. A geochemical tracer (radon-222) was selected to detect SGD using anRTM1688-2 radon sensor instrument. This sensor underwent experimental design performance tests to adopt to mobile offshore platform monitoring i.e Wave Glider (WG). Groundwater characteristics (i.e.Rn222, salinity, temperature) aquifers were first deter-mined by measuring Salalah’s coastal onshore groundwater wells. The preliminary radonactivity results offshore Salalah demonstrated a distinctive radon concentration gradient between groundwater and seawater with an enrichment factor of up to 4000, which is idealfor signal preservation in a freshwater plume until reaching the sea surface. Accordingto the collected morphological data and the hydrological data from literature, Salalahis found to be the best potential location for SGD. A numerical integral plume model(TAMOC - Texas A&M Oil spill Calculator) was used to investigate the detectability of single-phase freshwater plumes. The model considers the local freshwater characteristics, geomorphological and oceanographic constraints at different discharge rates anddischarge buoyancy. After validating the plume model with literature data derived from laboratory experiment, detection limit guidance for autonomous monitoring of SGDs was established from each parameter: ambient cross-flow velocity, initial discharge salinity/temperature, initial discharge velocity, initial diameter discharge and initial dischargeradon activity concentration. Moreover, groundwater flow rates of a recently investigated SGD (Dhalkut SGD) in the area could be estimated using the plume model. It is shown that measured physical and chemical oceanographic parameters’ combined with ground-water well data provide a well constrained data set to simulate SGD off Salalah quite well and could give realistic values for SGD plumes in the area. The outcome of the model was utilized to find best practices for SGD detection by using autonomous surface vehicles (i.e. Wave Glider).

Description: The study investigates salinity, depth and thickness of the isopycnals long term changes in the north northern tropical Atlantic through the density range 1026.8:1027.3 kg m-3, where oxygen minimum zone is situated. The salinity front is situated at around l8° N in the chosen density range and hydrographic property evolution is found being different to the north and south domains of the salinity front. Higher salinities during 70s than 2000-2009 period are estimated in the northern domain for all isoclines, but lower in the southern domain. Strong increase in density layer depths is observed in the southern domain intermediate water masses, i.e. Antarctic Intermediate water masses. The strongest correlation between depth and salinity anomalies is found to be -0.5 for the Antarctic Intermediate water masses, being about twice !arger than the estimates for any other water mass. In the end, the salinity front parameters such as width, gradient and variability are described over entire north tropical Atlantic.

Description: Large amounts of methane hydrate are thought to be stored in marine sediments. Natural methane hydrate deposits have been found along the world's continental margins as the prevailing low ocean temperatures and high pressures guarantee their stability. Climate change could induce a destabilization of marine hydrates due to changes in bottom water temperatures and/or sea level. Once the hydrates are destabilized they could release methane into the water column and potentially into the atmosphere, enhancing global warming. In this study a comprehensive model analysis is performed to evaluate the impact of destabilizing methane hydrates onto global climate within the next century. Additionally, the focus is set on changing bottom water temperatures to infer the response of the global methane hydrate inventory to future climate change. This study provides a new estimate of the global methane hydrate inventory based on a transfer function, which was recently developed by Wallmann et al. (2012). Global bottom water temperatures and their future evolution are analyzed in detail, as over the past few decades bottom water temperatures changed considerably along the continental margins, owing to natural, but also to anthropogenic climate variability. The current variability of the global bottom water temperatures is investigated in a hindcast simulation of the global ocean-sea ice model configuration ORCA025. The future temperature trend is analyzed by using an ensemble of 22 100-year-long global warming experiments of the Kiel Climate Model (KCM). The resulting warming trend is found to be mostly confined to shallow and mid-depth regions. Especially the warming at mid-depth could destabilize methane hydrates. As a consequence, methane could be released into the ocean and could potentially reach the atmosphere, leading to a strong positive carbon climate feedback. Based on the temperature analyses the changes in the global abundance and distribution of methane hydrates under future climate conditions are inferred. By applying the transfer function of Wallmann et al. (2012) the present-day world's total methane hydrate inventory is estimated to be 1146 Gt of methane carbon. In a worst-case scenario, where steady state is reached by 2100, the global inventory could be reduced by ~0.6%, resulting in an additional average annual methane flux of ~89 Mt from the seafloor. Based on the results of this study, the amount of methane released from melting hydrates by 2100 will not have a major impact on the global climate.

Description: Sahel summer precipitation is highly variable with floods and droughts occurring on a regular basis. Summer rainfall events were predicted using neural networks based on a collection of preseason climate indices. Understanding the prediction of a neural network is limited as this model type is a black-box. Layer-wise relevance propagation was applied to account for this issue. It allowed the attribution of single predictions to certain states in the underlying data. By that the trustworthiness of a neural network can be assessed and driving mechanisms of climate events can be pointed out. While the presented rainfall predictions are worse than earlier attempts from other studies, and consequently any event attribution is not reliable, this study demonstrates how layer-wise relevance can be implemented in climate-related contexts and illustrates the potetial of that technique in this field.

Description: While atmospheric CO2 concentrations have been increasing during recent decades due to anthropogenic emissions, the ocean has acted as a sink for atmospheric carbon. Essentially, the global air-sea flux of CO2 showed a trend towards more oceanic uptake as expected from increasing emissions. Yet, the oceanic CO2 uptake also responded to climate change and fluctuated due to climate variability and variations in the growth rate of atmospheric CO2. So far, the drivers of the variability in oceanic CO2 uptake are not conclusively understood. In this thesis, the global ocean biogeochemistry model FESOM-1.4-REcoM is used to quantify the effects of climate change and of the increasing atmospheric CO2 concentration on the trend in the oceanic carbon uptake during the period 1958-2019 (62 years). Two approaches are applied: (1) Offline diagnostics based on a linear approximation relating the trends in the sea surface temperature, dissolved inorganic carbon, alkalinity, salinity plus freshwater fluxes, wind velocity and sea-ice concentration to the trend in the CO2 flux and (2) a model experiment with the historical forcing fields compared to simulations in which certain forcing fields (e.g. winds and the atmospheric forcing fields that control the sea surface temperature) are replaced by a repeated year forcing in order to isolate their effects on the CO2 flux. In FESOM-1.4-REcoM, the ocean took up 1:85 PgCyr-1 of atmospheric CO2 on average during the simulated period. The ocean carbon sink increased with a trend of 23:8 TgCyr-1 per yr. In a simulation with rising atmospheric CO2 concentrations but without climate change and variability, the trend in oceanic carbon uptake was 27% higher than that, suggesting that climate variability has substantially reduced the uptake over the simulated period. Of this, a trend towards more outgassing of 2:9 TgCyr-1 per yr was driven by the change and variability in winds, which was particularly relevant in the polar and subpolar regions. Hereby, a comparison between the offline and online approach reveals that the effect of winds was dominated by wind-driven changes in the transport of natural carbon with the circulation. Global warming caused a trend towards more oceanic outgassing of 2:3 TgCyr-1 per yr, which mostly originated from the tropical and subtropical zone. The increasing sea surface temperature led to more outgassing due to the reduced solubility of CO2. The offine estimate for the effect of warming on the trend in the CO2 flux is much larger, which can be attributed to the neglect of compensating feedbacks. In particular, the simulated effect of global warming reveals that in response to the increasing temperature, the concentration of dissolved inorganic carbon in the mixed layer decreased, which attenuated the thermally-driven outgassing. Changes in all other variables were less important drivers of the trend in the CO2 flux.

Description: The main aim of this thesis is to investigate the responses of the Ocean Salt Content (OSC) change to perturbed surface flux forcings and its spread across climate models within the Flux-Anomaly-Forced Intercomparison Project (FAFMIP) experiments. The decompostion of the OSC into different physical processes as contributions to vertical salt transport reveals that the resolved mean and parameterised eddy advection are the main drivers for change in OSC, which are also the main processes of the global vertical salt balance. The regions of change are mainly found in eddy energetic and frontal regions such as North Atlantic Subpolar Gyre, Western Boundary Currents (WBCs) or the Antarctic Circumpolar Current (ACC). However, these are the locations where inter-model spread are also largest for control simulation as well as for the simulations with the applied surface flux perturbations. The changes and spread of OSC responses to perturbed surface heat fluxes are strongest followed by the experiment with altered surface freshwater fluxes. The signals are weakest in the experiment with changed wind stress forcing.

Description: The response of the Atlantic Meridional Overturning Circulation (AMOC) to heat flux forcing of the North Atlantic Oscillation (NAO) is analyzed. This is done with the Kiel Climate Model (KCM), a coupled ocean-atmosphere-sea ice model. The dynamical links of the atmosphere-ocean interaction in this experiment agree with other studies and theoretical considerations: In a positive (negative) NAO phase the enhanced (reduced) heat loss from the subpolar North Atlantic to the atmosphere leads to a deepening (shallowing) of the mixed layer, especially over the center of the subpolar gyre. This causes increased (decreased) convection and after 3 to 14 years a stronger (weaker) AMOC at 30°N. The NAO forcing has a slightly red spectrum. The AMOC, however, responds with lower frequencies. For the AMOC index at 30°N two modes of variability can be identified through Singular Spectrum Analysis (SSA): one with a period of 93 years and one with a period of 36 years. However, general differences of these two modes in the spatial structure of the overturning streamfunction cannot be identified. Though, the longer mode might be related to anomalies in the mixed layer depth and the strength of the subpolar gyre, which exhibit variability on similar time scales. To compare the conditions of an unforced simulation, a control experiment is also investigated. Here, the relation between the NAO and the AMOC is not as clear. The largest correlation is found when the AMOC 30°N index leads the NAO index by 1 year. Additionally, the temporal variability of the AMOC modes differs in this experiment. A low-frequency but unstable variability mode of about 120 years period is detected mainly for the first half of the simulation period. All other modes of the control experiment were interannual to decadal. This disagrees with a different control experiment version of the KCM, where a multi-centennial, a quasi-centennial, and a multi-decadal mode were dominant. Furthermore, one must bear in mind that the KCM sea surface temperatures of the North Atlantic are biased. Too low temperatures, the resulting coverage of the Labrador Sea with sea ice, but also further differences of the model compared to observations might cause the missing convection in the Labrador Sea. This region, however, is often highlighted to play a crucial role for deep water formation and therefore for the strength of the AMOC. A reduction of these errors might lead to a better understanding of the AMOC and its variability modes. This finally helps to evaluate which modes could be excited or damped under future climate conditions.

Description: Anomalies of the winter stratospheric polar vortex can propagate down to tropospheric levels and modulate variability patterns, such as the North Atlantic Oscillation (NAO), which is the leading mode of variability in the North Atlantic (NA) sector during boreal winter. Not only is the NAO important for European winter weather conditions, but the NAO related heat and freshwater fluxes, and the associated changes in westerly wind over the NA region, also influence the formation of deep water masses in the NA basin and can thereby influence the variability of the Atlantic meridional overturning circulation (AMOC). The northward transport of heat by the AMOC is very important for European climate and the variability of the AMOC is therefore of great interest. To investigate the role of the stratosphere for variability over the North Atlantic sector, two state-of-the-art ocean-atmosphere general circulation models are used: a high-top model (CESM1(WACCM)) and a low-top model (CCSM4). For each model, a Control simulation is analyzed and compared to a simulation under the Intergovernmental Panel on Climate Change (IPCC)’s RCP8.5 scenario, which represents the worst case scenario of greehouse gas (GHG) emissions. Strong and weak vortex events are defined using the Northern Annular Mode (NAM), which is also used to describe the downward propagation of these anomalies. In the low-top model the downward propagation of stratospheric NAM anomalies to the surface is not well captured, but it is very well represented in the high-top model. This simulated difference in stratosphere-troposphere coupling is also reflected in the simulated effects of the stratosphere on the surface atmosphere and ocean parameters. While stratospheric vortex events in the high-top model are connected to NAO-like anomalies at the surface (in sea level pressure, turbulent heat flux and surface wind stress), in the low-top model this connection is less pronounced. No significant changes in mixed layer depth (MLD), which is used as an indicator for deep water formation, are found in the low-top model. The high-top model, on the other hand, shows a strong connection between stratospheric polar vortex events and MLD anomalies (strong (weak) vortex events are connected to deeper (shallower) MLDs), especially in the Labrador Sea, which is an important area of deep water formation in the NA region. A cross-correlation analysis of the NAM/NAO and AMOC shows that the NAO leads the AMOC by about 4 years in both, the high and low-top model Control simulations. While the stratospheric NAM is also highly correlated with the AMOC in the high-top model (peaking when the NAM leads the AMOC by 2 years), there is no resonable correlation between NAM and AMOC in the low-top model. Under global warming the correlation between the AMOC and NAO decreases for both models. In the case of the high-top model, the NAM and AMOC are more strongly correlated than the NAO and AMOC under the GHG scenario.

Description: Atmospheric blocking is a persistent pressure pattern, which blocks the westerly flow in the mid-latitudes and strongly impacts weather both in its immediate vicinity and in regions upstream and downstream of it. Extreme events cause droughts in summer ( e.g. Russian heat wave 2010) and cold spells in winter ( e.g. European winter 1962\63). Blockings are also connected to Sudden Stratospheric Warmings (SSWs). In winter 1962\ 63 a blocking occured both before and after an SSW. However, blocking dynamics are not well understood and the phenomenon is not well-represented in climate models. To detect blockings, different definitions were derived, but deduced blocking frequencies are dependent on the blocking index. In the first part of this study, climatologies are derived from reanalysis data, to detect differences between index definitions. The indices by Lejenas and 0kland (1983) (LO*), Tibaldi and Molteni (1990) (TM) Pelly and Hoskins (2003) (PH) detect patterns, which show a reversal of the climatological flow. Anomalous high pressure systems are detectable with the index of Sausen et al. (1995) (SA). All indices show frequency maxima over the Euro-Atlantic region and the Pacific. The highest frequencies occur in winter. The TM and LO* indices show further maxima in spring and the PH index shows an additional frequency maximum in summer. Short blockings are more evenly distributed in location and season than long-lasting blocking episodes, which are accumulated over the ocean basins in winter. The second aim of the study is the investigation of the insufficient representation of blocking in climate models. Hence, climatologies are derived for CESM(WACCM). The model underestimates blockings, especially over the Euro-Atlantic region in winter. The frequency changes over the Pacific are dependent on the index. Discrepancies between the sectors are probably caused by the misrepresentation of transient eddies in the Atlantic region, which are neccessary for the formation and persistence of blockings. It is suggested that the model resolution is not high enough to simulate transient eddies, but further investigation is needed. A realistic mean flow in the model is also required. Mean flow biases could cause the underestimation over the Pacific. To investigate influences of external forcings, different model experiments were used. Increasing greenhouse gases could lead to a decrease in blocking frequency; El Nifio Southern Oscillation events probably cause a decrease over the Pacific. But these results have to be handled with care as the misrepresentation of blocking frequencies predominates differences between model runs.

Description: The decadal variability of El Niño / Southern Oscillation (ENSO) is investigated in the preindustrial control run of the GFDL-ESM2M fully coupled climate model. Overall, the climate model has quite a realistic representation of relevant ENSO properties: the probability distribution of Niño3.4 sea surface temperature (SST) anomalies is positively skewed, the highest equatorial Pacific SST variability is observed in boreal winter with the corresponding decrease in variability during spring, and the decadal climate variability shows a shift of the ENSO spatial pattern. Nevertheless, compared to the ERA-20C reanalysis product, the model shows problems most climate models have: the anomalous cold equatorial Pacific SST with the largest bias located on the eastern side, strong easterly winds over the western equatorial region, the rising branch of the Walker Circulation located too far west and the too strong subsidence regime east of the date line. Two main periods of about 60 years with high and low ENSO amplitudes are observed, ranging between l.5° C and 0.7°C. Here it is shown, that the High and Low epochs have remarkably different mean states, which can explain the differences in simulated ENSO amplitudes. The High epoch is characterized by a weaker zonal equatorial SST gradient and a warmer Niño3 SST. The less intense Walker Circulation reduces the subsidence branch, and the negative shortwave (SW) feedback during El Niño events is extended over the Niño3 domain. The stronger convective response over the eastern equatorial Pacific enhances the SST variability, increasing considerably during boreal winter, and the strong non-linearities in atmospheric feedbacks are kicked forming strong East Pacific-like (EP) El Niño events. Hence, the ENSO asymmetry is remarkably incremented. During the Low epoch, the zonal equatorial SST gradient is increased with cooler Niño3 SST. The Walker Circulation is intensified and the subsidence branch over the Niño3 region is strengthened. The Niño3 domain also coincides with the reduction of the negative SW feedback during El Niño events, as well as the incapability of the atmospheric regime to turn into a convective state, when SST anomalies are turned positive. In addition, the Niño3.4 SST variability and the wind feedback are considerably decreased during boreal winter. There are indications that the reduced SST variability of the Low epoch is caused by the too strong subsidence branch over the Niño3 region, which restricts the seasonal southward migration of the Intertropical Convergence Zone (ITCZ), and hampers the evolution of strong EP El Niño events. However, the convective response is maintained over the western equatorial Pacific, outside of the strongest mean subsidence region, as shown by the highest negative SW feedback. Therefore, during this time period the frequency of Central Pacific-like (CP) El Niño events is increased, shifting the ENSO spatial pattern, and reducing SST variability in lack of strong EP El Niños. Correspondingly, the non-linearities between the positive and negative phases of ENSO are reduced, diminishing the ENSO asymmetry. In summary, these results show how important the mean state is for the ENSO amplitude and asymmetry.

Description: In the Labrador Sea, mesoscale eddies have been identified as a major exchange agent between the fast flowing boundary currents and the quiescent interior Labrador Sea. This way, the eddies contribute to heat, freshwater and property fluxes and impact deep convection and carbon uptake. It is therefore of interest to carefully analyse the occurrence, dynamics and water mass characteristics of mesoscale eddies in the Labrador Sea. Here, four years of moored instrument time series data are analyzed for eddy occurrences. A semi-automatic method for eddy detection in moored velocity data was developed and the eddie’s time series data were fit to a Rankine vortex model in order to estimate eddy characteristics. Over the four years, three cyclonic and seven anticyclonic eddies have been detected with this method. Surprisingly, most eddies differ in their characteristics and structures from the eddies reported in earlier studies, namely Irminger Rings, Convective Eddies, and Boundary Current Eddies. In particular, no Irminger Ring was found but a cyclonic, bottom-intensified warm core eddy, which has not been reported in this area before.

Description: The Agulhas Leakage (AL) transports warm and salty Indian Ocean waters into the Atlantic Ocean and as such is an important component of the global ocean circulation. These waters are part of the upper limb of the Atlantic Meridional Overturning Circulation (AMOC) and AL variability has been linked to AMOC variability. The AL is expected to increase under a warming climate due to a shift in the Southern Hemisphere westerlies, which could further influence the AMOC dynamics. This study investigates the AL transport variability on long time scales in the pre-industrial and under a warming climate and its relation to the AMOC. It uses a high-resolution configuration of the Community Earth System Model (CESM) with a nominal horizontal resolution of 0.1° for the ocean and sea-ice and 0.25° for the atmosphere and land, which resolves the necessary spatial scales. The simulated AL transport of 19.7 ± 3 Sv lies well within the observed range of 21.3 ± 4.7 Sv. A positive correlation between the Agulhas Current and the AL is shown, meaning that an increase of the Agulhas Current transport leads to an increase in AL. Furthermore, the salt flux associated with the AL influences AMOC dynamics through the salt-advection feedback by reducing the AMOC’s freshwater transport at 34°S. In a warming climate, the AL transport was indeed found to increase due to strengthened and southward shifting winds while the Agulhas Current transport was found to decrease. Consequently, a larger fraction of the Agulhas Current will flow into the Atlantic Ocean rather than being recirculated into the Indian Ocean. The increase in AL is accompanied by a higher salt flux into the Atlantic Ocean, which destabilises the AMOC within the salt-advection-feedback. But whether and to what extent this additional salt advected to the North Atlantic could also dampen an AMOC weakening induced by increased meltwater input under climate change still needs further research.

Description: Seamounts, such as the Ewing Seamount in the south-east Atlantic, offer a habi- tat for Vulnerable Marine Ecosystems (VMEs) with increased biological activity compared to the surrounding deep sea. This is due to characteristic circulation features, such as a ’Taylor cap’ with a horizontal anticyclonic cell and vertical cir- culation cells that transport nutrient-rich water from the depths to the slopes and summit of the seamount. With the help of ocean model data and Lagrangian simu- lations, favourable conditions for marine species can be observed and quantified. One method for this is the description of the ’retention potential’, the likelihood of particles, for example nutrients or larvae, to remain in a certain area. Experiments with Parcels were carried out to investigate how the different model formulations and resolutions of INALT20 and ROMS-AGRIF affect the particle spreading at the Ewing Seamount. These model configurations differ both in the horizontal and ver- tical resolution, the formulation of the vertical axis and the atmospheric and tidal forcing. With coarser resolution and a vertical axis with levels of constant depth, there is no anticyclonic cell around the summit. Additionally, velocities near the bathymetry are generally low compared to when using terrain-following coordi- nates. As a result, particles are carried away from the seamount at a slower rate, which increases the retention potential. Tides, in turn, reduce this potential, and an atmospheric forcing with interannual variability contributes to greater variabil- ity in the particle distribution.