TARANTO - Short-term climate dynamics and biogeochemical processes in the Gulf of Taranto: From regional proxy variations to climate records

Past and ongoing climate changes influence the physical and biogeochemical functioning of the ocean. Such responses leave their imprints in the marine sediment. As a consequence, records from marine sedimentary cores can be used to reconstruct past ocean conditions and the related climate variability by using transfer functions. Traditionally, these transfer functions are constructed by evaluating the spatial distribution of present-day sediment proxy data and environmental parameters. This approach is based on the implicit assumption that temporal variations leave the same signature in the proxy as spatial variations. Usually this cannot be shown to be the case due to the coarse temporal resolution of the proxy records and the shortness of the instrumental climate record. In this project we will try to test this assumption and attempt to derive a more appropriate transfer function. Based on this result our goal is to reconstruct the climate variability and related biogeochemical cycles for the past few centuries and the late Holocene by taking advantage of the high-quality sediment observations.

The Gulf of Taranto (southern Italy, part of the Eastern Mediterranean Sea) is a highly suited region for this research because the marine environments are closely linked to large scale climate variations. The high rate of sediment accumulation permits the reconstruction of environmental variability at a sub-decadal resolution. To achieve the goal, we implement a high-resolution Mediterranean setup of the physical-biogeochemical ocean model MPIOM-HAMOCC with a sediment module to obtain a clear understanding of the climate imprint on the sediment. Processes of sediment resuspension and transport have been included. The model system is applied in a regional stand-alone version with prescribed atmospheric forcing and river input. The model domain and grid configuration can be seen in Fig. 1. The simulation focuses on the present-day time slice from 1900 to 2010 where high-quality reanalysis data are available to force the model.


                       Fig.1: Model domain and the grid configuration of the regional ocean model MPIOM-HAMOCC.
                                         Red box shows the location of the Gulf of Taranto.

The model simulates synthetic sediment core data, which allow a direct comparison to the proxy records measured at the corresponding core sites provided by cooperating partners at the Universities of Hamburg and Bremen. Comparison will be done for the composition of the sediment (organic matter, calcite, opal and terrestrial material) and the stable isotope signatures of the sediment organic matter and shell material. An additional new feature is the temperature recording for planktonic matter at its production. This enables us to get insight into the quality of one of the key derivatives from transfer functions - the SST. The model also provides information on primary production (primary results see Fig.2) and sediment fluxes, which help to understand mechanisms driving the sedimentation processes and biogeochemical cycling. Detailed analysis of the model results can be seen here.



Fig.2: (a) Simulated spatial distribution of vertically integrated primary production in the middle Mediterranean Sea and around the Gulf of Taranto averaged over 1901-2010. (b) Time series of vertically integrated primary production (Int PP) averaged over the Gulf of Taranto, as shown in the red box in Fig.1.

A series of sensitivity tests will be carried out with the model setup to identify the anthropogenic and natural control mechanisms driving the sedimentation processes and biogeochemical cycling. In detail, we will perform sensitivity studies with respect to changes in the riverine nutrient load (e.g. anthropogenic fertilization) and changes in the strength of the physical forcing (e.g. extreme NAO).

Within the model framework in combination with the observed proxy records, we expect to be able to define a transfer function. Two approaches will be utilized here: The ‘classical’ geological one estimated from spatial variations of observed proxy and climate data, and an approach exploiting temporal variations. The latter may result in a spatially variable transfer function. Both approaches will be tested against the set of surface sediment samples covering the present-day time slice.

In a next step we will apply the derived transfer function to reconstruct the regional climate and biogeochemical variability for the instrumental period (since at least 1850, individual records even longer). The skill of the transfer function will be assessed by an independent validation with published climate records. Subsequently, we will extend the reconstruction to the full length of the available proxy records (approximately 5000 years). The aim is to study the sensitivity of the Mediterranean marine environments to large-scale climate patterns of the Northern hemisphere such as the North Atlantic Oscillation (NAO) during the late Holocene.

Applicants: Gerhard Schmiedl, Uwe Mikolajewicz, Katharina Six and Kay Emeis

This research is funded by the Deutsche Forschungsgemeinschaft.