Investigation of Primary Production in the German Bight : a study on the influence of sea surface temperature

The German Bight is a shallow region of the North Sea, which is intensely exploited by human activities and under positive temperature trends. A broad-scale investigation of the changes in the surface chlorophyll-a (Chl-a) concentration related to the observed warming of above 1°C since 1962 is vital to understanding the primary production response in the German Bight. This thesis investigates changes in the German Bight primary production associated with increasing temperatures using Chl-a concentration as an indicator. The specific objectives to fulfil this goal are to (1) define the influence of the increasing temperatures at various spatial and temporal scales on the Chl-a concentration, (2) investigate changes in Chl-a temporally and spatially, (3) analyse Chl-a surface concentration fields to estimate the present situation, and (4) evaluate the use of machine learning (ML) algorithms to predict the Chl-a time series. The conclusions are summarized as: (1) evaluations and observed changes in variability (seasonal and interannual) cannot be ignored in temperature considerations because they are part of the significant changes occurring in temperature that affect ecological systems; 
(2) the assessment of Chl-a variability reveals that seasonal variations are dominant in the German Bight, regulated by two observed phytoplankton blooms. The nonseasonal variability is also defined by the interannual and intra-annual variability of the spring blooms, dominated by non-seasonal forcings, such as wind variability connected to NAO and the inflow of Atlantic water types that are warmer and poorer in nutrients from the English Channel and freshwater rivers and; (3) this work demonstrates the ability of ML models to use environmental in situ time series to predict Chl-a concentration with significant accuracy while also defining the best predictors. This work is a step towards using ML algorithms in marine areas based on long-term time series.

Rechte

Nutzung und Vervielfältigung:


CC BY 4.0

Bitte beachten Sie, dass einzelne Bestandteile der Publikation anderweitigen Lizenz- bzw. urheberrechtlichen Bedingungen unterliegen können.

Zitieren

Zitierform:
Zitierform konnte nicht geladen werden.