A current critical issue in climate change studies is how temperature changes and shifts on different spatial and temporal scales can affect organisms in terms of trends, variability and frequency of extremes. In this paper, we analysed marine temperature data on different temporal and spatial scales. We related the sea surface temperature data from the Helgoland Roads Time Series, one of the most important and detailed long-term in situ marine ecological time series, to the Sylt Roads, North Sea, Germany, Europe, North Atlantic and Northern Hemisphere surface temperatures. All time series showed a distinct upwards shift in temperature in the late 1980s, early 1990s, with positive trends in overall for the period between 1962 and 2019 ranging from 1 to 2 °C over 57 years. We quantified changes in temperature variability by comparing the years before and after 1990, on both long-term and seasonal scales. At Helgoland and Sylt, an increase in the number of warmer days in summer and a decrease in extremely cold days in winter are the new characteristics of the temperature pattern after 1990; higher than expected temperatures now also occur earlier during the year. For these locations, we observed the highest trends overall, i.e. of around 0.3 °C/decade. The observed bimodal shape of the probability density functions, characterized by winter and summer modes, had become more heterogeneous, with the cold mode peak moving to higher values and the steepness to the peak increasing, which is a consequence of a decrease in extremely cold days. North Atlantic Oscillation (NAO) and Multidecadal Oscillation (AMO) large-scale phenomena had no significant correlations or, for the NAO, were limited to the winter season at the regional and local scales. The closest landmass (mainland Germany) temperature was highly correlated with the North Sea sites. Taken together, our results suggest that marine pelagic ecosystems and their species are subject to temperature shifts with similar patterns but with variations in magnitude at the different scales. Temperature is one of the main drivers of species diversity and distribution, and this manifests on different spatial and temporal scales depending on population growth, life stages, cycles and habitat. Accordingly, we here present the temperature changes on the appropriate spatio-temporal scales, and thus provide the suitable and useful fundament for studies on the effects of warming on marine ecosystem function and biodiversity.