Nat. Hazards Earth Syst. Sci., 23, 415–428, 2023 https://doi.org/10.5194/nhess-23-415-2023 © Author(s) 2023. This work is distributed under the Creative Commons Attribution 4.0 License. Detecting anomalous sea-level states in North Sea tide gauge data using an autoassociative neural network Kathrin Wahle 1 , Emil V. Stanev 1,2,3 , and Joanna Staneva 1 1 Helmholtz Zentrum Hereon, Geesthacht, Germany 2 Research Department, University of Sofia “St. Kliment Ohridski”, Sofia, Bulgaria 3 Department of Meteorology and Geophysics, University of Sofia “St. Kliment Ohridski”, Sofia, Bulgaria Correspondence: Kathrin Wahle (kathrin.wahle@hereon.de) Received: 24 June 2022 – Discussion started: 29 June 2022 Revised: 8 November 2022 – Accepted: 29 December 2022 – Published: 2 February 2023 Abstract. The sea level in the North Sea is densely mon- itored by tide gauges. The data they provide can be used to solve different scientific and practical problems, includ- ing the validation of numerical models and the detection of extreme events. This study focuses on the detection of sea-level states with anomalous spatial correlations using autoassociative neural networks (AANNs), trained with different sets of observation- and model-based data. Such sea-level configurations are related to nonlinear ocean dynamics; therefore, neural networks appear to be the right candidate for their identification. The proposed network can be used to accurately detect such anomalies and localize them. We demonstrate that the atmospheric conditions under which anomalous sea-level states occur are characterized by high wind tendencies and pressure anomalies. The results show the potential of AANNs for accurately detecting the occurrence of such events. We show that the method works with AANNs trained on tide gauge records as well as with AANN trained with model- based sea surface height outputs. The latter can be used to enhance the representation of anomalous sea-level events in ocean models. Quantitative analysis of such states may help assess and improve numerical model quality in the future as well as provide new insights into the nonlinear processes involved. This method has the advantage of being easily applicable to any tide gauge array without prepro- cessing the data or acquiring any additional information. 1 Introduction The dynamics of sea level in tidal basins are one of the most addressed topics in physical oceanography. Theoreti- cal prediction of tidal motion was pioneered by the applica- tion of a Fourier analysis by Lord Kelvin (Thomson, 1880) and later improved by Doodson (1921), who developed the tide-generating potential in harmonic form. Analysis and in- terpretation of tidal observations by Proudman and Doodson (1924) enhanced the understanding of sea-level fluctuations due to winds and changes in atmospheric pressure. The de- velopment of numerical 2D storm surge models by Peeck et al. (1983) and Flather and Proctor (1983) led to early warn- ing systems for coastal flooding. With increasing computa- tional power and the availability of satellite data, sea-level predictions have been continuously improved. However, cur- rent model predictions are not always perfect (Stanev et al., 2015a; Sandery and Sakov, 2017; Staneva et al., 2016; Ponte et al., 2019; De Mey-Frémaux et al., 2019; Jacobs et al., 2021), which emphasizes the need for further understand- ing of sea level. A recent important evolution in predicting sea level in the North Sea was achieved in the framework of the development of the northwest European shelf forecasting system (e.g., O’Dea et al., 2012; Tonani et al., 2019) by enhancing the model resolution to 1.5 km. Thus, dynamical features such as coastal currents, fronts, and mesoscale eddies are better re- solved, and the model results are improved, especially when compared to high spatial–temporal resolution observations. Satellite altimetry has added critical information in the last 30 years (Madsen et al., 2015). Notably, different Published by Copernicus Publications on behalf of the European Geosciences Union.