Abstract. Sea level in the North Sea is densely monitored by tide gauges. These measurements are used for validation as well as for the detection of extreme events. Here, we focus on the detection of sea level states with anomalous spatial correlations. An autoassociative neural network emulates the spatial correlations of gauges in a lower dimensional subspace. Anomalous sea-level states are defined by means of failure of the reconstruction model. Using spatially distributed data as input to the network thus reveals sea-level conditions with unusual spatial correlations. The corresponding atmospheric conditions indicate high wind tendencies and pressure anomalies. Quantitative analysis of such states might help assess and improve numerical model quality in the future as well as provide new insights into the nonlinear processes involved. The method has the advantage of being easily applicable to any tide gauge array without preprocessing the data or acquiring any additional information.