A large storm surge event occurred on 2007 November 2009 in the southern North Sea where strong winds caused the sea level to rise drastically by up to 3 m within several hours. Based on the Proudman Oceanographic Laboratory storm surge model, the predicted loading displacements at coastal stations can reach a few centimetres in the vertical and several millimetres in the horizontal directions. In this study, we used two-hourly global positioning system (GPS) positions at 26 stations around the southern North Sea to identify the loading displacements caused by this storm surge event. We find that the mean rms of the differences between the estimated and predicted displacements are 4.9, 1.3 and 1.4 mm, which are insignificant compared to the one-sigma GPS positioning errors of 5.1, 2.0 and 2.4 mm for the Up, East and North components, respectively. More interestingly, in both vertical and horizontal directions, the estimated displacements successfully tracked the temporal evolution of the storm surge loading effects. In addition, within the whole of 2007 November, we used the predicted displacements to correct the two-hourly GPS positions, and consequently reduced the rms of the estimated displacements on average from 9.3, 3.0 and 2.9 mm to 7.8, 2.8 and 2.8 mm for Up, East and North components, respectively. Therefore, subdaily loading effects due to storm surges should be paid attention to in the GPS positioning that contributes to crustal-motion studies around shallow seas such as the North Sea, the Baltic Sea and the Gulf of Mexico.