The Rotenburg area was monitored by a small-scale, high-resolution seismic network comprising 3c, sparse array, and classical array stations. Installation was challenging due to high local noise levels, but finally three years of continuous monitoring could be acquired. Data screening was performed by advanced automated processing, e.g., pattern recognition, cross-correlation, Fisher ratios, and source scanning. In addition, data were also screened manually by visual inspection of super-sonograms. In summary, we achieved a detection threshold of ML 0.5 for the whole region. The detected and located seismicity showed significant differences between regional clusters, and their vicinity to producing gas fields. Just three gas fields have neighbouring earthquakes at all: Walsrode, Rotenburg/Taaken and Söhlingen. The Walsrode cluster features ‘standard’ magnitude distribution of many small, and some larger events, while in the Rotenburg/Taaken and Söhlingen region just two earthquakes each with magnitudes ML 1 to 3 were recorded despite a much better monitoring sensitivity. Hypocenter depths were determined by an improved 1-D velocity model compensating for the observed misfit of S phases by constant vP/vS ratios. Since depths are in the range of gas reservoirs, these events must probably be associated with gas production as depletion-induced seismicity. Additionally, four deep-crustal earthquakes were identified indicating natural, intraplate tectonics eventually related to (post-) glacial isostatic adjustment. Precise location of the ML 1.5 Söhlingen event in February 2017 allowed for reprocessing the aftershock-sequence of the 2004 Rotenburg earthquake (ML 4.5) by waveform comparisons. Evaluation of S and converted phases indicate a split of aftershock series into two distinct depth ranges, one at 12 km close to the main shock, the other at 5 km within the reservoir, eventually indicating a triggered stress release at a secondary fault.weiterlesen