hide
Free keywords:
-
Abstract:
Integrated processing of high-rate GNSS and accelerometer data can overcome the disadvantage of each individual sensors and to increase the quality of derived co-seismic displacement. However, the contribution of accelerometer is usually underestimated by estimating baseline shifts epoch-wise which in fact happened very rarely. To take full advantage of both sensors, we propose a sliding window based Kalman filter to detect baseline shifts according to the disagreement of GNSS and accelerometer data and to estimate only the detected baseline shifts. The relationship of the window width, minimal detectable baseline shift and the displacement accuracy is investigated. The performance of the proposed approach is demonstrated by datasets collected during Samos Island earthquake (Mw 7.0, 30th October 2020). The results show that the baseline shifts in accelerometers can be precisely detected and estimated according to the very good agreement of the displacement integrated from accelerometer data after applying baseline shift corrections and that estimated from high-rate GNSS. Furthermore, the baseline corrected accelerometer data provides tight and reliable constraints on position and velocity to facilitate correct PPP ambiguity resolution. Thanks to the proposed approach, the complementary of GNSS and accelerometers is fully employed, consequently the co-seismic displacements of the tightly integrated processing are significantly improved.