ALBERT

Description: 〈span〉〈div〉ABSTRACT〈/div〉Interstation correlation is the basic operation in seismic noise and coda‐wave interferometry for signal extraction in imaging and monitoring applications. Conventional cross‐correlations evaluate the similarity between two signals along lag time, and they are efficiently computed by the fast Fourier transform (FFT), valuable to manage the large data volumes that ambient noise applications demand. The phase cross‐correlation (PCC) method contributes to increase convergence, a key issue in seismic ambient noise imaging and monitoring; however, it is much more computationally demanding. PCC evaluates similarity by subtracting the modulus of the sum and difference of the instantaneous phase of two signals. We introduce solutions to dramatically reduce the high‐computational cost of PCC. We show that PCC can be rewritten as a complex cross‐correlation and computed by the FFT when the moduli are raised to the power of 2, and we demonstrate PCC can improve waveform coherence and increase convergence compared with the default processing flow of 1‐bit amplitude normalization and standard cross‐correlation. Moreover, we develop a graphics processing unit implementation to accelerate computations when using powers other than 2 and particularly when using the power of 1. Finally, we extract Rayleigh‐ and body‐wave signals from many years of data from seismic stations distributed worldwide using PCC without a significant increase in computational cost compared with conventional cross‐correlation.〈/span〉

Description: 〈span〉〈div〉SUMMARY〈/div〉We use seismic noise cross-correlations to obtain a 3-D tomography model of 〈span〉SV〈/span〉-wave velocities beneath the western Indian Ocean, in the depth range of the oceanic crust and uppermost mantle. The study area covers 2000 × 2000 km〈sup〉2〈/sup〉 between Madagascar and the three spreading ridges of the Indian Ocean, centred on the volcanic hotspot of La Réunion. We use seismograms from 38 ocean bottom seismometers (OBSs) deployed by the RHUM-RUM project and 10 island stations on La Réunion, Madagascar, Mauritius, Rodrigues, and Tromelin. Phase cross-correlations are calculated for 1119 OBS-to-OBS, land-to-OBS, and land-to-land station pairs, and a phase-weighted stacking algorithm yields robust group velocity measurements in the period range of 3–50 s. We demonstrate that OBS correlations across large interstation distances of 〉2000 km are of sufficiently high quality for large-scale tomography of ocean basins. Many OBSs yielded similarly good group velocity measurements as land stations. Besides Rayleigh waves, the noise correlations contain a low-velocity wave type propagating at 0.8–1.5 km s〈sup〉−1〈/sup〉 over distances exceeding 1000 km, presumably Scholte waves travelling through seafloor sediments. The 100 highest-quality group velocity curves are selected for tomographic inversion at crustal and lithospheric depths. The inversion is executed jointly with a data set of longer-period, Rayleigh-wave phase and group velocity measurements from earthquakes, which had previously yielded a 3-D model of Indian Ocean lithosphere and asthenosphere. Robust resolution tests and plausible structural findings in the upper 30 km validate the use of noise-derived OBS correlations for adding crustal structure to earthquake-derived tomography of the oceanic mantle. Relative to crustal reference model CRUST1.0, our new shear-velocity model tends to enhance both slow and fast anomalies. It reveals slow anomalies at 20 km depth beneath La Réunion, Mauritius, Rodrigues Ridge, Madagascar Rise, and beneath the Central Indian spreading ridge. These structures can clearly be associated with increased crustal thickness and/or volcanic activity. Locally thickened crust beneath La Réunion and Mauritius is probably related to magmatic underplating by the hotspot. In addition, these islands are characterized by a thickened lithosphere that may reflect the depleted, dehydrated mantle regions from which the crustal melts where sourced. Our tomography model is available as electronic supplement.〈/span〉

Description: 〈span〉〈div〉Summary〈/div〉Primary microseism is the less studied seismic background vibration of the Earth. Evidence points to sources caused by ocean gravity waves coupling with the seafloor topography. As a result, these sources should be in water depth smaller than the wavelength of ocean waves. Using a state-of-the-art ocean wave model, we carry out the first global-scale seismic modeling of the vertical-component power spectral density of primary microseisms. Our modeling allows us to infer that the observed weak seasonality of primary microseisms in the southern hemisphere corresponds to a weak local seasonality of the sources. Moreover, a systematic analysis of the source regions that mostly contribute to each station reveals that stations on both the East and West sides of the North Atlantic Ocean are sensitive to frequency-dependent source regions. At low frequency (i.e., 0.05 Hz), the dominant source regions can be located thousands of kilometers away from the stations. This observation suggests that identifying the source regions of primary microseisms as the closest coasts can be misleading.〈/span〉

Description: 〈span〉〈div〉Summary〈/div〉We use seismic noise cross-correlations to obtain a 3-D tomography model of 〈span〉SV〈/span〉-wave velocities beneath the western Indian Ocean, in the depth range of the oceanic crust and uppermost mantle. The study area covers 2000×2000 km〈sup〉2〈/sup〉 between Madagascar and the three spreading ridges of the Indian Ocean, centred on the volcanic hotspot of La Réunion. We use seismograms from 38 ocean bottom seismometers (OBSs) deployed by the RHUM-RUM project and 10 island stations on La Réunion, Madagascar, Mauritius, Rodrigues, and Tromelin. Phase cross-correlations are calculated for 1119 OBS-to-OBS, land-to-OBS, and land-to-land station pairs, and a phase-weighted stacking algorithm yields robust group velocity measurements in the period range of 3-50 s. We demonstrate that OBS correlations across large interstation distances of 〉2000 km are of sufficiently high quality for large-scale tomography of ocean basins. Many OBSs yielded similarly good group velocity measurements as land stations. Besides Rayleigh waves, the noise correlations contain a low-velocity wave type propagating at 0.8-1.5 km/s over distances exceeding 1000 km, presumably Scholte waves travelling through seafloor sediments. The 100 highest-quality group velocity curves are selected for tomographic inversion at crustal and lithospheric depths. The inversion is executed jointly with a data set of longer-period, Rayleigh-wave phase and group velocity measurements from earthquakes, which had previously yielded a 3-D model of Indian Ocean lithosphere and asthenosphere. Robust resolution tests and plausible structural findings in the upper 30 km validate the use of noise-derived OBS correlations for adding crustal structure to earthquake-derived tomography of the oceanic mantle. Relative to crustal reference model CRUST1.0, our new shear-velocity model tends to enhance both slow and fast anomalies. It reveals slow anomalies at 20 km depth beneath La Réunion, Mauritius, Rodrigues Ridge, Madagascar Rise, and beneath the Central Indian spreading ridge. These structures can clearly be associated with increased crustal thickness and/or volcanic activity. Locally thickened crust beneath La Réunion and Mauritius is probably related to magmatic underplating by the hotspot. In addition, these islands are characterised by a thickened lithosphere that may reflect the depleted, dehydrated mantle regions from which the crustal melts where sourced. Our tomography model is available as electronic supplement.〈/span〉

Description: 〈span〉〈div〉SUMMARY〈/div〉Primary microseism is the less studied seismic background vibration of the Earth. Evidence points to sources caused by ocean gravity waves coupling with the seafloor topography. As a result, these sources should be in water depth smaller than the wavelength of ocean waves. Using a state-of-the-art ocean wave model, we carry out the first global-scale seismic modelling of the vertical-component power spectral density of primary microseisms. Our modelling allows us to infer that the observed weak seasonality of primary microseisms in the southern hemisphere corresponds to a weak local seasonality of the sources. Moreover, a systematic analysis of the source regions that mostly contribute to each station reveals that stations on both the east and west sides of the North Atlantic Ocean are sensitive to frequency-dependent source regions. At low frequency (i.e. 0.05 Hz), the dominant source regions can be located thousands of kilometres away from the stations. This observation suggests that identifying the source regions of primary microseisms at the closest coasts can be misleading.〈/span〉

Description: 〈span〉〈div〉ABSTRACT〈/div〉Interstation correlation is the basic operation in seismic noise and coda‐wave interferometry for signal extraction in imaging and monitoring applications. Conventional cross‐correlations evaluate the similarity between two signals along lag time, and they are efficiently computed by the fast Fourier transform (FFT), valuable to manage the large data volumes that ambient noise applications demand. The phase cross‐correlation (PCC) method contributes to increase convergence, a key issue in seismic ambient noise imaging and monitoring; however, it is much more computationally demanding. PCC evaluates similarity by subtracting the modulus of the sum and difference of the instantaneous phase of two signals. We introduce solutions to dramatically reduce the high‐computational cost of PCC. We show that PCC can be rewritten as a complex cross‐correlation and computed by the FFT when the moduli are raised to the power of 2, and we demonstrate PCC can improve waveform coherence and increase convergence compared with the default processing flow of 1‐bit amplitude normalization and standard cross‐correlation. Moreover, we develop a graphics processing unit implementation to accelerate computations when using powers other than 2 and particularly when using the power of 1. Finally, we extract Rayleigh‐ and body‐wave signals from many years of data from seismic stations distributed worldwide using PCC without a significant increase in computational cost compared with conventional cross‐correlation.〈/span〉

Description: In December 2018, the National Aeronautics and Space Administration (NASA) Interior exploration using Seismic Investigations, Geodesy and Heat Transport (InSight) mission deployed a seismometer on the surface of Mars. In preparation for the data analysis, in July 2017, the marsquake service initiated a blind test in which participants were asked to detect and characterize seismicity embedded in a one Earth year long synthetic data set of continuous waveforms. Synthetic data were computed for a single station, mimicking the streams that will be available from InSight as well as the expected tectonic and impact seismicity, and noise conditions on Mars (Clinton et al., 2017). In total, 84 teams from 20 countries registered for the blind test and 11 of them submitted their results in early 2018. The collection of documentations, methods, ideas, and codes submitted by the participants exceeds 100 pages. The teams proposed well established as well as novel methods to tackle the challenging target of building a global seismicity catalog using a single station. This article summarizes the performance of the teams and highlights the most successful contributions.