Stations that belong to National Observatory of Athens, Institute of Geodynamics (HL) (http://bbnet.gein.noa.gr/) network have been used for this study along with other HUSN stations from HA, HP and HT networks. Table S1 describes briefly sensor, vault and thermal insulation type at each site.
Table S1. HL Broadband Stations: Name, Location, Last sensor, Vault type, Insulation.
Trnkoczy, A., Bormann, P., Hanka, W., Holcomb, L. G., Nigbor, R. L., Shinohara, M., Suyehiro, K. and Shiobara, H. (2012). Site Selection, Preparation and Installation of Seismic Stations. In P. Bormann (Ed.), New Manual of Seismological Observatory Practice 2 (NMSOP-2) (pp. 1-139). Potsdam: Deutsches GeoForschungsZentrum GFZ. doi:10.2312/GFZ.NMSOP-2_ch7
Wielandt, E. (2012). Seismic Sensors and their Calibration. In P. Bormann (Ed.), New Manual of Seismological Observatory Practice 2 (NMSOP-2) (pp. 1-51). Potsdam: Deutsches GeoForschungsZentrum GFZ. doi:10.2312/GFZ.NMSOP-2_ch5.
We explore the correlation between significant wave height (SWH) data from 10 buoys deployed and operated by the Hellenic Center for Marine Research (Poseidon System) with average noise levels at the DF band for the same time period between 2007 and 2010 (Figure S1). We calculate for each buoy and seismic station pair the statistically significant correlation coefficient for these two independent variables using the function "corrcoef" in MATLAB with p-values in all cases much lower than 0.05.
Figure S1. Daily average PSD at the DF band (3-10 sec) versus daily average significant wave height (SWH) measurements from buoys located within the Aegean and the Ionian seas. The top two panels show seismic stations and buoys in the Aegean Sea and Crete, whereas the lower two panels show seismic stations located in mainland Greece and buoys located on both seas. Colored triangles on the inset maps are seismic station locations corresponding to the colored average PSDs. Similarly, colored stars are buoy locations corresponding to the SWH colored dots.
We explore, non-quantitatively, the origin of the SF microseism recorded across HUSN by focusing on extremely high noise days for this band. The highest noise levels for the examined four year period were recorded on 01/18/2009. SWH measurements show considerable wave heights in the North Atlantic on this date. Moreover, the SF level of noise across HUSN stations shows a clear increasing pattern from SE to NW towards the North Atlantic direction. On the same date the local weather was mild, with relatively low SWH in the Ionian and the Aegean seas. A similar trend of increasing noise from SE to NW is also observed during an eight day period (01/15/2009-01/23/2009), following the development of this significant meteorological event (Movie S1)). The level of SF microseismic noise is reduced at the end of this period. We also show two other dates that significant high noise levels are observed across this band; on 12/09/2009 (Figure S2) and 11/22/2009 (Figure S3). We observe the same noise pattern and considerable wave heights in the North Atlantic.
Movie S1 [h.264-encoded MP4 movie file; 1 MB]. SF noise for an eight day period from 15th to 22nd of January 2009. Top left: Western hemisphere validated SWH measurements from at least 4 satellite altimeter missions (Queffeulou, 2004). Top right: smooth SWH map from measurements of boys (stars) located within the Aegean and the Ionian seas. Bottom center: a smooth map of the SF level of noise (10-16 sec) from measurements of HUSN seismic stations on the same date (triangles). Areas in (top right) and (bottom center)) that are located further than 100 km from the nearest buoy or seismic station are masked out to avoid unrealistic oscillations and extrapolations.
Figure S2. Top left: Western hemisphere validated SWH measurements from at least 4 satellite altimeter missions (Queffeulou, 2004) on 22nd of November 2009. Top right: smooth SWH map from measurements of boys (stars) located within the Aegean and the Ionian seas. Bottom center: a smooth map of the SF level of noise (10-16 sec) from measurements of HUSN seismic stations on the same date (triangles). Areas in (top right) and (bottom center)) that are located further than 100 km from the nearest buoy or seismic station are masked out to avoid unrealistic oscillations and extrapolations.
Figure S3. Top left: Western hemisphere validated SWH measurements from at least 4 satellite altimeter missions (Queffeulou, 2004) on 9th of December 2009. Top right: smooth SWH map from measurements of boys (stars) located within the Aegean and the Ionian seas. Bottom center: a smooth map of the SF level of noise (10-16 sec) from measurements of HUSN seismic stations on the same date (triangles). Areas in (top right) and (bottom center)) that are located further than 100 km from the nearest buoy or seismic station are masked out to avoid unrealistic oscillations and extrapolations.
Queffeulou, P. (2004), Long term validation of wave height measurements from altimeters, Marine Geodesy, 27, 495–510, doi:10.1080/01490410490883478
This noise model for Greece is based on the statistical mode of the PDF noise levels of all HUSN stations (Figure S4). It is calculated in a similar procedure with McNamara and Bulland (2004), who computed a new noise model for the continental U.S.A. The statistical mode represents the highest probability noise level at each given station. The minimum value of all station modes per octave is the new representative noise model for Greece (HMLNM) (Figure S5 and Table S2).
Figure S4. Spatial distribution of each seismic sensor used in this study.
Figure S5. HUSN PDF mode noise model (HMLNM, red line) calculated from the minimum of all station PDF mode noise levels. The corresponding PDF mode noise level for continental U.S.A. (UMLNM), as is estimated by McNamara and Bulland (2004), is also plotted (blue dashed line). The gray shadowed area marks the area between the minimum 10th and 90th percentiles of all HUSN station PSD distributions, representing the 80% confidence interval of the minimum noise levels in Greece. Thin solid lines represent individual modes for each HUSN station.
Table S2. HUSN mode noise model (HMLNM).
McNamara, D. E., and R. P. Bulland (2004), Ambient Noise Levels in the Continental United States, Bull. Seismol. Soc. Am., 94 (4), 1517–1527, doi:10.1785/012003001
[ Back ]
