slow TrAnsienT recorded by The cGPs fredneT neTwork AT …In the last years the attention to GPS measurements increased enormously, for the capability of reconstructing plate boundaries

slow TrAnsienT recorded by The cGPs fredneT neTwork AT The AdriA norThern TiP (ne iTAly)G. rossi, d. zuliani, P. fabrisOGS (Istituto nazionale di Oceanografia e di Geofisica Sperimentale), Trieste, Italy

Introduction. In the last years the attention to GPS measurements increased enormously, for the capability of reconstructing plate boundaries and their movements, as well as faults activity at a more local scale. This implied also a continuous effort in increasing the reliability and accuracy of the measurements and of the calculus of the relative linear trends. Nowadays, having at disposal time series long enough (more than ten years) it is possible to begin to distinguish other terms that are superimposed to the linear trend, and hence to the rigid plate motion. In fact, Dong et al. (2002) and Blewitt and Lavallee (2002) identified annual and seasonal effects, while Smalley et al. (2005), Calais et al. (2005) evidenced some characters of the North-American plate movements that differ from the simple linear motion, and the interpretation of which is vividly debated. Also in our regions, Devoti et al. (2008; 2011) evidenced slow varying patterns in Italy at long-scale length (> 100 km) that cannot be explained by simple block models, and Nocquet (2012) interprets the observed velocity field in the whole Mediterranean region as a combination of localized and distributed deformation.

These considerations are at the basis of the research here presented. Having a disposal a data set of continuous GPS measurements, of decadal average length, in an area of intense and complex tectonic phenomena, we tried to inquire whether deviations from the linear trend are present, and the possible physical origin.

The data. The northern tip of the Adria microplate (NE-Italy) is continuously monitored by the Friuli Regional Deformation Network (FReDNet) of OGS (Istituto Nazionale di Oceanografia e Geofisica Sperimentale), consisting of 15 GPS/GNNS stations, the first 8 of which were installed between 2002 and 2004. The 10 stations of the Marussi network of the Friuli-Venezia Giulia regional council, some of which record continuously since 1999, provide additional information on the strain field in the region.

Fig. 1 – Vectors of the yearly horizontal velocities for the sites of the two networks (yellow, FReDNet stations, green, Marussi network stations), calculated from the linear trend. Dashed: the vectors calculated on the shortest time series.

198

GNGTS 2013 SeSSione 1.2

131218 - OGS.Atti.32_vol.1.30.indd 198 01/11/13 08:41

We considered the GPS data of the longest time series from both networks, starting from 2002 till present. The resulting data set is composed by the data of: ZOUF, AFAL, ACOM, MPRA, UDI1, MDEA, TRIE from FReDNet, and AMP0, MOG0, POR0, PAL0 and TRI0 from the Marussi network.

We processed the data using GAMIT/GLOBK, eliminated the outliers, and filled the eventual short gaps in the data through linear interpolation. A strong annual component is present, as expected, due to seasonal variations of hydro-meteorological parameters (e.g., Blewitt and Lavallee, 2002). We tested, however, whether this annual term is affected by the so-called draconitic term, i.e. the time between two passages of the object through its ascending node, or the point of its orbit where it crosses the ecliptic from the southern to the northern hemisphere. From an accurate spectral analysis of the time series, it resulted that only a couple of station show, on one of the components, relevant amplitude associated to the draconitic term. A low-band pass filter allowed obtaining the time-series cleaned from the components with frequencies higher than 1.5 year, so to eliminate the annual and quasi-annual terms, and the highest frequencies. The so-obtained time-series for the two horizontal components are dominated by a linear trend, as expected, to which clear oscillations of apparent period of a few years (2-4) are superimposed. Oscillations are present also in the vertical component.

The data are shown in European plate reference frame, obtained from ITRF08 after rotation using the Euler vectors of Altamimi et al. (2012). The resulting velocity field from the analysis of the linear trend suggest crustal shortening, with values ranging between 0.6 and 2.8 mm/year, decreasing from South to North and, less pronounced, from East to West. This is in agreement with preceding observations and with the geodynamic character of the region, located in the area of convergence between Adria microplate and Eurasia (Fig. 1).

Deviations from the linear trend: characteristics and analysis. A said above, in the various sites, superimposed on the linear trend, 2-4 year period oscillations are present, and show higher amplitude with respect to the annual terms (Fig. 2).

To better analyse the oscillations, we first applied a band-pass filter (1.5–3 years) to the data, and then calculated the signal component along directions, spaced 15°, from N to N165E. This procedure allowed to evidence a sort of transient, of “period” of roughly 2.0 years, causing a bending in all the stations considered, distributed over the whole region, mainly along a direction about coincident with the Dinaric trend, N120E. Only in a few cases, the transient causes a bending toward an anti-dinaric direction which is also a dominant tectonic direction in the region (e.g. Bressan et al., 2003, 2007). Fig. 3 shows the curves relative to the signal along the direction in which the maximum is recorded at each station, as well as the correspondent vertical GPS signal.

Fig. 2 – Three of the amplitude spectra of the longest time series from the FReDNet and Marussi networks: the numbers on the peaks are indicative of the periods in years and fractions of year.

199

GNGTS 2013 SeSSione 1.2

131218 - OGS.Atti.32_vol.1.30.indd 199 01/11/13 08:41

As it may be seen, the transient causes positive displacement up-wards and along the directions indicated in the legend between 2007 and 2008, and opposite trend in 2009.

In order to state, whether the transient is due to tectonic phenomena or, vice versa, has an hydrologic origin, as found by Zerbini et al. (2010) in a neighboring area, we calculated the seismic energy released in the region in the same periods (e.g. Franceschina et al., 2006), as well as the hydrological balance.

The first is defined as:

where E is the seismic energy and MD is the earthquake duration magnitude.For the calculus of the hydrological balance, we started from the meteorological stations

of the regional council networks nearest to each of the GPS stations and corrected the de-trended cumulative curves for the estimated evapo-transpiration, using the Thornthwaite (1948) formula:

where

and

with PET = estimated potential evapotranspiration (mm/month), Ta is the average daily temperature, N is the number of days in the month, L is the average length of the day, and l is the heat index. We compared GPS displacements, seismic energy, and hydrological balance in time, at time interval of about two months, starting from 2005 to 2010. There is a certain correlation between the time variations of the seismic energy distribution in the region and the ones of the deformation field induced by the transient. On the contrary, the variations in time of the hydrological balance distribution appear less in agreement, varying more slowly.

Fig. 3 – Solid lines: horizontal displacements along the directions shown in the legend box, along which the displacement is maximum for each station; dashed lines: vertical components. Constant values are applied to the curves to enable the comparison between the various sites. Grey dashed, vertical lines indicate the years.

200

GNGTS 2013 SeSSione 1.2

131218 - OGS.Atti.32_vol.1.30.indd 200 01/11/13 08:41

Discussion and conclusions. In other parts of the world GPS networks recorded creep phenomena or silent or slow earthquakes (rupturing over a period of hours or days instead of seconds), in some cases accompanied by tremors, as observed, e.g., by Dragert et al. (2001), Lowry et al. (2001). In the region under study, after the tremors recorded by the horizontal pendulums of Grotta Gigante in the three years preceding the Friuli earthquake (Chiaruttini and Zadro, 1976), interpreted a-posteriori as slow earthquakes (Dragoni et al., 1984/5), tiltme-ters and strainmeters evidenced the presence of quasi-periodic signals of longer-term (9 years and 33 years) with respect to the ones here analysed, and acting along directions about N70°E and N-S oriented respectively, hence, normal to the two main tectonic systems present in the region (the Alpine and the Dinaric ones) (Rossi and Zadro, 1996; Braitenberg and Zadro, 1999).

The present study focused on short period variations, recognizable on all the time-series considered, and causing bending from the linear trend mainly in the Dinaric direction. The data were compared both with the seismic energy released in the region and with the distribution of the hydrological balance, as representing possible hydrological effect. The results of the comparison are slightly more in favour of a tectonic origin of the transient, but to come to a definitive conclusion requires more analyses in the next future. referencesAltamimi, Z., L. Metivier, and X. Collilieux; 2012): ITRF2008 plate motion model, J. Geophys. Res., 117, B07402,

doi:10.1029/2011JB008930.Blewitt, G., and Lavallee D.; 2002: Effect of annual signals on geodetic velocity, J. Geophys. Res., 107(B7),

10.1029/2001JB000570.Braitenberg C. and Zadro M.; 1999: The Grotta Gigante horizontal pendulums – instrumentation and observations. Boll.

Geof. Teor. Appl., 40, 3/4, 577-582.Bressan G., Bragato P.L. and Venturini C.; 2003: Stress and Strain Tensors Based on Focal Mechanisms in the

Seismotectonic Framework of the Friuli–Venezia Giulia Region (Northeastern Italy), Bulletin of the Seismological Society of America, 93, 3, 1280–1297.

Bressan G., Kravanja S., Franceschina G.; 2007: Source parameters and stress release of seismic sequences occurred in the Friuli-Venezia Giulia (Northeastern Italy) and in Western Slovenia, Phys. Earth Planet. Inter., 160, 192-214.

Calais, E., Mattioli G., DeMets C., Nocquet J.-M., Stein S., Newman A., Rydelek P.; 2005: Tectonic strain in plate interiors? Nature 438, doi: 10.1038/nature04428.

Chiaruttini C., Zadro M.;1976: Horizontal pendulum observations at Trieste. Udine Meeting on the Friuli Earthquake, 1976. Boll. di Geof. Teorica ed Applicata, Vol.XIX, n.72, 441-455.

Devoti R., Riguzzi F., Cuffaro M. and Doglioni C.; 2008: New GPS constraints on the kinematics of the Apennines subduction. Earth Planet. Sci. Lett. 273, 163-174, doi:10.1016/j.epsl.2008.06.031.

Devoti R., Riguzzi F., Cuffaro M. and Doglioni, C.; 2011: New GPS constraints on the kinematics of the Apennines subduction, Earth and Planetary Science Letters, 273, 163-174.

Dong, D., Fang P., Bock Y., Cheng M. K., and Miyazaki S.; 2002: Anatomy of apparent seasonal variations from GPS-derived site position time series, J. Geophys. Res., 107(B4), doi:10.1029/2001JB000573.

Dragert, H., Wang, K., James T.S.; 2001: A Silent Slip Event on the Deeper Cascadia Subduction Interface, Science, 292, 1525-1528, DOI: 10.1126/science.1060152.

Dragoni M., Bonafede M., Boschi E.; 1984/85: On the interpretation of slow ground deformation precursory to the 1976 Friuli earthquake, Pure Appl. Geophys., 122, 781-792.

Franceschina, G., Kravanja S.and Bressan G.; 2006: Source parameters and scaling relationships in the Friuli-Venezia Giulia (Northeastern Italy) region, Phys. Earth. Planet. Int., 154, 2, pp. 148-167.

Nocquet J-M.; 2012: Present-day kinematics of the Mediterranean: A comprehensive overview of GPS results, Tectonophysics, 579, 220-242.

Rossi G. and Zadro M.; 1996: Long-term crustal deformations in NE-Italy revealed by tilt-strain gauges, Phys. of the Earth and Plan. Int., 97, 55-70.

Smalley, R. Jr, Ellis M. A., Paul J. and Van Arsdale R. B.: 2005: Space geodetic evidence for rapid strain rates in the New Madrid region, central USA, Nature 435, pp. 1088-1090.

Thornthwaite, C.W.; 1948: An Approach toward a Rational Classification of Climate, Geogr. Rev., 38, 1., pp. 55-94.Zerbini, S., Raicich F., Richter B., Gorini V. and Errico M.; 2010: Hydrological signals in height and gravity in Northeastern

Italy inferred from Principal Components analysis, J. Geodyn., 49, 3-4, pp. 190-204.

201

GNGTS 2013 SeSSione 1.2

131218 - OGS.Atti.32_vol.1.30.indd 201 01/11/13 08:41