High resolution deformation measurements at active volcanoes: a new remote sensing technology

It is known from observations at different volcanoes using ULP seismic observations that the volcanic edifice deforms slightly prior to an eruption. It can be expected that immediately prior to an eruption the largest deformation should occur in the vicinity of the vent. However, placing instruments at the vent is impossible as they will be destroyed during an eruption. Here we present new, high temporal resolution (up to 300Hz) deformation measurement that utilizes the phase information of a frequency modulated Doppler radar system. We decompose the Doppler signal into two parts, one part which allows us to measure speeds significantly above 0.5m/s (i.e. the movement of volcanic ash and clasts). The other part utilizes the slow phase changes of the signal reflected from non-moving objects, i.e. the volcanic edifice. This signal is used to measure very slow and longer term deformations, which are the main subject of this study. The method has been tested measuring the displacement of high rise buildings during strong winds. It can be shown that displacements down to 50 μm can be resolved without a problem. We apply this method to different data sets collected at Stromboli volcano, Italy, as well as Santiaguito volcano, Guatemala. At Stromboli we observed the NE crater once in 2008 and once in 2011. During both campaigns we observe on average a displacement between 1 and 5mm before different eruptions. This displacement can be interpreted as a widening of the conduit prior to an eruption. In a couple of cases even an oscillatory movement is observed with frequencies of about 0.5Hz. Finite element modeling of the rise of a pressurized slug indicates that deformations at the crater rim on the order of a 1mm or less are certainly reasonable. In the case of Santiaguito volcano prior to an eruption we observe a pre eruptive displacement 5-15mm and after the end of an eruption a displacement of up to 1m before the next eruption occurs. This can be interpreted as in initial uplift of the dome prior to an eruption and a general subsidence of the dome after an eruption. Very simple modeling of the subsidence of the dome indicates a viscosity of the underlying conduit material on the order of 6x107 Pa s. However both interpretations of the signals given above are certainly not unique as temperature and humidity along the beam affect the refractive index and therefore the propagation speed and wavelength. This is especially important in the case of long term measurements (days and weeks) due to the daily and seasonal variations in temperature. In addition changes in the index of refraction due the expulsion of hot moist gases prior to an eruption may affect the measurements.