Page 404 - Caldera Volcanism Analysis, Modelling and Response

A New Uplift Episode at Campi Flegrei Caldera (Southern Italy) 379

2. Recent Ground Deformation Data at Campi Flegrei
Caldera

The recent vertical ground deformation at Campi Flegrei has been measured by
precision leveling since 1969, and by GPS since 2000. The present leveling network is
made by 330 benchmarks over a total length of about 135km in 14 circuits; in
addition, nine continuous GPS stations operate in the area. The location of GPS
benchmarks (Figure 1) has been determined taking into account previous evidence in
the area, which shows remarkably constant displacement patterns (De Natale et al.,
1997). Such consistency allowed the location of station RITE to be chosen as the
expected point of maximum vertical displacement, whereas ACAE, ARFE, and
BAIA should give a good sampling of the range of expected maximum horizontal
displacement; the remaining points are meant to sample larger distances. During the
whole period of CGPS operation, the maximum horizontal displacement has always
been recorded at ACAE, thus confirming the high stability of the source axis. Figure 3
summarizes vertical displacements measured from precise leveling in the campaigns
from May 2004 to September 2006 at the three common and most significant
subnetworks (Solfatara circuit, Coast line and South–North line). The leveling
measurements have been carried out using automatic optic levels Leica NA2 with
parallel-plate micrometer (0.01mm resolution) and using invar rods, calibrated before
the survey. The double-run field procedures were utilized, with a maximum closure
between forward and backward running of each section of less than 2.5 (L km ) 1/2 mm,
where L represents the length of the leg in kilometers. For leveling loops, it has
been assumed that the closure error resulted in less than 2.0 (C km ) 1/2 mm, where C
represents the length (in kilometers) of the circuit. After such control, the observed
height differences have been adjusted by the least squares method, deriving for every
benchmark the most probable quota and its standard error.
In May 2004, a measurement of the whole leveling network was performed. The
field measurements have been compensated by least squares, showing a standard
deviation per unit weight s 0 =1.02 mm, confirming good quality data (Pingue et al.,
2006). The compensated heights have been referred to the benchmark located in
Naples-Mergellina, also shown in Figure 1. Compared to the May and November
2003 data, the May 2004 data show that very minor subsidence (less than 1 cm/year,
almost within the errors) is still present in the area (Del Gaudio et al., 2005). In
November 2004 another measurement was performed, but only on part of the
network (circuits in cyan, Figure 1). This campaign still shows very slight subsidence,
with a maximum of less than 1 cm at Solfatara crater, which is very close to the 2s
error limits (Pingue et al., 2006). Two further field measurement campaigns were
performed in 2005 (one each in May and October) showing a slight uplift with
respect to the 2004 measurements (maximum uplift of 1 cm in Pozzuoli harbor),
although still in the limits of 2s errors. A new leveling measurement was conducted
in March 2006, with a s 0 =0.63 mm, showing a clear uplift of up to 2 cm at Pozzuoli
harbor with respect to May 2004 (yellow curves in Figure 3). The last partial leveling
survey was conducted in September 2006 (with a standard deviation per unit weight
s 0 =0.40 mm), which has underlined an uplift increase of almost 4 cm (green curve in