*3.2. Pore-Water Pressure Distribution at Sliding Surface Depth*

The characterization of pore-water pressure distribution during different seasons in correspondence of the typical depth of sliding surface was performed through the monitoring of data acquired at 1.2 m from ground level, in the test-site slope, in the period March 2012–December 2018. Averagely, the soil horizon kept in unsaturated conditions during the year, reaching minimum values during summer warm months (till −993 kPa), when strong evapotranspiration was not compensated due to limited rainfall amounts. During cold months, soil horizon re-wetted due to a more significant infiltration of rainwater and a more limited evapotranspiration. In these timespans, pore-water pressure grew to values close to 0 kPa, testifying conditions close to saturation in this soil level. In several periods during the cold time span of a year, pore-water pressure could reach positive values (till 12 kPa), especially when several rainfall events were spaced out by limited dry periods.

The distribution of the measured values of pore-water pressure (Table 3 and Figure 8) presents a certain degree of Gaussian trend, as confirmed by the values of the skewness very close to 0 and by the results of Shapiro–Wilk test, whose WS-W statistic did not allow us to reject the null hypothesis of gaussianity at 95% confidence level (WS-W = 0.92; *p*-value = 0.06). The first quartile of this distribution was equal to −846 kPa, while the third one was of −20 kPa.

**Table 3.** Main statistics of the distribution of the pore-water pressure values at typical depth of shallow-landslide sliding surface (1.0–1.2 m from the ground level) during the monitored time span (March 2012–December 2018) at the test-site slope: (Sd) standard deviation; (Median) median; (Min) minimum value; (Max) maximum value; (I quart) first quartile; (III quart) third quartile; (Skew) skewness; (WS-W) statistic of the Shapiro–Wilk test, applied to test the gaussianity of the distribution; (*p*-value) confidence level of the statistic of the Shapiro–Wilk test.


**Figure 8.** Histogram of distribution of the pore-water pressure values at typical depth of shallow-landslides sliding surface (1.0–1.2 m from the ground level) during the monitored time span (March 2012–December 2018) at the test-site slope.

Monitoring data allowed to exploit information on triggering events occurred in cold periods. Bordoni et al. [24] showed that during the observed event of 28 February–2 March 2014 at the test-site slope, pore-water pressure was about 0 kPa, at the beginning of the rainfall event which caused the shallow landslide triggering. Only this information is not enough to characterize exhaustively the antecedent hydrological conditions immediately before a rainfall able to provoke landslides in the study area. To analyze a higher range of soil hydrological conditions causing shallow landslides triggering and to test the effect of initial pore-water pressure on the definition of a threshold, physicallybased thresholds were then estimated by modeling the response of the soil to different rainfall events, starting

from an initial pore-water pressure condition of −20, −10, or 0 kPa. For simplicity, they are named TRIGRS/-20, TRIGRS/-10, and TRIGRS/0, respectively. In this way, a significant amount of the typical pore-water pressure values at depths of 1.0–1.2 m was considered in the definition of physicallybased threshold, as the third quartile of the measured values was in fact equal to −20 kPa.
