*3.4. Nivological Analysis*

The local nivological analysis was based on a detailed dataset manually collected at the nivo-meteorological station of the Meteomont service (https://www.sian.it/infoMeteo, accessed on 15 February 2021). It is located at the base of the slope in the Prati di Tivo area (Figure 3), at an elevation of 1450 m a.s.l. It features a northern exposure similar to avalanche-prone regions located at higher elevations. The available historical data for this station begins from the 1977/1978 winter season (from November to April) for 32 surveying seasons. The series is nearly uninterrupted, with a few gaps mostly occurring in correspondence of the beginning/end of seasons. However, data related to the 1992/1994 seasons are completely missing. More in detail, the considered dataset shows several temporal gaps, since it is deeply affected by the irregularity in the opening/closing dates of ski facilities—the former occurring after the first significant snowfall events and the latter during the spring period, usually in the presence of a thick snow cover. This condition was widely relevant before the 1986/1987 winter season and after the 2008/2009 one; consequently, the amounts of seasonal new snow were not correctly computed in these temporal intervals. To reduce this underestimation, we tried to derive good-quality data about the potential snowfall events by computing thermo–pluviometric records at gauges located at a comparable elevation not far from the Prati di Tivo area (e.g., Campotosto gauge, 1344 m a.s.l.—yellow dot in Figure 4). Nevertheless, to deduce a general nivometric trend and better define the nivometric regime of the study area, data belonging to a 20-year time period (1986/1987–2008/2009) were considered and thoroughly analyzed.

#### *3.5. Snow Avalanche Hazard Assessment*

This analysis was performed following a stepwise methodological approach that involved the snow avalanche inventory analysis, the analysis and mapping of snow avalanches' paths, the elaboration of a snow avalanche hazard map, and the definition of numerical models.

The snow avalanche inventory was retrieved from the State Forestry Corps of Italy and the Abruzzo Region (http://opendata.regione.abruzzo.it/content/carta-storica-dellavalanghe, accessed on 15 May 2021) and allowed us to clearly describe the avalanches' spatial distribution over the study area. Moreover, it was integrated with information derived from the available literature and technical reports [44,75,99].

The analysis of snow avalanches' paths was achieved by combining the literature data, specific site investigations, investigations of the snow-covered ground, interviews of witnesses to past avalanche events, and studying of previous events recorded in various historical and technical archives [44,100].

The evaluation of the snow avalanche hazard map was carried out according to the Swiss mapping criteria [101,102] and thematic guidelines provided by AINEVA (Italian lnterregional Association for Snow and Avalanche) [31,103]. Avalanche-exposed zones were defined and annexed within the Avalanche Hazard Exposure Zones Plan—PZEV (Piano delle Zone Esposte a Valanghe in Italian). Generally, this evaluation is fixed through mathematical parameters, which quantified the velocity and flow height, transmitted pressures, and stopping distances of the avalanches [31,102,104,105]. In the invasion zones, as reported in Table 1, some areas are identified and marked with different colors according

to the estimated avalanche hazard—i.e., high hazard with red, moderate hazard with blue, and low hazard with yellow. Town planning and land use prescriptions are fixed for each of the identified zones.

**Table 1.** Synthesis of the AINEVA criteria [31] for the delimitation and the use of areas with different degrees of exposure to avalanche hazards (T = return time of the avalanche (years) and Pimp = impact pressure (kPa)).


In the PZEV's framework, morphometric and nivometric data are generally combined to define the degree of exposure of a specific area in terms of the frequency and intensity of avalanche events. This detailed analysis is usually expressed through:


The obtained maps effectively identify the avalanche sites and their expansion in the accumulation zones. This has proven to be most helpful in defining these zones in terms of avalanche frequency and dynamic pressure, thus determining the magnitude/frequency distribution in the runout zones [106–108].

The criteria established and reported in the Avalanche Artificial Detachment Intervention Plan—PIDAV (Piano di Intervento di Distacco Artificiale di Valanghe in Italian) [109] were followed to develop prevention and managemen<sup>t</sup> activities in the study area. Generally, the main objective of these protection measures is to minimize negative consequences due to snow avalanche risk for people and goods in their settlements and along traffic lines, as well as for skiers [32,110]. The PIDAV plan is a tool, eventually complementary to the aforementioned PZEV, which refers to an area open to the public, clearly defined in space and time, where an artificial release of unstable snow masses is performed to reduce avalanche hazards and risks [109,111]. In case of an urban zone or a ski facility to be protected, as in the study area, it is necessary to define a managemen<sup>t</sup> measures plan to protect the ski lift. It should include the plan for meteo-nivological conditions monitoring—which are in constant evolution during climatic events—and describes activities to be exerted to learn about this evolution at the meso- and microscale to evaluate snow cover stability conditions and their potential evolution.

In conclusion, this stepwise sequence was completed by avalanche simulation models. In detail, 1- and 2-dimensional avalanche simulation models (e.g., AVAL-1D and RAMMS [39,112]) were applied both to back-analyze documented avalanche events at a particular site, as well as to estimate the consequences of possible hazard scenarios. According to the literature and technical data [102,111,113], the main nivometric parameters required for dynamic avalanche modeling are represented by the maximum height of the

snow cover (Hs) and the increase of the snow cover height over three consecutive days (Dh3gg). For developing the present study, an increase of 5 cm of new snow and a snow cover for every 100 m of elevation was proposed, taking into account the aforementioned literature data and the nivological expert judgment. These physical–mechanical characteristics, together with ancillary information concerning the physiography, steepness, and roughness of the ground, the presence of infrastructures s.l. were reported on a 5-m grid DTM base map and elaborated in a GIS environment.

AVAL-1D is a numerical avalanche dynamics program developed by the Swiss Federal Institute for Snow and Avalanche Research [112]. It allows the simulation of avalanches in one dimension from the starting zone to the runout one. It reproduces runout distances, flow velocities, and impact pressures of both flowing and powder snow avalanches along a specified avalanche track. It consists of two modules: FL-1D (dense flow model) for dense flow avalanches and SL-1D (powder snow model) for powder snow avalanches. It cannot reproduce the whole set of dynamical parameters, since it is a one-dimensional formulation that combines the internal distribution of flowing variables into basic ones controlled by two frictional parameters [114]. In order to supply this not accurately modeling, the RAMMS (RAapid Mass MovementS) code [115] was mainly used to calculate the pressure values on a site-specific avalanche path (such as Vallone della Giumenta) from initiation to runout in a three-dimensional terrain. It is a practical tool for avalanche practitioners, which requires a complete procedure to fulfill the morphological features and release parameters. Moreover, it can be used to estimate runout distances, flow velocities, flow heights, and impact forces [116–118].
