Spatial distribution of natural debris-flow impact

. The high destructive potential of debris flows poses a challenge for the design of structural mitigation measures to protect vulnerable areas. An essential part of designing such structures is to determine the magnitude and spatial distribution of the impact forces. The impact is expected to be related to the composition of the flow, which can vary from clay to large boulders along a flow event and between events. Experimental studies are prone to scaling bias and benchmark observations at the full scale are rare so far. Here we present measurements of the temporal and spatial variations of the impact of a natural debris flow in the Gadria creek, IT, onto an instrumented barrier structure. The flow event was preceded by a precursory surge, which was then followed by the debris flow with multiple surges. The flow height reached up to 2.3 m. We found that the impact of boulders occurs primarily in the upper half of the flow profile, the highest forces were measured in first part of the flow, and deposition has a strong influence on the lower part of the structure by damping and/or redirection.


Introduction
Debris flows exhibit the ability to transport large boulders of several meters, trapped in a matrix with a high concentration of solids that can shape steep fronts with high velocities that pose severe potential for destruction of settlements and infrastructure. The interaction of this mass wasting process with mitigation measures is a fundamental problem for the design and a key aspect for the functional reliability of mitigation structures. Measuring events on a real scale is a challenge for science. Only some efforts were done in the past to quantify these parameters in real scale [1][2][3][4]. Due to the low data availability of natural events, most investigations to enhance our understanding of the impact process are based on laboratory experiments [e.g. [5][6][7]. Field measurements offer a complementary method with no scaling difficulties and provide quantitative observations of flow dynamics and interaction. Considering that debris flows show a great heterogeneity in their temporal and spatial evolution [8][9][10], this is particularly evident for impact forces.
Impact forces of debris flows are often divided into two different contributions depending on the duration of impact (see Fig. 1). The impact of individual boulders exhibits a high contact force in a short period of time, while the bulk pressure of the mixture exerts a more uniform force on the obstacle over a longer period of time [2]. For the planning of mitigation measures, it is * Corresponding author: georg.nagl@boku.ac.at necessary to determine the magnitude, location and duration of the impact forces exerted onto the structure. In this contribution, we will present in-situ measurements of impact forces of a natural debris and we show the temporal and spatial distribution of impact forces of a natural debris-flow on a monitoring structure. The full-scale test site is situated at the Gadria creek (Northern Italy), draining a catchment area of about 6.3 km² with an altitude difference from 2,950 m a.s.l. to 1,394 m a.s.l. at the fan apex. The steep headwater catchments, metamorphic rock and a high degree of fragmentation and glacial deposits provide the basis for frequent rainfall-triggered debris flows in the Gadria catchement [11,12]. The channel has a trapezoidal form with a fixed channel bed of a longitudinal slope of 6° and a rock-riprap channel bank. The channel bed is 40 m upstream and 10 m downstream from the barrier secured with concrete riprap to prevent erosion during a debrisflow event. The monitoring structure represent a single part of a structure used in alpine regions to mitigate mass wasting processes, as debris flows. The barrier structure is one element of the construction system and consists of a concrete element of 1 m width vertical orientated on the center of the channel (see Fig.  2) and 3 m above ground, connected to a foundation below the channel bed. Fourteen load cells (HBM C6A from 1 to 2 MN capacity) are installed on the front of the structure to measure the spatial distribution of impact. Each sensor is protected by a steel plate with a circular area of 0.2 m in diameter. Upstream of the barrier, a traverse check dam unconnected to the wall is flush to the ground and builds up the second part of the system with two weighing systems, one infront if the barrier and und aside.

Analysis
For the analysis of the impact force, the time series were processed as follows:  We processed the data with a wave-let analysis by using a Daubechie function (db4) and decomposed into 8 levelswith a soft threshold to eliminate random noise [3,13]. For the load cells, we found no measureable resonance frequency.
 The impact force of debris flows can be divided into two different forms of impact depending on the duration of impact. In order to obtain bulk force we separate boulder impact forces and bulk pressure by calculating a binned maximum and median value over one second [14].

Results and interpretation
Stage hydrographs recorded directly in front of the barrier (sensor 2) and next to the barrier (sensor 1) illustrate not only the interesting flow dynamics of individual events, but also reflect the interaction of the barrier with the flow and the complex erosiondeposition pattern during the events.
On August 2 nd , 2020, the debris flows was triggered by intensive rainstorm. After increasing discharge, a sediment-laden precursory surge (debris flood) hit the barrier at 19:29:00 UTC (307 sec. on Fig.4). The flow developed with moderate velocity. A big boulder deposited in front of the barrier at 19:30:02 UTC (413 sec.), and blocked the flow height sensor two (Fig. 3a). . After this surge, the flow height decreased continuously. Some minor surges could not remove the boulder in front of the barrier and deposited more debris in the area around the barrier. At 19:51:18 UTC, (1700 sec.) a surge with 2.2 m flow height re-activated the deposited material and removed all boulders in front the barrier (see Fig. 3b). Afterwards several surges hit the barrier. During this period, the velocity approached zero in between surges and the material stopped at rest with flow heights of 1 to 0.8 m and was reactivated by the following surges (see Fig 4). To illustrate the temporal and spatial impact behaviour, we calculated the relative impact force as median binned value of force /maximum binned value for each second and compared this with the relative height defined as the height of each sensor normalized by the flow height, in Fig.5. The symbol size represents the absolute magnitude of the force and the colour (from blue to yellow) represents the time. Hence, if the point is close to 1, the maximum value and the mean value are almost the same. This represents a more fluid like impact. Instead, if the relative impact force tents to zero, the maximum and mean value is far apart, which we interpret as a bouldery impact. We found the highest relative impact forces were concentrated in the upper regions of the flow. This may be connected to a wedge of deposited material that damped the impact of single boulders. Additionally, the highest values were found in the first phase of the debris flow, which is consistent with the arrival of a boulder front.

Conclusions
Our in-situ measurements demonstrate the temporal and spatial distribution of impact forces of natural debris flows, which can subsequently be useful for appropriate design.