F. Bettella1, T. Michelini1, V. D’Agostino1 and G. B. Bischetti2
1Department TESAF, University of Padova, Viale dell'Università 16, 35020 Legnaro, Padova, Italy 2Department of Agricultural Engineering, Università degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy
Abstract
The runout prediction is a key factor in debris-flows hazard assessment. Despite its importance, few literature studies can be found about the forest capacity in reducing debris-flows mobility in the depositional zones. For this reason, the aim of this research is to explore the forest capacity to reduce debris-flows runout through small-scale experiments, in order to provide guidelines for the management of the protection forests. The laboratory investigations were conducted using the flume of the Department of Agricultural Engineering (University of Milan). A geometric scale of 1:50 was defined for the laboratory model construction. Three scenarios were simulated: no elements (i.e. without vegetation), high forest, and coppice. High forest was simulated using rigid steel screws, while coppice was simulated with flexible wicker. Elements on the depositional plane caused changes in the deposit shapes; in general, shorter runouts, and larger deposit widths were measured. The observation of debris-flow mobility suggests that high forest management is not able to offer a very significant contribution in reduction of debris flow motion for the highest sediment concentrations; instead, coppice seems able to offer a remarkable contribution, which increases for the debris flow with highest solid concentration.
Keywords: debris flow; forest protection; small-scale experiment; forest management. Introduction
The prediction of runout characteristics (including runout distance, area, and width) is a key factor for hazard assessment in mountain areas (e.g. Fell et al., 2008). It is widely recognized that the presence of forest can prevent slope failures and debris-flow generation for events of small/medium size by reinforcing soil through root strength (e.g. Nonoda et al., 1994), suppressing debris-flow movement, and promoting debris flow deposition by the resistance provided by trunks and branches (e.g. Mizuyama et al., 1989). The ratio between the height lost and the distance travelled by the centre of mass of landslides (H/L) is often used as an indicator of debris flows mobility in empirical methods, which have been developed to predict runout and inundation areas (e.g. Rickenmann, 2005). The capacity of forest to suppress sediment movement has been studied by some Authors trough field survey (e.g. Ishikawa et al., 2003), numerical models (e.g. Irasawa et al., 1991), and laboratory investigations (Morlotti, 2010). There are, instead, few literature studies about the aptitude of forest buffer to reduce debris-flows movement on the fan.
The aim of this research is to explore the ability of the forest to reduce debris-flows mobility through small-scale laboratory experiments, in order to provide guidelines on the protection forests management. Materials and Methods
A series of laboratory experiments were performed using a small-scale flume 200 cm long, 15 cm wide, and 40 cm deep, which can be set at different slopes (15° to 45°). At the upper end of the flume, there is a tank for the mixture accumulation, whereas a runout area with an adjustable inclination (0° to 10°) is located at the toe of the experimental channel for the observations of the deposit characteristics. On the depositional plane,
some threaded holes (10 cm grid spaced) allow placing vertical elements to simulate the presence of the trees. The experimental mixtures were obtained mixing water with a dry solid material sampled in real debris flow deposits, which occurred along Gadria creek (Silandro, Bz) in 2014, sieved at 19 mm. In each experiment the granular material was weighed, and mixed with water to obtain 4000 cm3 of mixture. The inclination of the flume was set equal to 20° for the channel (Sc), and equal to 3° for the depositional area (Sd); these
parameters were kept constant in each test. The interaction between the flow in motion and the elements were examined using mixtures with four different sediment volume concentrations (CV): 0.50, 0.55, 0.60, and
0.65. Three different scenarios have been finally set in order to explore the forest management effects on debris-flow depositional behaviour: 1) free depositional plan (i.e. without elements), 2) presence of rigid elements on the depositional plane to simulate high forest, 3) presence of groups of flexible elements on the depositional plane to simulate coppice. A geometric scale was defined for the laboratory model with a scale factor of approximately 1:50. The high forest scenario was simulated using rigid steel screws 10 cm long and with a diameter of 0.6 cm. Aiming to model the shoot flexibility, the coppice scenario was simulated using for each stump 8 pieces of wicker 6-8 cm long, and of 0.2 cm in diameter. The elements allowed to simulate real forests with a density of 400 stands per hectare, with stand diameter of 0.30 m for the high forest case (basal area equal to 28 m2ha-1) and 0.10 m for coppice (basal area equal to 25 m2ha-1). At the end of each test, the
following geometric characteristics of the deposit have been collected: a) maximum length reached by the debris flow (runout, R); b) maximum width (W); c) maximum deposit thickness (smax); d) weight of deposited
material (Pdep); e) weight of remained material in the tank (Pret).
Results
A total of 29 tests were carried out and the geometric characteristics of the deposit were measured. Runout (R) ranged from 37.0 to 119.0 cm, maximum width (W) ranged from 34.0 to 52.0 cm, maximum deposit thickness (smax) from 0.60 to 3.30 cm, deposited area (A) from 1450 to 4730 cm2 and travel angle (equal to the
arctangent of H/L ratio) from 15.40° to 19.56°. As shown by the box and whiskers plots (Figure 1) the main geometric features change over the three scenarios. R decreases from the no elements condition, to high forest, to coppice scenarios, while smax and H/L show an increasing trend. W, instead, seems to be not
affected by the presence of elements on the depositional plane.
Figure 1: box plots of the geometric features for the three scenarios (no elements, high forest and coppice; samples size 13, 8, and 8
respectively): runout R, maximum width W, maximum deposit thickness smax, and travel angle (equal to the arc tangent of the ratio H/L).
The reduction of R, in the cases of the presence of trees respect to the no elements scenario, ranged from 15.0% (CV = 0.50) to 2.2 % (CV = 0.65) in high forest scenario, and from 13.9% (CV = 0.50) to 16.5% (CV =
0.65) in coppice scenario. The increase of the H/L ratio was analysed in terms of its excess passing from presence of elements to the no elements scenario. In high forest scenario, H/L excess ranged from 4.9% (CV
test (level of significance 0.10) was then performed to check if significant differences exist between the depositional sizes of paired scenarios. Results showed the coppice statistically differs from no element scenario for all the analysed variables (R, W, smax, H/L), whereas the high forest presented differences only in
deposit width and thickness when compared with the no element scenario. Only coppice, then, seems to be able to significantly affect the debris-flow mobility at the depositional stage.
Discussions and Conclusions
Our small-scale debris flow showed that the presence of vegetation elements on the fan area causes changes in the deposit shape. In general, shorter runouts, and larger deposit widths have been observed.
Our observations on debris flow mobility suggest that high forest management is not able to significantly contribute to reduce debris flow motion at the highest solid concentrations, while a certain reduction can be achieved for the lowest concentrations. On the contrary, coppice seems to provide a notable contribution, which increases as the solid concentration raises. At the lowest concentration, high forest offered an H/L excess increment, if compared to no element scenario, 1.1 times greater than coppice, while at the highest sediment concentration, coppice offered an increment of H/L excess 8.1 times greater than high forest. The same behaviour can be observed for R reduction: at the lowest concentration, high forest and coppice showed a similar reduction capacity of R with respect to no element scenario, whereas at the highest concentration, coppice offered a reduction of R values 16.5 times greater than high forest.
Our interpretation of the better protective function exhibited by coppice setting is that coppice stocks manifest a more effective obstruction action compared to high forest trunks. In fact, the coppices in our laboratory model were composed by eight elements (coppice shoots) and occupied in the merging zone (coppice stocks) about 1 cm transverse to the flow direction, whereas high forest trunks occupied 0.6 cm per element. In addition, at the higher concentration of the mixtures, largest particles floating in the matrix are easily trapped by the upper part of the simulated coppice forests, where the shots form a sort of retaining rake. At lower concentrations, such an effect is reduced because both low flow depths do not interact with the ‘rake’ and the debris-flow tail exerts a washout effect against the sediments previously deposited behind the elements. This fact explains the highest variability of R and H/L in the coppice data.
To consolidate such findings, more work is needed to further investigate the validity of using the H/L ratio as a valid indicator for different soil uses and vegetation covers of the fan area as well as to draw guidelines on the protection efficiency of the forest against gravity-driven natural hazards.
Acknowledgments - This study was funded by the PRIN2010-11Project ITSE: “National network for monitoring, modelling and sustainable management of erosion processes in agricultural land and hilly-mountainous area”, prot. 20104-ALME4.
References
Fell, R., J. Corminas, C. Bonnard, L. Cascini, E. Leroi, and W.Z. Savage (2008). Guidelines for landslide susceptibility, hazard and risk zoning for land-use planning, Q. J. Eng. Geol., 102, 99-111.
Irasawa, M., Y. Ishikawa, A. Fukumoto, and T. Mizuyama (1991). Control of debris flows by sabo tree zones (in Japanese). J. Comput. Civil. Eng., Public Works Research Institute, Ministry of Construction, 33(5), 30-37.
Ishikawa, Y., S. Kawakami, C. Morimoto, and K. Mizuhara (2003). Suppression of debris movement by forests and damage to forests by debris deposition. J. For. Res., 8(1), 0037-0047, doi: 10.1007/s103100300004.
Legros, F. (2002). The mobility of long-runout landslides. Engineering Geology, 63(3), 301-331.
Mizuyama, T., Nakano Y., and Suzuki H. (1989). Control of running landslide and/or debris flow with trees (in Japanese). J. Jpn. Soc. Erosion Cont. Eng., 42(1), 34-36.
Morlotti, E. (2010). Funzione del bosco nella fase di arresto dei debris flow. Università degli studi di Milano. Doctoral Thesis, 23° Cycle. Rickenmann, D. (2005). Runout prediction methods. In Debris-flow Hazards and Related Phenomena. Springer Berlin Heidelberg, 305-324.