Presentazione su invito alla Giornata di studio AIGEO - Università La Sapienza "Monitoraggio e valutazione dei processi erosivi"

Invited talk nell'ambito del Workshop Italy-Chile CONICYT project "Natural Risks in Mountain Areas and their Effects in River Morphology"

The increasing availability of high-resolution Digital Terrain Models (DTMs) in recent years represents a great opportunity for Geosciences being this input information a fundamental prerequisite for an accurate representation of the surface and of the processes acting on it. Nowadays, a number of techniques in order to produce high-resolution DTMs are available. Depending on the scale of analysis, the most used survey techniques are_ i) Radar, ii) LiDAR and iii) Photogrammetry. The raw survey-derived products are often in need of a careful post-processing in order to filter out unwanted features. This, for example, could be the case of vegetation removal for surface processes modeling. After the filtering procedures, usually either the point density under filtered areas drops down or missing data areas are created. Resulting interpolated surfaces, in particular over these areas, could vary significantly, mainly depending on the selected interpolation algorithm, sometimes leading to the creation of artifacts that are not able to mimic the original surface trend and texture. In this work, we devised an experiment in order to compare the results of an image inpainting technique over missing data holes against commonly used interpolators (IDW, Spline, Kriging, Natural Neighbor) quantifying the accuracy of the approaches. The image inpainting technique has demonstrated in all the cases a significantly better performance in reconstructing the original surface. An assessment of the extent up to which such an approach could be regarded as robust has also been carried out. Selected applications of surface propagation models to the reconstructed surfaces have been reported so as to highlight how surface variability and uncertainty in surface reconstruction can influence (positively/negatively) model results. Applications of such an approach could pave the way for accurate surface interpolation and/or surface reconstruction in case of the need to remove selected features to characterize several modeling scenarios or to fill the gaps to reproduce the original surface in a consistent way.

The study of a landslide runout (i.e., the distance traveled by sediment detached by a landslide) is fundamental for a comprehensive understanding of the risk associated with landslide occurrence since sediment mobilized from landslides can often reach downstream areas causing extensive damages. The spatial characterization of sediment dynamics through geomorphic systems (i.e., sediment connectivity) is a key property to define sediment transfer paths. Sediment connectivity is mainly controlled by the morphological complexity of the catchment and the spatial organization of vegetation. In particular, vegetation plays an important role on sediment connectivity, since it influences surface roughness and local sediment retention capacity. Moreover, vegetation changes due to land use and management practice modifications can introduce major changes in sediment delivery dynamics, leading to important effects on earth surface processes, especially in relation to slope instability phenomena. Within this framework, the aim of this work is to evaluate how vegetation characteristics and its modifications over time can affect sediment connectivity in response to shallow landslides occurrence. A GIS-based index of connectivity (IC) was computed for a hilly catchment located in the northern Apennines in Italy. This area was characterized by a high degree of land abandonment, leading to important landscape modifications over the last 60 years. Moreover, the studied catchment was highly affected by shallow landslide phenomena, mainly triggered by an intense rainfall event in 2009. Land use changes characterizing the study area in the last 60 years were integrated into IC computation. The Overland Flow Manning's n Roughness Values was used in order to represent the impedance to water and sediment fluxes related to each land use class. This parameter was integrated into a revised version of IC weighting factor (W). In particular, both a linear e non-linear relationship for the Manning's n based W was computed. This allowed integrating the hydraulic efficiency of overland runoff flow in the analysis framework. Finally, the results of IC implementing the revised W, based on land use change dynamics, was integrated into a data-driven model to evaluate the probability that the sediments produced by instability phenomena could reach downstream areas. In particular, this analysis allowed identifying those areas in which the mobilized sediment could damage the road network blocking traffic, isolating villages and stopping activities.

A significant number of mountain catchments have suffered important land use changes due to agricultural abandonment during the second half of the 20th century. These changes do not only induced an increase in the vegetation cover but also a series of natural and anthropic morphological changes (e.g. terraces) that influence the water and sediment flux through catchments. In the present study the effect of these changes (i.e. land cover and topography) on sediment connectivity is analyzed in different scenarios of change in the Upper Cinca Basin (Southern Pyrenees). Potential sediment connectivity can be assessed by a morphometric indicator named Index of Connectivity (IC) developed by Cavalli et al (2013) based upon the original approach by Borselli et al (2008). The input variables of IC are determined by the land use or landscape roughness, and topography. In this work we present a methodological workflow to obtain such variables from historical aerial photos using digital photogrammetry through Structure from Motion and Multi View Stereo algorithms (SfM-MVS). A prerequisite of the analysis is to reconstruct the land use and the landscape properties at the period in which the IC is estimated. The analysis consists of three interrelated steps_ (a) extraction of historical ortophotomaps and point clouds from historical photos (aerial photos from 1957 and 1977) using SfM-MVS, (b) derivation of land use maps and topographic models for those periods to parametrize IC, and (c) assessment and comparison of historical sediment connectivity in each period. The workflow was tested in different contrasted sub-catchments representative of three main scenarios_ (1) Changes on land cover; (2) Topographic changes on agricultural fields (terracing); (3) Topographic changes associated to infrastructures (road construction). From the methodological point of view, results highlight as the selection of the suitable resolution of the Digital Elevation Models used in each scenario is crucial for the correct analysis of IC changes. Terraces affect connectivity in a major degree than land use changes, following a more spatially concentrated pattern, while the morphological variation due to road construction led to an increase of erosional activities uphill the road. In the light of these results, the developed workflow to assess sediment connectivity at multiple temporal scales has proved to be a promising and useful methodology to plan and asses potential risks associated to morphological and land use changes.

Debris-flow catchments are characterised by remarkable geomorphic changes that can occur with high frequency. Accordingly, monitoring topographic changes induced by these processes requires high-resolution surveys acquired with high frequency. Recent photogrammetric techniques, such as Structure from Motion (SfM) and Multi-View Stereo (MVS), represent a low-cost opportunity for acquiring multi-temporal high-resolution topography. However, these techniques need important steps of data processing and uncertainty analysis to identify and filter erroneous or unwanted data, that may have a significant effect on the estimates of topographic changes. Within this context, in this paper we present a methodological and standardized workflow for_ i) data-acquisition and post-processing to obtain usable and accurate Digital Terrain Models (DTMs) and ii) error analysis to assess uncertainty in the study of topographic changes through DEM-differencing analysis (DEMs of Difference or DoD). This workflow is based in the application of SfM and MVS techniques and was developed and tested in a small area of the Moscardo catchment (eastern Italian Alps). Multi-temporal SfM-MVS photogrammetry based on images taken from the ground and by means of UAV were carried out before and after three debris-flows occurred between June and July 2016. The developed workflow involves the following steps_ (a) point cloud generation through the application of SfM-MVS; (b) analysis of the precision and accuracy of the Ground Control Points (GCPs) and the resultant point clouds by means of Check Points (CPs); (c) data filtering, cleaning and decimation through geostatistical tools; (d) evaluation of the need to carry out the alignment of multi-temporal point clouds; (e) DTMs generation; (f) assessment of a minimum Level of Detection (minLoD) based on data precision; (g) DoDs preparation and thresholding based on the minLoD; (h) assessment of topographic changes.

Il presente studio nasce dall'esigenza degli uffici della Regione Veneto di predisporre una procedura finalizzata alla valutazione della pericolosità sui conoidi alluvionali classificati come zone d'attenzione nei PAI Piave, Brenta-Bacchiglione e Livenza. Al fine di individuare un percorso oggettivo per la valutazione della pericolosità sono state identificate e proposte metodologie di raccolta dati e di analisi atte ad individuare la tipologia del processo da investigare distinguendo, in particolare, tra colate detritiche e piene con trasporto solido, essendo questa distinzione di fondamentale importanza al fine di orientare le successive analisi. Nell'ambito dello studio sono stati sviluppati due strumenti per condurre e facilitare la raccolta dati del sistema bacino-conoide: i) una scheda, in formato digitale editabile, strutturata in diverse sezioni per la raccolta di una serie di informazioni che concorrono a caratterizzare oggettivamente la morfologia, la genesi, i processi attualmente attivi e lo stato di funzionamento del sistema bacino-conoide; ii) un database geografico delle aree di conoide che integra le informazioni riportate nelle schede e la documentazione su eventi storici reperita da diverse fonti documentarie. Nell'ambito della procedura per la valutazione della pericolosità sono stati sviluppati diversi software liberamente disponibili per l'identificazione lungo il reticolo idrografico delle aree di potenziale innesco, rallentamento e deposito di colate detritiche e per il calcolo della distanza di arresto delle colate detritiche finalizzato all'assegnazione di una priorità d'indagine dei conoidi. Questi applicativi costituiscono un prezioso ausilio per condurre le analisi previste dalla procedura sviluppata. Il presente contributo presenta inoltre i risultati delle analisi in tre aree test situate nelle province di Vicenza, Treviso e Belluno.

The paper proposes a modeling analysis of debris flows in anomalous basin-fan systems. Previous studies pointed out that the main geo-hydrological hazard in anomalous basin-fan systems is really related to the occurrence of debris flows, favored by the presence of large amount of loose debris on high slopes, which also favor the predominance of the entrainment process in determining the magnitude of the events. Consequently, the pseudo-3D model RAMMS DEBRIS FLOW was used in order to simulate the debris flow dynamic in anomalous systems. At this aim, a back analysis was carried out on the debris flow that occurred in July 2013 in the Gadria catchment (Val Venosta, northern Italy). This event was characterized by a quite small source area located in the head zone of the basin, whereas the volume that reached the basin outlet was quite large, because of entrainment process. The frictional parameters were calibrated using the entrainment simulation, obtaining values more similar to the ones typical of rock avalanches than the values generally used for the simulation of debris flows.

This paper describes the application of a commercial distributed optical fiber sensing system to a large scale physical model of landslide. An optical fiber cable, deployed inside the landslide body, is interrogated by means of optical frequency domain reflectometry with very high spatial density. A shallow landslide is triggered in the physical model by artificial rainfall and the evolution of the strain is measured up to the slope failure. Precursory signs of failure are detected well before the collapse, providing insights to the failure dynamic.

Hydro-geomorphic connectivity has significantly emerged as a new concept to understand the transfer of surface water and sediment through landscapes. A further scientific challenge is determining how the concept can be used to enable sustainable land and water management. This research proposes an interesting approach to integrating remote sensing techniques, connectivity theory, and geomorphometry based on high-resolution digital terrain model (HR-DTMs) to automatically extract landslides crowns and gully erosion, to determine the different rate of connectivity among the main extracted features and the river network, and thus determine a possible categorization of hazardous areas. The study takes place in two mountainous regions in the Wellington Region (New Zealand). The methodology is a three step approach. Firstly, we performed an automatic detection of the likely landslides crowns through the use of thresholds obtained by the statistical analysis of the variability of landform curvature. After that, the research considered the Connectivity Index to analyse how a complex and rugged topography induces large variations in erosion and sediment delivery in the two catchments. Lastly, the two methods have been integrated to create a unique procedure able to classify the different rate of connectivity among the main features and the river network and thus identifying potential threats and hazardous areas. The methodology is fast, and it can produce a detailed and updated inventory map that could be a key tool for erosional and sediment delivery hazard mitigation. This fast and simple method can be a useful tool to manage emergencies giving priorities to more failure-prone zones. Furthermore, it could be considered to do a preliminary interpretations of geomorphological phenomena and more in general, it could be the base to develop inventory maps. References Cavalli M, Trevisani S, Comiti F, Marchi L. 2013. Geomorphometric assessment of spatial sediment connectivity in small Alpine catchments. Geomorphology 188_ 31-41 DOI_ 10.1016/j.geomorph.2012.05.007 Sofia G, Dalla Fontana G, Tarolli P. 2014. High-resolution topography and anthropogenic feature extraction_ testing geomorphometric parameters in floodplains. Hydrological Processes 28 (4): 2046-2061 DOI_ 10.1002/hyp.9727 Tarolli P, Sofia G, Dalla Fontana G. 2012. Geomorphic features extraction from high-resolution topography_ landslide crowns and bank erosion. Natural Hazards 61 (1): 65-83 DOI_ 10.1007/s11069-010-9695-2

Unravelling the spatio-temporal variability of the sediment transfer within a catchment represents a challenge of great importance to quantify erosion, soil redistribution and their impacts on agricultural landscape. Structural and functional connectivity have been identified as useful aspects of connectivity that may clarify how these processes are coupled or decoupled in various types of catchment sediment cascades. In this study, hydrological and sediment connectivity in a Mediterranean agricultural catchment (1.4 km2) modified through traditional drainage systems (i.e. ditches and subsurface tile drainages) was assessed during two contrasted rainfall events occurred in October 2016 (20 mm in 24 h -return period < 1 yr-, I30 6.6 mm h-1 with 32 mm accumulated in 14 days) and in December 2016 (99 mm in 24 h -return period 25 yr-, I30 23 mm h-1 with 39 mm accumulated in 14 days). A morphometric index of connectivity (IC) was calculated to study the spatial patterns of structural connectivity. The identification of the main sediment pathways -in terms of functional connectivity- was conducted by field mapping, whilst the estimation of erosion and deposition rates by the analysis of high resolution digital terrain models (i.e. 5 cm pix-1; RMSE < 0.05 m) obtained from automated digital photogrammetry and unmanned aerial vehicle (UAV). The IC estimations allowed the identification of the most (dis-)connected areas related with the anthropogenic control in the resisting forces of the catchment. On the one hand, in the upper part of the catchment, depositional compartments were created by dry-stone walls that separate agricultural properties laminating flash floods. On the other hand, in the lower part of the catchment these depositional compartments were generated by an orthogonal network of ditches situated topographically above the natural thalwegs. In its turn, the most connected areas are located in the steepest parts of the catchment under rainfed herbaceous crops without dry stone walls and also within the lowland depositional compartments where the pathways are diverted generating parallel concentrated flows because of the greater elevation of these ditches. The observed spatial patterns of functional connectivity showed significant differences between the two events, although well fitted with IC as a clear evidence of anthropogenic controls in the resisting forces. During the October 2016 event -representative of high frequency-low magnitude events in the catchment- traditional drainage systems controlled the water and sediment transfer which was mainly concentrated within the ditches. By contrast, during the event of December 2016 -representative of extreme events- this transfer process was controlled by the natural morphology of the catchment, which activated coupling mechanisms between different compartments, increasing the effective area and triggering erosion processes including the formation of rills and incipient gullies. The spatial location of the sediment mobilization and deposition areas during the extreme event in December 2016 is well fitted with the IC estimations. The application of IC, therefore, may provide useful information to improve the drainage systems design and the implementation of measures to prevent soil losses.

In a fluvial system, mountain basins control sediment export to the lowland rivers. Hence, the analysis of the erosion processes and sediment delivery patterns that act in mountain basins is important. Several studies have investigated the alterations triggered by recent climatic change on the hydrological regime, whilst only a few works have explored the consequences on the sediment dynamics. Here we combined and analyzed the quasi-unique dataset of climatic conditions, landscape response, and sediment export produced, since 1986 in the Rio Cordon basin (5 km2, Eastern Italian Alps) to examine the sediment delivery processes occurring in the last three decades. The temperature, precipitation, and fluvial sediment fluxes in the basin were analyzed using continuous measurement executed by a permanent monitoring station, while the landscape evolution was investigated by three sediment source inventories established in 1994, 2006, and 2016. Thus, the analysis focused on the trends exhibited during the periods 1986-1993, 1994-2006, and 2007-2015. In terms of climatic conditions, three distinct climate forcing stages can be observed in the periods analyzed_ a relatively stable phase (1986-1993), a period characterized by temperature and rainfall fluctuations (1994-2006), and a more recent warmer and wetter phase (2007-2015). In the 1986-1993 period, the fluvial sediment fluxes reflected the stable trend exhibited by the climatic conditions. In the subsequent 1994-2006 period, the average temperature and precipitation were in line with that previously observed, although with higher interannual variability. Notwithstanding the climate forcing and the occurrence of high magnitude/low frequency floods that strongly influenced the source areas, between 1994 and 2006 the Rio Cordon basin showed relatively limited erosion activity. Hence, the climatic conditions and the landscape response can only partially explain the strong increase of sediment export recorded in the 1994-2006 period. In this sense, the sediment availability resulting from armour layer and bedform removal appears crucial to describing the sediment fluxes during this period, stressing the key role of the in-channel sediment supply. In the recent period 2007-2015 a marked climate warming accompanied by increased precipitation was observed. This climate forcing did not affect the landscape evolution, with sediment source extent remaining substantially in line between 2006 and 2016. The absence of a significant landscape response and the restoration of the channel armour layer can describe the limited sediment fluxes observed during the last decade. In particular, the increased temperature and precipitation were not accompanied by an increase in flood occurrence and magnitude, stressing the evident absence of hillslope-channel network coupling. This research was funded by the University of Padova Research Projects 'Sediment transfer processes in an Alpine basin_ sediment cascades from hillslopes to the channel network-BIRD167919'.

Land uses changes interfere on sediment production and delivery in fluvial channel networks. The study of the evolution of sediment connectivity associated with different land use changes is prerequisite for a better understanding of sediment budgets and associated processes. Previous studies examined historical changes of sediment connectivity, but most are based on indices of sediment connectivity (IC) estimated by means of_ (i) a single Digital Elevation Model (DEM), usually the most recent; and (ii) a weighting factor parameter, used in IC as a proxy of the impedance to sediment fluxes, that is assessed based on land use properties. However, some structural or geomorphological elements determined by both natural processes (e.g. rock falls) and human impacts (e.g. land uses changes) may have fundamental influences on connectivity, especially in a mountain areas typically affected by mass movements and strong land crop abandonment during the 20th century. Therefore, all these elements are able to modify landscape properties and, consequently, sediment connectivity. Within this context, the objective of this work is to develop and apply a workflow to extract historical IC maps using different information that can be obtained from historical aerial photos. A prerequisite of the analysis is to reconstruct the land use and the landscape properties at the period in which the IC is estimated. The analysis consists of three interrelated steps_ extraction of historical ortophotomaps and point clouds from historical photos (aerial photos from 1957 and 1977) in the Upper River Cinca (Southern Pyrenees), derivation of land use maps and topographic models for those periods, and assessment and comparison of historical sediment connectivity in contrasted sub-catchments exhibiting variable degrees of land use change. The study of changes on sediment connectivity through time may provide valuable information to understand some of the floodplain and channel adjustments that are observed in the majority of the fluvial systems experiencing massive land use changes in their headwaters during the middle of the 20th century.

The Moscardo Torrent (eastern Italian Alps) is a small rugged catchment (drainage area 4.1 km2, range in elevation between 890 and 2043 m) frequently affected by debris flows that deliver large amounts of sediment to the receiving stream, and cause concerns for infrastructures located on the alluvial fan and near the confluence. Over the last decades, hydraulic control works were implemented in the main channel to limit bed erosion and to stabilize channel banks. Although the objectives of training works have been only partly achieved, check dams and hillslope stabilization works have affected the sediment transfer from hillslopes to the channels and along the main channel. The effects of hydraulic control works were investigated by means of multi-temporal Structure from Motion (SfM) surveys based on images taken from the ground and UAV. The ground and air based surveys were carried out over a channel reach in which two check dams have recently been built. SfM surveys were taken before and after three debris-flow events (occurred between June and July 2016), allowing the generation of four high-resolution Digital Elevation Models (DEMs). Geomorphic changes caused by the debris-flow events have been assessed in order to produce the DEM of Differences (DoDs with a 0.2 m spatial resolution) that allowed estimating erosion and deposition volumes in the study area. Furthermore a debris-flow monitoring system has been in operation in the Moscardo Torrent; the analysis of the videos and of the hydrographs recorded by ultrasonic sensors permitted to assess the debris-flow volumes. These estimates were used to characterize the magnitude of events in support of the topographic analysis. By examining the changing pattern of erosion and deposition over time it was possible to understand the check dams' effects on sediment dynamics. The results show that the new check dams effectively stored sediment transported by the three debris flows. However, once the check dams have been completely filled, they lost their functionality, letting sediment flow downstream along paths drawn accidentally by the torrent control works and by the morphology of debris-flow deposits. Moreover, debris-flow lobes deposited upstream of the check dams could act as sediment sources further increasing downstream debris-flow magnitude.

As a consequence of a changing climate the occurrence of unexpected events, like flooding and erosion, that affects urban areas will most likely increase. The infrastructure is especially vulnerable to heavy rainfall events due to high costs and long term investments. Accumulation of water and sediments thus has a large impact on the consequences of such events, and it is therefore essential to identify factors that influence the catchment and the occurrence of flooding. Both spatial and temporal characteristics of the patterns of sediment connectivity is important for estimating the sediment contribution and transfer paths in a catchment. In recent years several approaches have been developed to assess sediment connectivity, as for example the geomorphometric indices of sediment connectivity which mainly picture a static frame of the system. With the development of remote sensing technology and the growing availability of satellite images the opportunity to consider also temporal variability and hydrological parameters as soil moisture within this kind of indices is increasing (e.g., Foerster et al., 2014; Gay et al., 2015). However, there is still a knowledge gap in considering the potential of soil moisture satellite imagery in assessing sediment connectivity at the catchment scale. This study aims to integrate spatial and temporal soil moisture properties in the index of sediment connectivity by Cavalli et al. (2013), which can be used to predict where flood events will have the strongest impact. The results will provide decision makers with a prediction tool to identify road-intersections that are prone to flood risk at the catchment scale. The method developed in this study would increase awareness and be a basis for decision makers and stakeholders to promote action towards enabling sustainable water and land management. References Cavalli, M., Trevisani, S., Comiti, F., Marchi, L. , (2013) Geomorphometric assessment of spatial sediment connectivity in small Alpine catchments. Geomorphology 188_ 31-41, doi_10.1016/j.geomorph.2012.05.007. Foerster, S., Wilczok,.C., Brosinsky, A., Segl, K., (2014) Assessment of sediment connectivity from vegetation cover and topography using remotely sensed data in a dryland catchment in the Spanish Pyrenees. J Soils Sediments 14_1982-2000 Gay, A., Cerdan, O., Mardhel, V., Desmet, M. (2015) Application of an index of sediment connectivity in a lowland area. J. Soils Sediments 16(1), 280-293, doi_10.1007/s11368-015-1235-y,.