High-resolution topography for understanding Earth ...

Geomorphology 216 (2014) 295–312

Contents lists available at ScienceDirect

Geomorphology journal homepage: www.elsevier.com/locate/geomorph

Invited review article

High-resolution topography for understanding Earth surface processes: Opportunities and challenges Paolo Tarolli ⁎ Department of Land, Environment, Agriculture and Forestry, University of Padova, Agripolis, viale dell'Università 16, 35020 Legnaro (PD), Italy

a r t i c l e

i n f o

Article history: Received 4 February 2013 Received in revised form 21 February 2014 Accepted 3 March 2014 Available online 12 March 2014 Keywords: Earth surface processes High-resolution topography Lidar Geomorphic signatures Anthropogenic signatures Anthropocene

a b s t r a c t In the last decade, a range of new remote-sensing techniques has led to a dramatic increase in terrain information, providing new opportunities for a better understanding of Earth surface processes based on geomorphic signatures. Technologies such as airborne and terrestrial lidar (Light Detection and Ranging) to obtain highresolution topography have opened avenues for the analysis of landslides, hillslope and channellization processes, river morphology, active tectonics, volcanic landforms and anthropogenic signatures on topography. This review provides an overview of the recent flourishing literature on high-resolution topographic analyses, underlining their opportunities and critical issues such as their limitations. The goal is to provide answers to questions such as what kind of processes can be analyzed through high-resolution topographic data and how to do it. The review focuses on two different environments: natural and engineered landscapes. In both contexts, high-resolution topography offers opportunities to better understand geomorphic processes from topographic signatures. Particular attention is given to engineered landscapes in which the direct anthropic alteration of processes is significant. The last part of the review discusses future challenges. © 2014 Elsevier B.V. All rights reserved.

1. Introduction

1.1. Lidar and high-resolution topography

From the analysis of the surface morphology of the Earth, we are able to learn a lot about its history and processes. Just to make a comparison, let us consider the skin of a man: his age and “processes” (e.g., disease or accidents) may leave significant signatures. These signatures are key elements to analyze the life of the man. The same is true for the surface of the Earth. Fig. 1 illustrates two mountain regions, located in two different parts of the world: the Uintas, Utah, USA (Fig. 1a) and the Himalayas (Fig. 1b). They present different morphologies: smooth in the Uintas, and rough in the Himalayas. The latter is characterized by narrow valleys, steep hillslopes, and a higher drainage density, and the former by gentler hillslopes, and a less complex drainage system. The two areas are also characterized by different ages, climate and tectonic forcing. Fig. 2 shows a more local phenomenon: a hillslope affected by a shallow landslide in the Rio Cordon basin, Dolomites, Italian Alps. In this case, the portion of hillslope without severe erosion is characterized by a smooth topography, while the portion with landsliding processes presents a rougher terrain. It is clear that surface signatures (e.g. roughness) allow us to understand what kind of process affects a certain region, although the quality of such topographic information including resolution is critical.

In the last decade, new remote-sensing techniques have led to a dramatic increase in terrain information, providing a basis to develop new methodologies for analyzing Earth surfaces (Tarolli et al., 2009). Among the available remote-sensing technologies, airborne and terrestrial lidar (Light Detection and Ranging) has a significantly rising number of applications and innovations (Roering et al., 2013). Lidar provides highresolution topographic data with notable advantages over traditional surveying techniques (Slatton et al., 2007). Valuable characteristics of this technology, compared to more traditional photogrammetric techniques, are: 1) the capability to derive topographic data related to the bare ground surface by automatically filtering vegetation or other objects on the surface; and 2) the capability to produce sub-meter resolution Digital Terrain Models (DTMs) and Digital Surface Models (DSMs) over large areas. The question is whether earth scientists really need sub-meter digital terrain resolution, and if yes, for what purpose. According to previous studies, a 10 m DTM has a resolution suitable for analyzing major hydro-geomorphic processes. Zhang and Montgomery (1994) found that a grid size of 10 m is sufficient for many DEM-based applications of geomorphic and hydrologic modeling. Tarolli and Tarboton (2006) also recognized that a 10 m DTM should be considered as optimal for shallow landslide modeling, overcoming the local slope noises detected with higher-resolution topography. At scales finer than 10 m, small-scale errors in the determination of slope may result in an increased number of spurious potential landslide initiation locations, reducing the discriminating capability of the model. These

⁎ Tel.: +39 049 8272677; fax: +39 049 8272686. E-mail address: [email protected].

http://dx.doi.org/10.1016/j.geomorph.2014.03.008 0169-555X/© 2014 Elsevier B.V. All rights reserved.

296

P. Tarolli / Geomorphology 216 (2014) 295–312

Fig. 1. Satellite images from Google Earth related to the Uinta (UT) (a), and Himalaya (b) mountain ranges.

results are in line with Claessens et al. (2005), who stated that a ‘perfect’ DEM resolution does not exist; the ‘optimal’ window size for a study is directly related to the average size of the landslides in that region. Milledge et al. (2012) discussed the validity of the infinite length assumption for modeling shallow landslides, and found results similar to those presented by Tarolli and Tarboton (2006). Although a coarser cell size may be ineffective for identifying the minimum landslide area, higher-resolution data can be problematic because the identified landslides may violate the infinite slope length assumption. Their results indicate that the infinite length assumption is valid for many of the existing modeling studies, which have used coarse (N25 m) resolution data. For models with a finer resolution (b 10 m), the assumption of infinite length is less valid, and depends on the assumed landslide failure plane depth and surficial material properties (Milledge et al., 2012). However, these studies based their analyses on the modeling of a process, not on geomorphic signatures.

Fig. 3 shows the morphology of an upper part of the Rio Cordon basin. Fig. 3a shows a high resolution (15 cm) aerial photograph, while shaded relief maps derived by 10 and 0.5 m lidar DTMs are shown in Fig. 3b, c. The DTMs are from a high quality lidar dataset with a point density greater than 5 points m−2, recording up to four returns with an absolute vertical error less than 0.3 m in flat areas (Pirotti and Tarolli, 2010). Fig. 3b, c clearly shows that only the 0.5 m grid cell size permits the detailed recognition of geomorphic signatures related to channel heads and colluvial channels. The colluvial channels of this area have a bankfull width of 1 to 1.5 m; therefore a highresolution topography with grid cell sizes smaller than 1 m is the right choice. Literature published during the last 10 years includes a lot of applications of lidar data, and this trend is becoming more prominent due to the wider availability and the lower cost of lidar DTMs for the scientific community and environmental agencies. Such lidar applications


297

Fig. 2. Landsliding area located in the Rio Cordon basin (Dolomites, Eastern Italian Alps).

include automatic detection of landslide scars and channel heads, landslide movement analysis, automatic channel network extraction, channel bed morphology analysis, evaluation of fault and volcanic activities, quantification of sediment production and deposition, detailed numerical flood modeling, especially that related to the micro-topography of urban areas, and extraction of anthropogenic features. Their novelty compared with the analysis conducted in the last 30 years using coarser topographic information allows us to say that we are experiencing a “new chapter” of geosciences in terms of the analysis of Earth surface processes. The following sections summarize the most recent advances in the applications of high-resolution topography, in relation to natural landscape, engineered landscape, and future challenges. The goal is to present the state of the art, providing useful guidelines about analytical methods and types of process to be analyzed using high-resolution topography. 2. Natural landscapes 2.1. Landslides Among the applications of lidar data, the analysis of landsliding processes has made the greatest advances. Derron and Jaboyedoff (2010), and Jaboyedoff et al. (2012) provided useful reviews about the use of lidar-derived high-resolution topography in the detection, characterization and monitoring of mass movements (landslide, rockfall and debrisflow), hazard assessment, and susceptibility modeling. This section presents an update with more recent literature focusing mainly on the application related to automatic landslide feature recognition and its effectiveness in forested areas. One of the first and most interesting applications of lidar data in landslides feature extraction was presented by McKean and Roering (2004). In their work to detect landslide areas, lidar data were considered to evaluate the local surface roughness by measuring the variability in slope and aspect, and a two-dimensional topographic curvature through a Laplacian operator. The assumption behind this analysis is the same as that illustrated in Fig. 2: landslide areas tend to have rougher terrain. This roughness is correctly detected only with a lidar-derived DTM having a meter-scale resolution. The novelty of their work is that introducing simple topographic indexes of surface roughness enables the automatic mapping of landslide morphology and estimation of landslide activity, providing a preliminary landslide inventory map. Fig. 4 shows an example of the roughness index of residual topography (Cavalli et al., 2008) computed using 10 and 0.5 m DTMs for an area with shallow landslide scars. This example indicates that only with a 0.5 m DTM is it possible to correctly represent surface roughness related to landslide scars. Glenn et al. (2006) used

Fig. 3. Aerial photograph related to the upper part of Rio Cordon basin (Dolomites, Eastern Italian Alps) (a), and shaded relief maps obtained from 10 m (b) and 0.5 m (c) lidarderived DTM. Channel heads surveyed in the field by DGPS (Pirotti and Tarolli, 2010) are mapped with white circles. White and red arrows (a and c) indicate a colluvial channel (after Orlandini et al., 2011).

high-resolution topographic data to calculate the surface roughness, slope, semivariance, and the fractal dimension. Their work demonstrates that high-resolution topographic data have the potential to differentiate morphological components within a landslide, and they can support the analysis of the material type and landslide activity. Ardizzone et al. (2007) used shaded relief and slope maps obtained from a 2 m lidar DTM to update the available landslide inventory maps. They also compared the result with that from a coarser (10 m) DTM, underlining the effectiveness of the fine resolution data in analyzing surface morphology. Booth et al. (2009) presented two methods based on spectral analysis to quantify and automatically map the topographic signatures of deep-seated landslides: (1) power spectra production using the two-dimensional discrete Fourier transform, and (2) the two-dimensional continuous wavelet transform. Their goals were to identify the characteristic spatial frequencies of deep-seated landslides

298


Fig. 4. Aerial photograph of a shallow landslide surveyed in the Upper Rio Cordevole basin (Dolomites, Eastern Italian Alps) (a), and roughness index maps of residual topography (Cavalli et al., 2008) calculated from 0.5 m (b) and 10 m (c) lidar-derived DTMs.

and the related morphological features such as hummocky topography, scarps, and displaced blocks of material. Roering et al. (2009) combined Interferometric Synthetic Aperture Radar (DInSAR) and lidar data with historical aerial photographs to quantify landsliding and sediment transport. Similarly, Mackey and Roering (2011) used orthorectified historical aerial photographs and a 1 m lidar DTM to objectively map earthflow movements and calculate the sediment flux. The novelty of these two works is the combination of high-resolution lidar data with other sources. According to Roering et al. (2009), lidar also facilitates the accurate rectification of historical air photos, providing a template for mapping deformation and erosional features that cannot be otherwise identified. Ventura et al. (2011) presented the results of four aerial lidar surveys during 2006–2010 for an active landslide area in Italy. The interpretation and analysis of surface roughness and residual topographic surfaces derived from 1 m DTMs and their temporal variations allowed a reconstruction and tracking of the landslide activity. DeLong et al. (2012) analyzed an active earthflow surveyed in 2003 and 2007 by Airborne Laser Swath Mapping (ALSM), enabling meter-scale quantification of landscape changes. They calculated a four-year volumetric flux from the earthflow and provided detailed maps of the changing

features to estimate flux velocity and the dominant geomorphic process. Tarolli et al. (2012) proposed an objective methodology for recognizing landslide crowns and bank erosion through landform curvature. The methodology analyzes the variability of landform curvature to test different thresholds for feature extraction in terms of standard deviation, interquartile range, mean absolute deviation, and the Q–Q plot, where the variable is plotted against the standard normal deviation of the same exceedance probability, and the deviation from a straight line indicates a deviation from the Gaussian distribution (Lashermes et al., 2007; Passalacqua et al., 2010a,b). They filtered the extracted features using a slope threshold based on field observations, and successfully detected landslide crowns and bank erosion. The proposed method is rapid and useful in detecting terrain (landslide) features. It also analyzes the scale dependency of curvature in relation to an appropriate scale to derive local morphology. Fig. 5 shows the result of Tarolli et al. (2012): landslide crowns and river bank erosion surveyed in the field, the maximum curvature computed according to the Evans (1979) approach, and the extracted features under the best performance. However, the proposed method is not optimal for all surface morphologies; for example, Fig. 5 shows features not necessarily related

Fig. 5. Landsliding area in the Rio Cordon basin (Dolomites, Eastern Italian Alps). The maps show (a) the field surveyed features (landslide crowns and river bank erosion), (b) landform curvature map derived by a 0.5 m DTM and obtained with a moving window of 21 × 21 cells, and (c) the geomorphic features corresponding to the best extraction (Tarolli et al., 2012) of all the proposed methodologies (threshold value of 1.5 interquartile range of maximum curvature calculated with a 21 × 21 cells moving window). The red arrows are related to the extracted features representing the main slope instabilities investigated, while the blue arrows are related to features not related to landsliding processes.


to landslides. These features are related to morphological complexity such as boulders on slopes. Lin et al. (2013) tested the performance of the Tarolli et al. (2012) method for large-scale deep-seated landslides in forested areas. The results indicate that the methodology performs better for larger deep-seated landslides with evident crowns. Iwahashi et al. (2012) evaluated the topographic differences between rainfalland earthquake-induced landslides using 2 m lidar DTMs. They used slope gradient and curvature from the DTMs with different window sizes, and indicated that an optimal scale factor is the key in assessing shallow landslides. The representative window sizes were approximately 30 m, which may be related to the average size of landslides in each region, being consistent with the findings of Tarolli et al. (2012). Van Den Eeckhaut et al. (2012) tested the potential of object-oriented analysis to map landslides using only single pulse lidar derivatives such as slope gradient, roughness and curvature. This represents one of the most relevant advances in landslide hazard mapping, since almost all previous methods are pixel-based. The results underline that, for forested landscapes, the proposed method is able to recognize and characterize the signatures of deep-seated landslides in detail: N90% of the main scarps and N 70% of the landslide bodies were accurately identified. There are also other useful applications using lidar to investigate landslides in forested areas with automatic and objective approaches (Sekiguchi and Sato, 2004; Van Den Eeckhaut et al., 2005, 2007; Schulz, 2007; Kasai et al., 2009; Burns et al., 2010; Razak et al., 2011). Landslide detection through high-resolution topography should be considered an important advance not only from the scientific point of view but also for risk management and environmental planning. The limits of these approaches include that DTMs, even high-resolution ones, do not directly represent geology, so applications to geologically complex areas may be difficult. 2.2. Channel network 2.2.1. Hillslope and channellization processes The analysis of hydrological and erosional processes in a catchment requires a differentiation between hillslopes and channels in terms of runoff generation and erosion mechanisms (Montgomery and Foufoula-Georgiou, 1993). In this context, a detailed representation of the hillslope-to-valley transition morphology is relevant for the accurate recognition and extraction of a channel network, and, consequently, for an accurate runoff prediction (Tarolli and Dalla Fontana, 2009). An effective connection of hillslopes with the channel network results in highly efficient sediment transfer processes such as debris flows (Cavalli et al., 2013). However, interactions between hillslope and channel processes may play an important role in controlling geomorphic responses to changes in rock uplift rates (Hilley and Arrowsmith, 2008). Montgomery and Foufoula-Georgiou (1993) proposed a partitioning of a landscape into drainage and slope regimes that include hillslopes, unchannelled valleys, debris flow-dominated channels, and alluvial channels. This partitioning has been presented in a log–log drainage area–slope relation, and it has been widely used in the last two decades. The relation shows two inflections: the first is related to the hillslopeto-valley transition, while the second correlates well with the transition from debris flow-dominated colluvial channels to gentler channels flowing over alluvial deposits (Montgomery and Foufoula-Georgiou, 1993). These transitions most likely depend on climate, uplift rates, rock strength, and the history of the fluvial system such as glaciation. Montgomery and Foufoula-Georgiou (1993) found that in moderately steep topography, a DTM finer than 30 m is required to accurately identify the hillslope-to-valley transition. Ijjasz-Vasquez and Bras (1995) used a 30 m DTM and identified four scaling regimes in the slope–area diagram that depict the change from diffusive to fluvial sediment transport processes. Stock and Dietrich (2003) interpreted the curved shape of the relation between slope and area on log–log diagrams as the topographic signature of valley incision by debris flows. Tarolli and Dalla Fontana (2009) tested this diagram in a landslide-dominated region in

299

Fig. 6. Logarithmic diagrams of the local slope versus contributing area for the a) 1 m, b) 10 m, and c) 30 m DTMs based on binned and averaged data. The vertical gray line shows the slope–area reversal at the hillslope-to-valley transition. The vertical dashed gray lines show the other scaling regimes with respect to the changes in negative gradient. The study area is the Miozza basin (Eastern Italian Alps) (Tarolli and Dalla Fontana, 2009).

the eastern Italian Alps using a lidar-derived DTM with a progressively increasing grid cell size: from 1 to 30 m. Fig. 6 compares the slope– area relations obtained from 1, 10 and 30 m DTMs. The diagram for the highest grid resolution (1 m) presents four regions with different scaling responses in different trend gradients. In Region I, with small contributing areas, the area–slope scaling has a positive gradient, but it becomes negative in Region II. The negative trend is relatively indistinct in Region III, where the gradient seems to be positive, but it becomes distinct in Region IV. These four regions are similar to those observed by Ijjasz-Vasquez and Bras (1995). The changes between Regions II, III, and IV are related to dominant sediment transport processes: the less distinct negative gradient in Region III is related to the dominance of debris flows and landslides (Montgomery and FoufoulaGeorgiou, 1993) that led to valley incision (Tucker and Bras, 1998; Stock and Dietrich, 2003). This analysis is consistent with the field evidence: the study area (Miozza basin) shows several areas affected by slope failures and valleys incised through debris flow (for details see Tarolli and Tarboton, 2006). Fig. 6 also shows the hillslope-to-valley transition morphology (change between Regions I and II) and the slope–area scaling of Region III that demonstrates the topographic signature of debris flow and landslides more evidently for finer DTMs. This result is consistent with Booth et al. (2013), who found the same slope–area scaling of Region III due to the topographic signature of deep-seated landslides. For the 30 m DTM, Region I cannot be identified (Fig. 6), highlighting that only high-resolution data provide a detailed and useful representation of the hillslope-to-valley transition morphology. 2.2.2. Channel network extraction The identification of a channel network is of fundamental importance in landscape-scale geomorphic and hydrologic analyses (Montgomery and Foufoula-Georgiou, 1993), and it is a key step when studying catchment hydrological responses to rainfall events (Tucker et al., 2001). Different methods have been proposed in the literature for channel network extraction: a constant critical support area (e.g., O'Callaghan and Mark, 1984; Band, 1986; Mark, 1988; Tarboton, 1989; Tarboton et al., 1991), a slope‐dependent critical support area (e.g., Montgomery and Dietrich, 1992; Dietrich et al., 1993), and also a threshold of local curvature (Rodriguez‐Iturbe and Rinaldo, 1997; Heine et al., 2004). The most classical methods generally follow the procedure of filling pits,

300


computing the flow direction according to different flow direction algorithms (O'Callaghan and Mark, 1984; Quinn et al., 1991; Tarboton, 1997; Orlandini et al., 2003), and computing the contributing area draining to each grid cell (Tarboton, 2003). It also requires the choice of a threshold to apply to label channels. However, these classical methodologies present two main disadvantages: i) the flow direction approach is unable to operate effectively in flat areas; and ii) the extraction from flow paths to the network requires the choice of a unique threshold that is generally not able to fully reproduce the actual network. In the alpine environment, different processes account for channel initiation. Passalacqua et al. (2010b) described two main processes for channel initiation, surface erosion/landsliding and perennial flow, for the Rio Col Duro basin (sub-basin of Rio Cordon, Dolomites, Eastern Italian Alps) on the basis of previous detailed surveys on the same area (Pirotti and Tarolli, 2010). Fig. 7 illustrates an example of two channels in the Rio Col Duro basin initiated by different processes: a) a channel formed by perennial groundwater seeping flow in a low‐slope area (critical support area A of 100 m2, and local slope S of 0.37), and b) a channel formed by surface erosion during runoff in a high-steep area (A of 1190 m2, and S of 0.7). Orlandini et al. (2011) highlight how classic approaches show variable reliability and sensitivity over different drainage basins and DTM grid cell sizes, with a general tendency to overestimate the network, and they do not provide reliable predictions of channel heads across drainage basins having a different morphology and channel initiation depending on different processes. Henkle et al. (2011) observed piping and springs in the semiarid Colorado Front Range, indicating the importance of groundwater flow and complex flow paths and the interactions between surface and subsurface processes leading to channel initiation. They suggested that neither topographic parameters nor surface/subsurface processes are the dominant control on channel initiation. The example shown in Fig. 7, among others reported in literature (Jaeger et al., 2007; Tarolli and Dalla Fontana, 2009; Jefferson and McGee, 2013), underlines that a unique threshold of A or A × S for channel head identification and channel network extraction does not exist. In other words, the classical approaches are unable to predict channel heads accurately if channel initiation depends on different processes. The availability of high‐resolution topography from laser scanning offers an opportunity

to reconsider the procedures for channel network extraction, and obtaining more detailed results than in the past. Several studies published in recent years point out that a robust delineation of stream networks should be based on direct detection of channel head morphology from the DTM (e.g. Lashermes et al., 2007; Tarolli and Dalla Fontana, 2009; Passalacqua et al., 2010a,b; Pirotti and Tarolli, 2010; Thommeret et al., 2010; Sofia et al., 2011). The core idea of these approaches is to automatically extract convergent cells and connect them using classical flow routing or cost function procedures. Fig. 8 shows a shaded relief map at 0.5 m resolution for the upper part of the Rio Cordon basin. In the small colored box, the mean curvature (after Wood, 1996) is shown with which channels and ridges can be identified. The automatic and objective recognition and extraction of convergent cells is based on the analysis of the statistical distribution of curvature values. Statistical operators such as the standard deviation (Tarolli and Dalla Fontana, 2009) and Q–Q plots (probability plots consisting of a variable plotted against the standard normal deviate of the same exceedance probability) (Lashermes et al., 2007; Passalacqua et al., 2010a), along with highresolution DTMs, facilitate the objective recognition of different types of landscapes based on the different shapes of their landform curvature distribution: channellized, unchannellized ridge, and non-channellized regular surfaces. Wavelet analysis to locally filter lidar elevation data has been proposed by Lashermes et al. (2007) to recognize valleys and portions of probable channellized areas within the valley. The Q–Q plot of the Laplacian curvature and a slope direction change was used to define the objective threshold for channel network extraction. They suggested that the deviation in the positive tail of the probability distribution of curvature from a Gaussian distribution corresponds to a critical threshold of curvature which identifies channellized surfaces. Similarly, thresholds in the slope distribution identify the non-channellized and channellized valley transition. Tarolli and Dalla Fontana (2009) used curvature to assess the capability of high-resolution topography to recognize the convergent hollow morphology at channel heads. They suggested an objective methodology to extract the channel network based on the threshold range identified as the n-times of the standard deviation of mean curvature. Pirotti and Tarolli (2010) used different kernels for curvature calculation, demonstrating that the performance of such

Fig. 7. Examples of (a) a channel head formed by groundwater seeping upward in a low‐slope area (white circle, and detail in the pop-up box); (b) a channel head formed by a combination of flow accumulation and slope, with soil erosion in the Rio Col Duro basin (subbasin of Rio Cordon, Dolomites, Eastern Italian Alps).


301

Fig. 8. Shaded relief map and mean curvature obtained from 1 m lidar-derived DTM (Rio Cordon basin, Dolomites, Eastern Italian Alps).

approaches is scale-dependent. The window size for curvature calculations has to correspond to the size of the features to be detected. In addition, small window sizes may be unsuitable as only a limited area is investigated, whereas larger window sizes may yield too smooth curvature maps. Larger window sizes may lead to misinterpretation of channels. Pirotti and Tarolli (2010) indicated that the best choice is three times the channel width. Thommeret et al. (2010) used a threshold based on DTM noise to extract a badlands network, identifying convergent areas from a combination of terrain morphology indices and a single flow drainage algorithm. Passalacqua et al. (2010ab) applied the GeoNet nonlinear diffusion filtering combined with a geomorphically-informed geodesic cost function to automatically identify channel initiation points and channel paths from lidar data. Passalacqua et al. (2010b) also compared the extracted channel network with that from classic methodologies. An example of the comparison is shown in Fig. 9. The classical methodologies tend to predict channels that are actually not present, particularly in one side of the basin (red arrows in Fig. 9b, c). The channel in the lower side of the basin detected by the geometric nonlinear methodology (black arrow in Fig. 9a) corresponds to an ancient channel mapped on historical maps. Passalacqua et al. (2010b) suggest that the local nonlinear filter together with the global geodesic optimization

used in GeoNet is robust and computationally efficient, while achieving better localization and extraction of geomorphic features. Some open questions still remain, such as how to identify thresholds and select a scale without depending on data for a particular area. Sofia et al. (2011) proposed a methodology relatively independent of the input dataset or of the size of the analyzed features, based on normalized topographic attributes such as openness (Yokoyama et al., 2002; Prima et al., 2006) and minimum curvature (Evans, 1979) as a weight for the upslope area. The openness was chosen because the curvature, being a direct surface derivative, can amplify the DTM errors, as demonstrated in a later study (Sofia et al., 2013). For these papers, the identification of the optimum scale to evaluate topographic parameters is based on distribution analysis, and statistical thresholds permit the choice of parameters controlling network extraction. As a final step for an optimal definition and representation of the whole network, noise-filtering and a connection procedure are applied. Pelletier (2013) proposed a simple and robust method for drainage network extraction from high-resolution DTMs. The method consists in six different steps: (1) a filter to remove microtopographic noise, (2) mapping of the contour curvature, (3) identification of valley heads using a user-defined contour-curvature threshold criterion, (4) routing of a unit discharge of water from each valley

Fig. 9. Channel networks extracted using different methodologies (yellow channels) and compared to the DGPS surveyed one (blue channels): (a) The GeoNet methodology, (b) an area threshold At = 3099 m2, and (c) a combination of area and slope ASy N C with y = 2 (Montgomery and Dietrich, 1992) and C = 221 m2. Values of At and C are given by the mean of the drainage areas (O'Callaghan and Mark, 1984) and the product of mean area and mean slope squared, respectively, both measured at the surveyed channel heads (Pirotti and Tarolli, 2010). The channels plotted in blue are surveyed in the field by DGPS. The black arrow in (a) indicates an ancient channel mapped by GeoNet and reported in historical maps of the area. The red arrows in (b) and (c) indicate the side of the basin where the two classical methodologies tend to predict channels, which are not present in the field (Passalacqua et al., 2010b).

302


head using a multiple-flow-direction algorithm, (5) removal of discontinuous reaches from the drainage network using a user-defined discharge-per-upstream-valley-head threshold criterion, and (6) thinning of the valley network to a single pixel width. The analysis suggested that for DTMs with a resolution of approximately 1 m, the method has the capability to produce accurate results for a variety of landscapes. The advances of all the described approaches are the objective and (semi-)automatic recognition of channel networks. Their limits are similar to those observed for landslide analysis in the previous section: they are suitable in areas where the surface morphology is not complex (e.g. landscapes without bedrock outcrops and vertical cliffs), and where there are no distinct features related to human activities such as roads or agricultural terraces. 2.3. River morphology Morphological characteristics of channels and their degree of alteration are basic data for the proper management of mountainous watersheds, and detailed topographic data are a fundamental requirement to analyze parameters and establish meaningful hydrological and geomorphological relationships (Vianello and D'Agostino, 2007). As suggested by Marcus and Fonsad (2010), high-resolution topography derived from technologies such as green lidar (~bathymetric lidar), digital cameras, multispectral scanners, and interferometric synthetic aperture radar (inSAR) is useful in river morphology analysis. Cavalli and Tarolli (2011) classified such applications to river morphology into five categories: 1) bank erosion analysis (Thoma et al., 2005; De Rose and Basher, 2011), 2) characterization of river morphology (Jones et al., 2007; Cavalli et al., 2008; Notebaert et al., 2009; Vianello et al., 2009; Trevisani et al., 2010; Legleiter, 2012), 3) evaluation of river morphological evolution (Magirl et al., 2005; Wheaton et al., 2010), 4) grain size diameter analysis (Heritage and Milan, 2009; Hodge et al., 2009; Nitsche et al., 2013), and 5) monitoring and analysis of river morphology (Hilldale and Raff, 2008) or monitoring and managing ecosystems through bathymetric lidar (McKean et al., 2009). 2.3.1. Bank erosion analysis Thoma et al. (2005) quantified the sediment production of banks in a 56 km reach of Blue Earth River (Minnesota, USA) from 2001 to 2002, using topographic lidar data collected at different time steps. The results are in line with those reported in literature, with the mass wasting volume varying depending on the range of textural material entrained in the river. De Rose and Basher (2011) evaluated the accuracy of volumetric rates of bank erosion derived from lidar DTMs and photogrammetric analysis of historical aerial photography. The results demonstrate that lidar-derived DEMs can assist with the orthorectification and construction of DTMs from historical aerial photography with high accuracy, providing more accurate measurements of bank erosion. 2.3.2. Characterization of river morphology Jones et al. (2007) characterized the morphology of the Dee River in Wales (UK) through a multiscale analysis of DTMs including the classification of elevation changes. Their work underlined how the combination of airborne lidar data and GIS technology facilitates the rapid production of geomorphological maps of floodplains. Cavalli et al. (2008) presented an analysis for recognizing channel-bed morphology using 0.5 and 1 m resolution lidar DTMs using a surface roughness index, which is the standard deviation of the residual topography. The calculation of this index applies a moving window average filter to the original DTM, providing a smoothed elevation model. The residual topography is then obtained by subtracting this smoothed DTM from the original DTM, and its standard deviation within a moving window is computed. The calculated surface roughness by means of the residual topography is independent of the effect of slope along the channel. Their results suggest that, in mountain streams with shallow water, it is possible to automatically recognize and differentiate various bed

morphologies. Fig. 10 shows the results of Cavalli et al. (2008), where the step-pool reach (Fig. 10b) presents higher values of roughness index than the riffle-pool reach (Fig. 10a). Vianello et al. (2009) compared different methods for computing the slope of colluvial and alluvial channels using lidar-derived DTMs with different resolutions. This work concludes that a reliable channel slope can be derived from a lidar DTM having an appropriate resolution, by applying a slope evaluation method that considers the channel width. Notebaert et al. (2009) explored the potential use of lidar data for fluvial geomorphological research. Qualitative analysis of lidar data allowed the identification of former channel patterns, levees, colluvial hillslope, and fan deposits. Field data, topographic surveys and historical maps confirmed these results. The cell size resolution was proved to be an important factor in the identification of small landforms. The lidar data, in the same work, were also used in a quantitative analysis of channel dynamics. Notebaert et al. (2009) suggested that sequential lidar data could be used to calculate vertical sedimentation rates, as long as there is a control on the error of the reference levels used. Trevisani et al. (2010) analyzed the channel-bed morphology with the calculation of different morphometric indexes derived from a 0.5 m DTM. Three geomorphometric indices were proposed: 1) a slope index computed on the whole width of the channel bed, 2) directional variograms computed along the flow direction and perpendicular to it, and 3) local anomalies, calculated as the difference between directional variograms at different spatial scales. Their results demonstrated the capability to recognize patterns associated with boulder cascades and rapids with steps, whereas they did not clearly differentiate between morphologies with less marked morphological differences, such as step pools and cascades. Legleiter (2012) developed and evaluated a hybrid approach to the remote measurement of river morphology that combined lidar topography with spectrally based bathymetry. The results underlined that fusion of lidar and passive optical image data provides an efficient tool to characterize river morphology that would not have been possible if either dataset had been used in isolation. 2.3.3. Evaluation of river morphological evolution Magirl at al. (2005) analyzed the long-term changes induced by debris flow activities, and by fluvial reworking of tributaries' alluvial fan material of the Colorado River along the Grand Canyon (USA). The longitudinal profile of the water surface surveyed in the field by USGS in 1923 has been compared with that derived from a topographic lidar survey in 2000. The average change for the measured rapids indicated the net aggradation of the coarse-grained alluvium forming the rapids throughout the Grand Canyon. In addition, the comparison of the two water surface profiles showed enhanced pool-and-rapid morphology. Their results demonstrated how the geomorphic conclusions drawn from the data comparison represented a real change in the morphology of the rivers. Wheaton et al. (2010) analyzed fluvial geomorphic changes through 5 years of high-resolution repeat annual GPS surveys of a partially-braided portion of the River Feshie, in the Cairngorm Mountains of Scotland. In this work a new method to estimate the uncertainties of high-resolution DEM of Difference (DoD) maps was tested. The presented method is of great interest since it offers a useful tool to obtain a suitable estimation of river sediment budget through multitemporal surveys. 2.3.4. Grain size diameter analysis Hodge et al. (2009) underlined the fact that terrestrial laser scanning technology offers the scientific community a useful tool for the rapid acquisition of high-resolution and high-precision surface elevation data from in situ sediments. Their paper presented a new methodology for collecting and processing close-range laser scanner data to enable the reconstruction of DTMs from fluvial sediment surfaces, which could also be applied to other surfaces. While the factor limiting the resolution of DTMs was found to be the relative sizes of the laser footprint and smallest grains, the methodology presents a great added value. DTMs


303

Fig. 10. Roughness index-elevation for a step-pool (a), and a riffle-pool channel reach (b) calculated in the Rio Cordon basin (Dolomites, Eastern Italian Alps) from a 0.5 m DTM (Cavalli et al., 2008). Pictures of the morphological units surveyed in the field are also displayed.

produced by this methodology can be used to further understand multiple aspects of the fluvial system (e.g., it should be useful in the calculation of entrainment shear stresses; Hodge, 2007). Heritage and Milan (2009) used terrestrial laser scanner data for grain roughness height recognition of an exposed bar and river bed surface. The novelty of this study is the application of TLS to determine the full population of grain roughness in gravel-bed rivers. Grain roughness was extracted through determination of twice the local standard deviation of all the elevations in a 0.15 m radius moving window over the data cloud. They then compared the TLS-derived grain roughness with a grid-bynumber sampling method. The results suggested that the proposed methodology could replace conventional grid-by-number sampling, which is subject to errors related to low sample size and sampler bias. Nitsche et al. (2013) evaluated range imaging (RIM) as an alternative method for obtaining surface data in such complex environments. The field measurements allowed the creation of a high-resolution DTM (2 cm), similar in detail to the DTM derived from TLS data for the same site. According to the authors, RIM could be considered as a substitute for terrestrial laser scanners or photogrammetric approaches, and one of the future challenges in small-scale applications such as grain size diameter calculations and micro-topography measurements. When the water is deep, the bathymetric lidar, also known as green lidar, overcomes the limitations of topographic airborne lidar. 2.3.5. Monitoring of river morphology and ecosystems with bathymetric lidar Hilldale and Raff (2008) analyzed the quality of the bathymetric lidar survey from the perspective of its application toward creating accurate, precise and complete streambed topography for numerical modeling and geomorphological assessment. McKean et al. (2009) used a narrow-beam, water-penetrating, green lidar system to map

continuously 10 km of a mountain stream channel, including its floodplain topography, and they considered wavelet analyses to investigate the spatial patterns of channel morphology and salmon spawning. The new terrestrial–aquatic lidar could catalyze rapid advances in understanding, managing, and monitoring of valuable aquatic ecosystems through unprecedented mapping and attendant analyses. 2.4. Active tectonics Tectonic processes may leave significant signatures on the Earth's surface. Reading tectonic deformation from topography is a complicated endeavor because landscapes reflect both tectonic deformation and the ensuing response of superficial processes (Hilley and Arrowsmith, 2008). However, high resolution topography can improve the accuracy of tectonic deformation recognition, and it can allow a better understanding of the processes behind it. Surface features such as troughs, scarps, benches, sags, and ridges that define the fault zone are readily visible from high- resolution lidar-derived topography (Oskin et al., 2007; Arrowsmith and Zielke, 2009). Fig. 11 shows the shaded relief and slope maps obtained from a lidar-derived 0.5 m grid cell size DTM of a typical tectonic landform (San Andreas Fault). One can easily recognize the signature of a fault given by the complexity of the surface morphology due to a well-defined sequence of channels, and fault scarps. Cunningham et al. (2006) presented the first airborne lidar survey ever flown in Europe for the purpose of mapping the surface expression of earthquake-prone faults. Detailed topographic maps of the Idrija and Ravne strike-slip faults (NW Slovenia) were derived from lidar, revealing geomorphological and structural features that shed light on the overall architecture and kinematic history of both fault systems. The DTMs extracted in forested areas reveal surface scarps and tectonic landforms in unprecedented detail. Their analysis highlighted the

304


Fig. 11. Shaded relief (a) and slope maps (b) (0.5 m DTM B4 project — Ohio State/NCALM/USGS/UNAVCO processed at OpenTopography) showing fine scale tectonic landforms along the San Andreas Fault (Hilley and Arrowsmith, 2008).

potential contribution of lidar surveying in both low-relief valley terrain and high-relief mountainous terrain for a regional seismic hazard assessment program. Chan et al. (2007) applied high-resolution airborne lidar data to study a segment of the Hsincheng fault in Northern Taiwan. They considered aerial photographs, topographic maps and a 1 m lidar-derived DTM for a highly detailed geomorphic survey. In this case it was possible to detect landforms and subtle but important geomorphic features with high precision and clarity. Kondo et al. (2008) used a high-resolution lidar DEM at 0.5 m for the analysis of the active faults within the urban district of Matsumoto City, Japan, along the Itoigawa–Shizuoka Tectonic Line active fault system. In urban areas, fresh fault scarps are difficult to recognize in the field and on aerial photographs because they are obscured by buildings and other man-made changes to the ground. Thanks to lidar data, however, these authors were able to identify a continuous scarp, up to ~2 m in height, proving the effectiveness of lidar mapping surveys for active fault mapping, and for the identification of surface rupturing associated with large earthquakes. Hilley and Arrowsmith (2008) combined high-resolution topography with geomorphic mapping of active surface processes, and geologic mapping for the analysis of the topographic and erosional response to rock uplift of a small drainage basin along the Dragon's Back pressure ridge along the San Andreas fault in the Carrizo Plain, California. Their results showed a progressive deformation and rock uplift corresponding to increases in channel steepness and basin relief. Their analysis also revealed that channels responded to changes in rock uplift rates over thousands of years, whereas hillslope processes could take more than an order of magnitude longer to adjust to changes in rock uplift rates. Arrowsmith and Zielke (2009) analyzed, using lidar data, the southcentral San Andreas Fault in California to characterize the tectonic geomorphology of this rapidly slipping (~35 mm yr−1) fault zone in a semiarid environment. The landforms clearly show both the most recent slip zones associated with the last several earthquakes and the broad belt of deformation active over the last few to tens of kyr. Their results also suggested that the DTMs derived by airborne laser scanners can provide, for this kind of analysis, useful and complementary information to aerial photography and field-based approaches. Wechsler et al. (2009) proposed a prototype analysis of asymmetry of rock damage based on hypothesized covariation in surficial processes manifest in geomorphic signals along the central San Jacinto Fault of southern California. Several morphometric parameters were compared, including drainage density, on both sides of the fault, using lidar and synthetic aperture radar data. The high-resolution lidar data permitted focus on a single fault trace, eliminating the effects of parallel nearby

faults. The results indicated that there is a correlation between drainage density and proximity to the fault. Zones of structural complexity along the fault display the highest drainage density. Begg and Mouslopoulou (2010) used high-quality lidar data collected across the active Taupo Rift in the Rangitaiki Plains (New Zealand) to recognize active fault traces, discuss the attributes associated with their geometry and kinematics (e.g., strike, vertical displacements and displacement rates), and define regional tectonic deformation. The lidar-derived shaded relief maps with underlying color-ramped DTMs proved to be an effective tool in visualizing landscape elevation variation. Slope maps and detailed contour maps were used in some areas to precisely identify the exact location of geomorphic features across the plains, such as beach ridges, channels and fault traces. Zielke et al. (2010, 2012) recently used a Lidar topographic dataset to obtain previously unavailable, high-resolution data along the 1857 rupture trace caused by the moment magnitude (Mw) 7.9 Fort Tejon earthquake that occurred along the south-central San Andreas Fault, California. Lidar-derived DTMs allowed the identification and measurement of subtle tectonic– geomorphic features, justifying the re-evaluation of surface slip along the Carrizo segment associated with the 1857 and preceding earthquakes (see also Salisbury et al., 2012 for the San Jacinto Fault). Lin et al. (2013) detected subtle tectonic–geomorphic features in densely forested mountains in Japan using a very high-resolution airborne lidar survey. They suggested that for a complete detection of small tectonic–geomorphic features a 0.5 m grid cell size is necessary, and that the DTM visualization “Red Relief Image Map (RRIM)” allows the mapping of all the small features having different sizes, orientations and morphologies, overcoming the major drawbacks of the classic DTM visualizations. 2.5. Volcanoes Few locations on Earth change as dramatically and frequently as active volcanoes (Neri et al., 2008). In similar scenarios, high-resolution topographies can improve the understanding of a variety of volcanic processes with significant implications in hazard and risk assessment (Bisson et al., 2009). Davila et al. (2007) used lidar, the Advanced Spaceborne Thermal Emission and Reflection Radiometer, and Landsat data to identify morphological changes in the drainage system, and map lahar emplacement, at Volcán de Colima (Mexico). Mazzarini et al. (2007) analyzed, through lidar, variations in the roughness and texture of surfaces at a meter scale as a result of two different processes, initial lava cooling and subsequent surface weathering within the area of Mount Etna (Italy). Csatho et al. (2008) used lidar to provide the


first high‐precision topographic map of an active crater, applying data for Erebus volcano (Antarctica). Ventura and Vilardo (2008) used a 1 m DTM derived from high-resolution lidar data in order to determine morphometric and physical parameters of a lava flow that erupted from the Somma-Vesuvius volcano in 1944. Morphometric parameters such as slope, aspect, relative relief, thickness, width, and crosssectional area were calculated, and then the changes in viscosity, velocity and flow rate were estimated. The goal was to recognize different flow surfaces, to reconstruct the flow kinematics, and obtain information on the mechanism of emplacement. Morris et al. (2008) proposed an analysis based on topographic variation (surface roughness), underlining how cell size is important for assessing surface processes, to assess flows at six field sites in Kilauea caldera (Hawaii). These authors considered three data sets at different resolutions: TOPSAR (10 m pixel−1), airborne lidar (1 m pixel− 1), and tripod-mounted lidar (0.02–0.03 m pixel− 1). Their results indicated that features formed during emplacement and modification of the flows exhibit statistically distinct roughness signatures. Favalli et al. (2009) introduced a technique to further refine the accuracy of lidar-derived topographies, and presented a variety of applications at Etna volcano, including degradation of scoria cones (see also Fornaciai et al., 2010), volume calculation of lava flows and surface changes due to tephra deposition. Fig. 12 illustrates an example of lava flows mapped using a combination of high-resolution DEM-difference and lidar intensity (Tarquini and Favalli, 2011). Tarquini et al. (2012) applied a lidar-derived DTM in order to improve the knowledge of lava flows' morphology and emplacement mechanisms. They presented a new semi-automatic procedure for the morphometric analysis of single lava flow units. Their method relies on the automatic processing of the elevation profiles obtained on transects orthogonal to the flow unit axis. A 2 m DTM served as the basis for such analysis. This method may improve the understanding of the emplacement mechanisms of lava flows on Earth and other planets. In the work of Kereszturi et al. (2012), morphometric parameters of lava flows, such as volume, length, thickness and area, were used to quantify the potential susceptibility to lava-flow inundation of New Zealand's densely populated area (Auckland). In this case high-resolution topography was also provided by airborne lidar.

305

Jessop et al. (2012) used a lidar-derived topography of Lascar Volcano (Northern Chile) to carry out a morphological analysis of the deposits formed during the 1993 pyroclastic flow which yield insights about flow dynamics and rheology. Deardorff and Cashman (2012) used a lidar-derived digital topography over a remote location (Collier Cone, Oregon, USA) to get an insight into the emplacement condition of a 1600-year BP lava flow. These authors collected morphometric measurements such as channel width, flow width, flow thickness, and slope all along the flow axis, suggesting the probable emplacement duration and the relative lava supply rate. As an outcome of the above, it is evident that volcanic processes involving topographic changes can be better understood, and therefore better modeled, when lidar-derived DTMs are available, and this is true for a variety of volcano-related mass wasting processes, but also for cinder fall-out or lava movements. Of particular interest is the progress which will be made possible in the near future by the use of high-resolution lidar-derived topographies in the understanding of extraordinarily complex phenomena such as the emplacement of lava flows at the many volcanoes on Earth. 3. Engineered landscapes Engineered landscapes cover as great an extent of the Earth's land surface as do many other globally important ecosystems (Achard et al., 2002; Ellis, 2004; Foley et al., 2005). In such environments, direct anthropic alteration of surface morphology and processes is significant, and it is aimed toward servicing the needs of human populations (Ellis et al., 2006; Ellis, 2011; Tarolli et al., 2014). 3.1. Agricultural and urbanized landscapes within floodplains In engineered landscapes within floodplains, human alteration is reflected by artificial earth-surface features, such as banks, levees and drainage networks (e.g. Dunn and Mackay, 1996; Moussa et al., 2002; Duke et al., 2006; Couturier et al., 2013). It is clear that the proper programming of measures for these areas relies on a correct characterization of the anthropogenic features that influence flood flows, starting

Fig. 12. Slope map obtained from a lidar-derived topography at Mount Etna (Italy). This map, combined with DTM-difference and lidar intensity maps, was used by Tarquini and Favalli (2011) to draw the outline of single lava flow units (reddish hue) emplaced during the 1999 effusive activity, and to check the results of lava flow numerical simulations, with straightforward implications for hazard assessment.

306


Fig. 13. Example of ditch surveyed in the field near Padova, north of Italy (photo courtesy of S. Calligaro).

from large-scale and up-to-date information about channels and ditch networks (Fig. 13), and positioning and geometry of banks. However, this information is usually lacking (Köthe and Bock, 2009). In these contexts, high-resolution digital terrain models have mainly been used for hydrological/hydraulic modeling purposes (e.g. Cobby et al., 2001; French, 2003; Dal Cin et al., 2005; Mandlburger et al., 2009; Sampson et al., 2012), while some other studies have focused on morphological aspects (e.g. Lohani and Mason, 2001; Challis, 2006; Challis and Howard, 2006; Nelson et al., 2006). Bailly et al. (2008) presented a method to delineate artificial network reaches in the agrarian landscape directly from the lidar point cloud. For cultivated landscapes, drainage algorithms used on DTMs are unable to represent anthropogenically modified overland flow paths (Garcia and Camarasa, 1999; Duke et al., 2006). The detection of the network in these contexts benefits from morphological approaches, whereas channellized elements are defined from local morphologies, and the hydrological connectivity is not accounted for. In the work of Passalacqua et al. (2012), the application of their automatic approach for channel network extraction GeoNet (Passalacqua et al., 2010a) has been tested in flat and engineered landscapes. They successfully performed automated channel morphometric analysis using a 1 m lidar-derived DTM including extraction of cross-sections, detection of bank locations, and identification of geomorphic bankfull water surface elevation, to differentiate between natural channels and man-made structures including artificial ditches, roads and bridges across channels. Cazorzi et al. (2013) presented a methodology for minor artificial drainage network (e.g. ditches and channels) detection. The method consists in applying a low-pass filter to the 1 m DTM, providing a smoothed elevation model; by subtracting the DTM to this smoothed map, an approximation of the local relief is achieved, where only small-scale topographic features are preserved. Fig. 14 shows an example of the network extracted according to this methodology. This network can be used to quantify the drainage density (km km−2) and water storage capacity (m3 ha−1) of agricultural lands. This assessment could be a useful tool for flood management: low values of channel storage capacity (and drainage density) can underline a deficit in the

network, and they can outline areas whose hydrological behavior is potentially critical during floods. At this regard, Fig. 15 shows an example of drainage density calculated using the network extracted from the 1 m DTM for an urbanized area in the Veneto floodplain (north of Italy): lower values of drainage density are related to urban areas, while the higher ones are related to agricultural lands. Sofia et al. (2014) tested, on the 1 m DTM, different topographic parameters to verify their suitability for feature extraction. The automatic recognition of levees and road scarps (an example is shown in Fig. 16) can offer a quick and accurate method to improve topographic databases for large-scale applications. The limit of these applications is that they are only suitable in areas where the investigated features represent a clear disturbance to the morphology. For landscapes with complex morphology, where several natural and anthropic features are mixed together, the feature recognition might not be as effective. These fast and automatic approaches can, however, offer tools to update the available databases or to assess measurements previously hard to obtain, such as the drainage density. The latter can also help in the analysis of the influence of changes in land use on changes in flood flows or terrain flood response. 3.2. Landscapes affected by road networks A large part of the anthropic landscape is affected by a dense road network. A well-designed road network allows fast communications, the efficient performance of every aspect of productive activities and good environmental management. Having said that, road networks and their related drainage systems may change the surface flow directions, affecting the surface morphology, and increasing the soil erosion or landslide risk (Reid and Dunne, 1984; Montgomery, 1994; Luce and Black, 1999; Wemple et al., 2001; Borga et al., 2004) for exposed infrastructures and socioeconomic activities. Thanks to high-resolution DTMs (~1 m grid cell size or less) derived from lidar, it is possible to detect in detail the road network, even in densely forested areas (White et al., 2010; Wang et al., 2013). Fig. 17 illustrates a map where a simple methodology (RPII, Relative Path Impact Index, Tarolli et al., 2013) for the automatic recognition of forest road-induced flow direction changes


307

was surveyed in the field by DGPS (white arrows in Fig. 17a). This is an example of how high-resolution topography can be used to interactively assist the final user in the task of surface instability analysis, erosion and landslide risk management for appropriate environmental planning. In the case of road networks, such information may also be useful for the design of structural (drainage cross-ditches) or nonstructural measures (different urban development) for risk mitigation. 4. Future challenges 4.1. Hydrogeomorphic processes High-resolution topography offers the scientific community a unique opportunity to develop new approaches and improve the understanding of Earth surface processes, starting from their typical geomorphic signatures. These advances will provide the basis to improve landscape evolution models, and to parametrize the relevant landscape features into geomorphic transport and hydrologic models (Passalacqua et al., 2010b). The question is whether the high-resolution topography can be integrated in a modeling approach that is able to improve the understanding and the prediction of dynamic hydrogeomorphic processes. Thanks to the availability of high‐resolution topography, the procedures for channel network extraction have been reconsidered, introducing new methodologies and reaching more detailed results than those obtained in the past. Nevertheless, the question is: does the improvement in channel network details reflect also an improvement in the analysis of catchment hydrological responses to rainfall events? If the drainage density is one of the key elements to better understand the climate, geology (e.g. tectonic) and relief (Kirkby, 1980, 1993; Howard, 1997; Tucker and Bras, 1998), what is the improvement in terms of the understanding of different landscapes and their related processes that can be given by such new methodologies for channel extraction from highresolution topography? Providing an answer to these open questions, of course, represents one of the future challenges for the Earth science community. 4.2. Concept of scale

Fig. 14. Aerial photograph (a), shaded relief map calculated from a 1 m DTM (b), and the artificial drainage network derived after thresholding according to the Cazorzi et al. (2013) methodology (c) (Veneto floodplain, north of Italy).

is tested using lidar data. The algorithm is based on the calculation of the drainage area variation in either the presence or absence of anthropic features on hillslopes. It is expressed in logarithmic form, in order to emphasize and map only such areas where an increase of the drainage area is observed due to the presence of roads. The higher the RPII value, the greater is the alteration induced by roads, and therefore the likely related surface instability. In Fig. 17 one can easily recognize the location of likely water flow alteration due to the road through RPII interpretation (blue arrow in Fig. 17c), exactly where such a process

Another critical issue, still unsolved, in the analysis of Earth surface processes is the concept of scale and feature size. An appropriate scale to derive and represent the local morphology has to be selected according to the scale of the features to be detected or the process to be analyzed through physical models. Hengl (2006) suggested that no ideal grid resolution exists, but rather a range of suitable resolutions: in general, one should avoid using a grid cell size that does not comply with the effective scale or inherent properties of the input dataset. The question is if high-resolution topography can be useful for the analysis of Earth surface processes, also when considering modeling approaches. In the case of landsliding, several researchers (Claessens et al., 2005; Tarolli and Tarboton, 2006; Milledge et al., 2012) suggested that the ‘optimal’ scale for an analysis is directly related to the average size of the landslides in that region, and the fact that a slightly coarse DTM (~ 10 m) performs better with the infinite length assumption in the case of shallow landslides. However, these analyses are not enough, and are too restricted to landsliding processes. More effort should be spent along this line, improving case studies, and also focusing on different physical processes. 4.3. Feature detection Many approaches for geomorphic feature extraction are rasterbased. So they are related to the limits and issues of pixel analysis. For example, considering landslide feature detection, it is hard to understand how a single feature (or feature type) may be indicative of the presence of a landslide (Guzzetti et al., 2012). Probably the methodology presented in the work of Van Den Eeckhaut et al. (2012), in which

308


Fig. 15. Drainage density (km km−2) calculated on the artificial drainage network (Veneto floodplain, north of Italy) extracted from a 1 m DTM, according to the Cazorzi et al. (2013) methodology.

landslide crowns have been investigated through automatic object shape recognition from high-resolution topography, will be the way to go in the future. In so doing, noises related to pixel-based analysis should be overcome. 4.4. Multi-temporal data The availability of high-resolution topographic data over the years for a given area offers the opportunity to better detect the geomorphic

changes of a surface. This is the case of landslide movements, detection of landslide mobilizable volumes, analysis of river bed morphology change, and estimation of river sediment budget. Tseng et al. (2013) considered multi-temporal DTMs, derived from lidar surveys carried out in 2005 and 2010, before and after the major typhoon Morakot. In their analysis it was possible to obtain accurate landslide-induced sediment volumes, highlighting the effectiveness of multi-temporal lidar data for this kind of analysis. More effort should be spent in this direction, to better understand the landscape evolution and validate

Fig. 16. Reference features (levee and scarp) and overview of the area (Veneto floodplain, north of Italy) (a), shaded relief map derived from a 1 m DTM (b), and best extraction (c), obtained by residual topography (Sofia et al., 2014) evaluated with a 23 m rectangular kernel.


309

Fig. 17. Field evidence of water flow intercepted by a forest road (Eastern Italian Alps). In detail, the figure illustrates: (a) aerial photograph related to a section of a forest road where there is a clear evidence of water flow interception (white surrounded arrow in the pop-up box), (b) shaded relief map calculated from a lidar-derived 1 m DTM, and (c) the RPII (Relative Path Impact Index) as proposed by Tarolli et al. (2013). The arrow indicates exactly the point on the road where evidence of surface runoff was observed (the color of the road is slightly darker on the runoff path).

modeling approaches. In the coming years there will be an increase of such multi-temporal analysis in different contexts, because of rising availability of data provided by different technologies mounted on different platforms. 4.5. Unmanned Aerial Vehicles An Unmanned Aerial Vehicle (UAV), commonly known as a drone, is an aircraft without a human pilot on board. They have integrated autopilot technology, which gives them semi- or fully-autonomous navigation, flight control and image acquisition capabilities (Hugenholtz et al., 2013). This recent remote sensing technology is growing fast and the scientific community is observing a significantly rising amount of Earth surface analysis using UAVs (e.g. Jaakkola et al., 2010). Hugenholtz et al. (2013) used a small unmanned aircraft system for the feature detection and accuracy assessment of a photogrammetrically-derived DTM. The results highlighted the effectiveness of this technology in Earth surface analysis: the vertical RMSE of the DTM was found to be equivalent to the RMSE of a bare earth lidar DTM for the same site. This suggests that UAV-acquired imagery may provide a low-cost, rapid, and flexible alternative to airborne lidar for geomorphological mapping. UAVs should really offer advances in the understanding of Earth surface processes. They can be easily and cheaply used before and after a natural event or also in real time in order to offer a basis upon which to test models, forecast and understand the evolution of the investigated process. Future challenges will, of course, be the improvement of vehicles (e.g. materials, hardware, capability to carry different sensors and weights), and the context of applications. 4.6. Structure-from-Motion (SfM) A real opportunity and challenge is presented by the low-cost and flexible photogrammetric technique called ‘Structure-from-Motion’ (SfM). SfM uses images acquired from multiple viewpoints in order to reconstitute the three-dimensional geometry of an object or surface (Fonstad et al., 2013). The novelty with respect to traditional photogrammetric methods is the fact that SfM is designed to reconstitute the three-dimensional geometry from randomly acquired images. The fact that with a user-friendly low cost technique such as SfM it is possible to obtain high-resolution and accurate topographic data represents a “revolutionary” advance compared with more expensive technologies and applications (e.g. TLS). Recently, Westoby et al. (2012) compared an SfM-derived DEM with a similar model obtained using terrestrial laser scanning. The results suggested that decimeter-scale vertical accuracy can be achieved using SfM also in areas with complex topography.

Fonstad et al. (2013) examined the applicability of SfM in a fluvial topographic environment. Several hundred images are used to create a DEM with point cloud densities comparable with airborne lidar, and horizontal and vertical accuracy in the order of centimeters. Javernick et al. (2014) as well used SfM in the analysis of a river environment, to model the topography of the complex shallow braided rivers. The results suggested that geo-registration errors of 0.04 m (planar) and 0.10 m (elevation) and vertical surface errors of 0.10 m in nonvegetation areas can be achieved from a dataset of photographs taken at 600 and 800 m above ground level. Of course, one should consider not only the context of river morphology; SfM should be useful, for example, in mapping landslides. One can arrange a lot of low-cost and logistically simple surveys in the same study area, obtaining an accurate multi-temporal DEM as the basis to better understand the landsliding process and quantify landslide volume. We are just at the beginning of SfM applications, but it should be not a surprise in a few years to see users who are able, with a simple smartphone camera, to create accurate high-resolution DEMs or DSMs for any purpose. 4.7. Earth surface processes in the Anthropocene High-resolution topography has been found to be useful not only for natural landscapes, but also for engineered landscapes, where the anthropic forcing related to human activities (e.g. urbanization, road network, and agricultural practices) may affect natural processes. The scientific community is debating the fact that we are now living in a new geological epoch, the Anthropocene (Zalasiewicz et al., 2008, 2011; Zalasiewicz, 2013), where human activities may leave a significant signature on the Earth, by altering its morphology, climate, and ecosystems. The increase of the population has been related to a progressive increase of intensive agriculture and urbanization. This anthropogenic forcing has deeply affected the environment, inducing or reducing erosion (Tarolli et al., 2014). In this respect, high-resolution topography could play a really strategic and helpful role, through the recognition of human-induced geomorphic and anthropogenic features, and the connected erosion. Great effort should be spent on this topic. This represents a real challenge for better understanding the landscape evolution in the epoch in which we are now living. 5. Final remarks The progressive rise of new remote sensing technologies (e.g. airborne and terrestrial laser scanners), and the related availability of high-resolution topographic data is offering the Earth science community great opportunities and challenges to better understand Earth

310


surface processes from their characteristic topographic signatures. In the last decade several researchers have used high-resolution topography to improve our knowledge of physical processes (e.g. geologic, hydrogeomorphic, volcanic and tectonic processes) in natural landscapes. New insights have been provided for landslide detection and quantification of landslide volume, recognition of river bed morphology, detection of faults, and lava flow path, and recognition of the real channel network. In addition to these advances, high-resolution topography also offers an opportunity to develop new methodologies to manage such detailed topographic information. New topographic indexes and algorithms have been tested, with the aim of providing the final user with a useful tool that is able to objectively process a surface, and then detect any topographic signatures. However, the scientific community has also focused its attention on engineered landscapes. Many regions of our planet are now significantly affected by human activities. These activities are making signatures on the morphology with direct consequences for Earth surface processes. High-resolution topography can really help to recognize in detail such signatures, providing useful data to better understand processes that are affected by humans. The recognition and analysis of these human-induced changes, signatures and processes probably represent this century's most interesting challenge for the scientific community. Such analysis will help in scheduling appropriate environmental planning for sustainable development, to mitigate the consequences of anthropogenic alteration and, of course, to better understand the evolution of our Planet. Acknowledgments The author wishes to thank Takashi Oguchi and Ching-Weei Lin for the comments raised during the review stage, and Ramon Arrowsmith, Simone Tarquini, and Paola Passalacqua for useful advices raised during the pre- and post-review stage. The author would also like to thank the colleague Giancarlo Dalla Fontana and collaborators (Giulia Sofia, Simone Calligaro, and Massimo Prosdocimi) involved in the digital terrain analysis group at the TESAF Department of the University of Padova, for stimulating the debate in the last few years. References Achard, F., Eva, H.D., Stibig, H.J., Mayaux, P., Gallego, J., Richards, T., 2002. Determination of deforestation rates of the world's humid tropical forests. Science 297, 999–1002. Ardizzone, F., Cardinali, M., Galli, M., Guzzetti, F., Reichenbach, P., 2007. Identification and mapping of recent rainfall-induced landslides using elevation data collected by airborne lidar. Nat. Hazards Earth Syst. Sci. 7, 637–650. Arrowsmith, J.R., Zielke, O., 2009. Tectonic geomorphology of the San Andreas Fault zone from high resolution topography: an example from the Cholame segment. Geomorphology 113, 70–81. Bailly, J.S., Lagacherie, P., Millier, C., Puech, C., Kosuth, P., 2008. Agrarian landscapes linear features detection from lidar: application to artificial drainage networks. Int. J. Remote Sens. 29, 3489. Band, L.E., 1986. Topographic partition of watersheds with digital elevation models. Water Resour. Res. 22, 15–24. Begg, J.G., Mouslopoulou, V., 2010. Analysis of late Holocene faulting within an active rift using lidar, Taupo Rift, New Zealand. J. Volcanol. Geotherm. Res. 190, 152–167. Bisson, M., Behncke, B., Fornaciai, A., Neri, M., 2009. Lidar‐based digital terrain analysis of an area exposed to the risk of lava flow invasion: the Zafferana Etnea territory, Mt. Etna (Italy). Nat. Hazards 50, 321–334. Booth, A.M., Roering, J.J., Perron, J.T., 2009. Automated landslide mapping using spectral analysis and high-resolution topographic data: Puget Sound lowlands, Washington, and Portland Hills, Oregon. Geomorphology 109, 132–147. Booth, A.M., Roering, J.J., Rempel, A.W., 2013. Topographic signatures and a general transport law for deep-seated landslides in a landscape evolution model. J. Geophys. Res. Earth Surf. 118, 603–624. Borga, M., Tonelli, F., Selleroni, J., 2004. A physically-based model of the effects of forest roads on slope stability. Water Resour. Res. 40, W12202. http://dx.doi.org/10.1029/ 2004WR003238. Burns, W.J., Coe, J.A., Kaya, B.S., Ma, L., 2010. Analysis of elevation changes detected from multi-temporal lidar surveys in forested landslide terrain in western Oregon. Environ. Eng. Geosci. 26, 315–341. Cavalli, M., Tarolli, P., 2011. Application of lidar technology for rivers analysis. Ital. J. Eng. Geol. Environ. 1, 33–44. http://dx.doi.org/10.4408/IJEGE.2011-01.S-03 (Special Issue). Cavalli, M., Tarolli, P., Marchi, L., Dalla Fontana, G., 2008. The effectiveness of airborne lidar data in the recognition of channel bed morphology. Catena 73, 249–260.

Cavalli, M., Trevisani, S., Comiti, F., Marchi, L., 2013. Geomorphometric assessment of spatial sediment connectivity in small Alpine catchments. Geomorphology 188, 31–41. Cazorzi, F., Dalla Fontana, G., De Luca, A., Sofia, G., Tarolli, P., 2013. Drainage network detection and assessment of network storage capacity in agrarian landscape. Hydrol. Process. 27, 541–553. Challis, K., 2006. Airborne laser altimetry in alluviated landscapes. Archaeol. Prospect. 13, 103–127. Challis, K., Howard, A.J., 2006. A review of trends within archaeological remote sensing in alluvial environments. Archaeol. Prospect. 13, 231–240. Chan, Y.C., Chen, Y.G., Shih, T.Y., Huang, C., 2007. Characterizing the Hsincheng active fault in northern Taiwan using airborne lidar data: detailed geomorphic features and their structural implications. J. Asian Earth Sci. 31, 303–316. Claessens, L., Heuvelink, G.B.M., Schoorl, J.M., Veldkamp, A., 2005. DEM resolution effects on shallow landslide hazard and soil redistribution modeling. Earth Surf. Process. Landf. 30, 461–477. Cobby, D.M., Mason, D.C., Davenport, I.J., 2001. Image processing of airborne scanning laser altimetry data for improved river flood modeling. ISPRS J. Photogramm. Remote Sens. 56, 121–138. Couturier, A., Daroussin, J., Darboux, F., Souchère, V., Le Bissonnais, Y., Cerdan, O., King, D., 2013. Improvement of surface flow network prediction for the modeling of erosion processes in agricultural landscapes. Geomorphology 183, 120–129. Csatho, B., Schenk, T., Kyle, P., Wilsond, T., Krabille, W.B., 2008. Airborne laser swath mapping of the summit of Erebus volcano, Antarctica: applications to geological mapping of a volcano. J. Volcanol. Geoth. Res. 177, 531–548. Cunningham, D., Grebby, S., Tansey, K., Gosar, A., Kastelic, V., 2006. Application of airborne lidar to mapping seismogenic faults in forested mountainous terrain, southeastern Alps, Slovenia. Geophys. Res. Lett. 33, L20308. http://dx.doi.org/10.1029/ 2006GL027014. Dal Cin, C., Moens, L., Dierickx, P., Bastin, G., Zech, Y., 2005. An integrated approach for realtime floodmap forecasting on the Belgian Meuse river. Nat. Hazards 36, 237–256. Davila, N., Capra, L., Gavilanes Ruiz, J.C., Varley, N., Norini, G., Vazquez, A.G., 2007. Recent lahars at Volcán de Colima (Mexico): drainage variation and spectral classification. J. Volcanol. Geotherm. Res. 165, 127–141. De Rose, R.C., Basher, L.R., 2011. Measurement of river bank and cliff erosion from sequential lidar and historical aerial photography. Geomorphology 126, 132–147. Deardorff, N.D., Cashman, K.V., 2012. Emplacement conditions of the c. 1,600-year BP Collier Cone lava flow, Oregon: a lidar investigation. Bull. Volcanol. 74, 2051–2066. DeLong, S.B., Prentice, C.S., Hilley, G.E., Ebert, Y., 2012. Multitemporal ALSM change detection, sediment delivery, and process mapping at an active earthflow. Earth Surf. Proc. Land. 37, 262–272. Derron, M.-H., Jaboyedoff, M., 2010. Preface “lidar and DEM techniques for landslides monitoring and characterization”. Nat. Hazards Earth Syst. Sci. 10, 1877–1879. Dietrich, W.E., Wilson, C.J., Montgomery, D.R., McKean, J., 1993. Analysis of erosion thresholds, channel networks and landscape morphology using a digital terrain model. J. Geol. 3, 161–180. Duke, G., Kienzle, S., Johnson, D., Byrne, J., 2006. Incorporating ancillary data to refine anthropogenically modified overland flow paths. Hydrol. Process. 20, 1827–1843. Dunn, S.M., Mackay, R., 1996. Modelling the hydrological impacts of open ditch drainage. J. Hydrol. 179, 37–66. Ellis, E.C., 2004. Long-term ecological changes in the densely populated rural landscapes of China. In: DeFries, R.S., Asner, G.P., Houghton, R.A. (Eds.), Ecosystems and Land Use Change. American Geophysical Union, Washington, DC, pp. 303–320. Ellis, E.C., 2011. Anthropogenic transformation of the terrestrial biosphere. Phil. Trans. R. Soc. A 369, 1010–1035. Ellis, E.C., Wang, H., Xiao, H., Peng, K., Liu, X.P., Li, S.C., Ouyang, H., Cheng, X., Yang, L.Z., 2006. Measuring long-term ecological changes in densely populated landscapes using current and historical high resolution imagery. Remote Sens. Environ. 100, 457–473. Evans, I.S., 1979. An integrated system of terrain analysis and slope mapping. Final Report on Grant DA-ERO-591-73-G0040. University of Durham, England. Favalli, M., Fornaciai, A., Pareschi, M.T., 2009. Lidar strip adjustment: application to volcanic areas. Geomorphology 111, 123–135. Foley, J.A., DeFries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter, S.R., 2005. Global consequences of land use. Science 309, 570–574. Fonstad, M.A., Dietrich, J.T., Courville, B.C., Jensen, J.L., Carbonneau, P.E., 2013. Topographic structure from motion: a new development in photogrammetric measurement. Earth Surf. Process. Landf. 38, 421–430. Fornaciai, A., Behncke, B., Favalli, M., Neri, M., Tarquini, S., Boschi, E., 2010. Detecting short-term evolution of Etnean scoria cones: a lidar-based approach. Bull. Volcanol. 72, 1209–1222. French, J.R., 2003. Airborne lidar in support of geomorphological and hydraulic modeling. Earth Surf. Process. Landf. 28, 321–335. Garcia, M., Camarasa, A., 1999. Use of geomorphological units to improve drainage network extraction from a DEM: comparison between automated extraction and photointerpretation methods in the Carraixet catchment (Valencia, Spain). Int. J. Appl. Earth Obs. Geoinf. 1, 187–195. Glenn, N.F., Streutker, D.R., Chadwick, D.J., Thackray, G.D., Dorsch, S.J., 2006. Analysis of lidar-derived topographic information for characterizing and differentiating landslide morphology and activity. Geomorphology 73, 131–148. Guzzetti, F., Mondini, A., Cardinali, M., Fiorucci, F., Santangelo, M., Chang, K.-T., 2012. Landslide inventory maps: new tools for an old problem. Earth Sci. Rev. 112, 42–66. Heine, R.A., Lant, C.L., Sengupta, R.R., 2004. Development and comparison of approaches for automated mapping of stream channel networks. Ann. Assoc. Am. Geogr. 94, 477–490. Hengl, T., 2006. Finding the right pixel size. Original Research Article Comput. Geosci. 32, 1283–1298.

P. Tarolli / Geomorphology 216 (2014) 295–312 Henkle, J.E., Wohl, E., Beckman, N., 2011. Locations of channel heads in the semiarid Colorado Front Range, USA. Geomorphology 129, 309–319. Heritage, G.L., Milan, D.J., 2009. Terrestrial Laser Scanning of grain roughness in a gravelbed river. Geomorphology 113, 4–11. Hilldale, R.C., Raff, D., 2008. Assessing the ability of airborne lidar to map river bathymetry. Earth Surf. Process. Landf. 33, 773–783. Hilley, G.E., Arrowsmith, J.R., 2008. Geomorphic response to uplift along the Dragon's Back pressure ridge, Carrizo Plain, California. Geology 36, 367–370. Hodge R. 2007. Grain-scale approaches to modelling and measuring fluvial sedimentology and bedload transport. Unpublished PhD Thesis, University of Cambridge; 315 pp. Hodge, R., Brasington, J., Richards, K., 2009. In situ characterization of grain-scale fluvial morphology using Terrestrial Laser Scanning. Earth Surf. Process. Landf. 34, 954–968. Howard, A.D., 1997. Badland morphology and evolution: interpretation using a simulation model. Earth Surf. Process. Landf. 22, 211–227. Hugenholtz, C.H., Whitehead, K., Brown, O.W., Barchyn, T.E., Moorman, B.J., LeClair, A., Riddell, K., Hamilton, T., 2013. Geomorphological mapping with a small unmanned aircraft system (sUAS): feature detection and accuracy assessment of a photogrammetrically-derived digital terrain model. Geomorphology 194, 16–24. Ijjasz-Vasquez, E.J., Bras, R.L., 1995. Scaling regimes of local slope versus contributing area in digital elevation models. Geomorphology 12, 299–311. Iwahashi, J., Kamiya, I., Yamagishi, H., 2012. High-resolution DEMs in the study of rainfalland earthquake-induced landslides: use of a variable window size method in digital terrain analysis. Geomorphology 153–154, 29–38. Jaakkola, A., Hyyppä, J., Kukko, A., Yu, X., Kaartinen, H., Lehtomäki, M., Lin, Y., 2010. A lowcost multi-sensoral mobile mapping system and its feasibility for tree measurements. ISPRS J. Photogramm. Remote Sens. 65, 514–522. Jaboyedoff, M., Oppikofer, T., Abellán, A., Derron, M.-H., Loye, A., Metzger, R., Pedrazzini, A., 2012. Use of lidar in landslide investigations: a review. Nat. Hazards 61, 5–28. Jaeger, K.L., Montgomery, D.R., Bolton, S.M., 2007. Channel and perennial flow initiation in headwater streams: management implications of variability in source-area size. Environ. Manag. 40, 775–786. Javernick, L., Brasington, J., Caruso, B., 2014. Modelling the topography of shallow braided rivers using Structure-from-Motion photogrammetry. Geomorphology 213, 166–182. Jefferson, A.J., McGee, R.W., 2013. Channel network extent in the context of historical land use, flow generation processes, and landscape evolution in the North Carolina Piedmont. Earth Surf. Process. Landf. 38, 601–613. Jessop, D.E., Kelfoun, K., Labazuy, P., Mangeney, A., Roche, O., Tillier, J.-L., Trouillet, M., Thibault, G., 2012. Lidar derived morphology of the 1993 Lascar pyroclastic flow deposits, and implication for flow dynamics and rheology. J. Volcanol. Geotherm. Res. 245–246, 81–97. Jones, A.F., Brewer, P.A., Johnstone, E., Macklin, M.G., 2007. High resolution interpretative geomorphological mapping of river valley environments using airborne lidar data. Earth Surf. Process. Landf. 21, 1574–1592. Kasai, M., Ikeda, M., Asahina, T., Fujisawa, K., 2009. Lidar-derived DEM evaluation of deepseated landslides in a steep and rocky region of Japan. Geomorphology 113, 57–69. Kereszturi, G., Procter, J., Cronin, S.J., Németh, K., Bebbington, M., Lindsay, J., 2012. Lidarbased quantification of lava flow susceptibility in the City of Auckland (New Zealand). Remote Sens. Environ. 125, 198–213. Kirkby, M.J., 1980. The stream head as a significant geomorphic threshold. In: Coates, D.R., Vitek, J.D. (Eds.), Thresholds in Geomorphology. Allen and Unwin, Winchester, Mass, pp. 53–73. Kirkby, M.J., 1993. Long term interactions between networks and hillslopes. In: Beven, K. , Kirkby, M.J. (Eds.), Channel Network Hydrology. John Wiley, New York, pp. 255–293. Kondo, H., Toda, S., Okumura, K., Takada, K., Chiba, T., 2008. A fault scarp in an urban area identified by lidar survey: a case study on the Itoigawa-Shizuoka Tectonic Line, central Japan. Geomorphology 101, 731–739. Köthe, R., Bock, M., 2009. Preprocessing of Digital Elevation Models – derived from Laser Scanning and Radar Interferometry – for Terrain Analysis in Geosciences. Proceedings of Geomorphometry, Zurich, Switzerland, 31 August–2 September. Lashermes, B., Foufoula-Georgiou, E., Dietrich, W.E., 2007. Channel network extraction from high resolution topography using wavelets. Geophys. Res. Lett. 34, L23S04. http://dx.doi.org/10.1029/2007GL031140. Legleiter, C.J., 2012. Remote measurement of river morphology via fusion of lidar topography and spectrally based bathymetry. Earth Surf. Process. Landf. 37, 499–518. Lin, C.W., Tseng, C.-M., Tseng, Y.-H., Fei, L.-Y., Hsieh, Y.-C., Tarolli, P., 2013. Recognition of large scale deep-seated landslides in forest areas of Taiwan using high resolution topography. J. Asian Earth Sci. 62, 389–400. Lin, Z., Kaneda, H., Mukoyama, S., Asada, N., Chiba, T., 2013. Detection of subtle tectonic– geomorphic features in densely forested mountains by very high-resolution airborne lidar survey. Geomorphology 182, 104–115. Lohani, B., Mason, D.C., 2001. Application of airborne scanning laser altimetry to the study of tidal channel geomorphology. ISPRS J. Photogramm. Remote Sens. 56, 100–120. Luce, C.H., Black, T.A., 1999. Sediment production from forest roads in western Oregon. Water Resour. Res. 35, 2561–2570. Mackey, B.H., Roering, J.J., 2011. Sediment yield, spatial characteristics, and the long-term evolution of active earthflows determined from airborne lidar and historical aerial photographs, Eel River, California. Geol. Soc. Am. Bull. 123, 1560–1576. Magirl, C.S., Webb, R.H., Griffiths, P.G., 2005. Changes in the water surface profile of the Colorado River in Grand Canyon, Arizona, between 1923 and 2000. Water Resour. Res. 41. http://dx.doi.org/10.1029/2003WR002519 (Art. No. W05021). Mandlburger, G., Hauer, C., Höfle, B., Habersack, H., Pfeifer, N., 2009. Optimisation of lidar derived terrain models for river flow modelling. Hydrol. Earth Syst. Sci. 13, 1453–1466. Marcus, W.A., Fonstad, M.A., 2010. Remote sensing of rivers: the emergence of a subdiscipline in the river sciences. Earth Surf. Proc. Land. 35, 1867–1872.

311

Mark, D.M., 1988. Network models in geomorphology. In: Anderson, M.G. (Ed.), Modeling Geomorphological Systems. John Wiley, New York, pp. 73–97. Mazzarini, F., Pareschi, M.T., Favalli, M., Isola, I., Tarquini, S., Boschi, E., 2007. Lava flow identification and aging by means of lidar intensity: Mount Etna case. J. Geophys. Res. 112, B02201. http://dx.doi.org/10.1029/2005JB004166. Mckean, J., Roering, J., 2004. Objective landslide detection and surface morphology mapping using high-resolution airborne laser altimetry. Geomorphology 57, 331–351. McKean, J., Nagel, D., Tonina, D., Bailey, P., Wright, C.W., Bohn, C., Nayegandhi, A., 2009. Remote sensing of channels and riparian zones with a narrow-beam aquatic–terrestrial lidar. Remote Sens. 1, 1065–1096. Milledge, D.G., Griffiths, D.V., Lane, S.N., Warburton, J., 2012. Limits on the validity of infinite length assumptions for modelling shallow landslides. Earth Surf. Process. Landf. 37, 1158–1166. Montgomery, D.R., 1994. Road surface drainage, channel initiation, and slope instability. Water Resour. Res. 30, 1925–1932. Montgomery, D.R., Dietrich, W.E., 1992. Channel initiation and the problem of landscape scale. Science 255, 826–830. Montgomery, D.R., Foufoula-Georgiou, E., 1993. Channel network source representation using digital elevation models. Water Resour. Res. 29, 3925–3934. Morris, A.R., Anderson, F.S., Mouginis-Mark, P.J., Haldemann, A.F.C., Brooks, B.A., Foster, J., 2008. Roughness of Hawaiian volcanic terrains. J. Geophys. Res. 113, E12007. http:// dx.doi.org/10.1029/2008JE003079. Moussa, R., Voltz, M., Andrieux, P., 2002. Effects of the spatial organization of agricultural management on the hydrological behaviour of a farmed catchment during flood events. Hydrol. Process. 16, 393–412. Nelson, P.A., Smith, J.A., Miller, A.J., 2006. Evolution of channel morphology and hydrologic response in an urbanizing drainage basin. Earth Surf. Process. Landf. 31, 1063–1079. Neri, M., Mazzarini, F., Tarquini, S., Bisson, M., Isola, I., Behncke, B., Pareschi, M.T., 2008. The changing face of Mount Etna's summit area documented with lidar technology. Geophys. Res. Lett. 35, L09305. http://dx.doi.org/10.1029/2008GL033740. Nitsche, M., Turowski, J.M., Badoux, A., Rickenmann, D., Kohoutek, T.K., Pauli, M., Kirchner, J.W., 2013. Range imaging: a new method for high-resolution topographic measurements in small- and medium-scale field sites. Earth Surf. Proc. Land. 38, 810–825. Notebaert, B., Verstraeten, G., Govers, G., Poesen, J., 2009. Qualitative and quantitative applications of lidar imagery in fluvial geomorphology. Earth Surf. Process. Landf. 34, 217–231. O'Callaghan, J.F., Mark, D.M., 1984. The extraction of drainage networks from digital elevation data. Computer Vision. Graph. Image Process. 28, 323–344. Orlandini, S., Moretti, G., Franchini, M., Aldighieri, B., Testa, B., 2003. Path-based methods for the determination of nondispersive drainage directions in grid-based digital elevation models. Water Resour. Res. 39, 1144. http://dx.doi.org/10.1029/ 2002WR001639. Orlandini, S., Tarolli, P., Moretti, G., Dalla Fontana, G., 2011. On the prediction of channel heads in a complex alpine terrain using gridded elevation data. Water Resour. Res. 47, W02538. http://dx.doi.org/10.1029/2010WR009648. Oskin, M.E., Le, K., Strane, M.D., 2007. Quantifying fault-zone activity in arid environments with high-resolution topography. Geophys. Res. Lett. 34, L23S05. http://dx.doi.org/10. 1029/2007GL031295. Passalacqua, P., Do Trung, T., Foufoula-Georgiou, E., Sapiro, G., Dietrich, W.E., 2010a. A geometric framework for channel network extraction from lidar: Nonlinear diffusion and geodesic paths. J. Geophys. Res. 115, F01002. http://dx.doi.org/10.1029/ 2009JF001254. Passalacqua, P., Tarolli, P., Foufoula-Georgiou, E., 2010b. Testing space-scale methodologies for automatic geomorphic feature extraction from lidar in a complex mountainous landscape. Water Resour. Res. 46, W11535. http://dx.doi.org/10. 1029/2009WR008812. Passalacqua, P., Belmont, P., Foufoula-Georgiou, E., 2012. Automatic channel network and geomorphic feature extraction in flat and engineered landscapes. Water Resour. Res. 48, W03528. http://dx.doi.org/10.1029/2011WR010958. Pelletier, J.D., 2013. A robust, two-parameter method for the extraction of drainage networks from high-resolution digital elevation models (DEMs): evaluation using synthetic and real-world DEMs. Water Resour. Res. 49, 1–15. Pirotti, F., Tarolli, P., 2010. Suitability of lidar point density and derived landform curvature maps for channel network extraction. Hydrol. Process. 24, 1187–1197. Prima, O.D.A., Echigo, A., Yokoyama, R., Yoshida, T., 2006. Supervised landform classification of Northeast Honshu from DEM-derived thematic maps. Geomorphology 78, 373–386. Quinn, P., Beven, K., Chevallier, P., Planchon, O., 1991. The prediction of hillslope flow paths for distributed hydrological modeling using digital terrain models. Hydrol. Process. 5, 59–79. Razak, K.A., Straatsma, M.W., Van Westen, C.J., Malet, J.-P., De Jong, S.M., 2011. Airborne laser scanning of forested landslides characterization: terrain model quality and visualization. Geomorphology 126, 186–200. Reid, L.M., Dunne, T., 1984. Sediment production from forest road surfaces. Water Resour. Res. 20, 1753–1761. Rodriguez‐Iturbe, I., Rinaldo, A., 1997. Fractal river basins. Chance and Self‐Organization. Cambridge Univ. Press, New York, p. 528. Roering, J.J., Stimely, L.L., Mackey, B.H., Schmidt, D.A., 2009. Using DInSAR, airborne lidar, and archival air photos to quantify landsliding and sediment transport. Geophys. Res. Lett. 36, L19402. http://dx.doi.org/10.1029/2009GL040374. Roering, J.J., Mackey, B.H., Marshall, J.A., Sweeney, K.E., Deligne, N.I., Booth, A.M., Handwerger, A.L., Cerovski-Darriau, C., 2013. ‘You are HERE’: connecting the dots with airborne lidar for geomorphic fieldwork. Geomorphology 200, 172–183 (l). Salisbury, J.B., Rockwell, T.K., Middleton, T.J., Hudnut, K.W., 2012. Lidar and field observations of slip distribution for the most recent surface ruptures along the Central San Jacinto Fault. Bull. Seismol. Soc. Am. 102, 598–619.

312


Sampson, C.C., Fewtrell, T.J., Duncan, A., Shaad, K., Horritt, M.S., Bates, P.D., 2012. Use of terrestrial laser scanning data to drive decimetric resolution urban inundation models. Adv. Water Resour. 41, 1–7. Schulz, W.H., 2007. Landslide susceptibility revealed by lidar imagery and historical records, Seattle, Washington. Eng. Geol. 89, 67–87. Sekiguchi, T., Sato, H.P., 2004. Mapping of micro topography using airborne laser scanning. Landslides 3, 195–202. Slatton, K.C., Carter, W.E., Shrestha, R.L., Dietrich, W.E., 2007. Airborne laser swath mapping: achieving the resolution and accuracy required for geosurficial research. Geophys. Res. Lett. 34, L23S10. http://dx.doi.org/10.1029/2007GL031939. Sofia, G., Tarolli, P., Cazorzi, F., Dalla Fontana, G., 2011. An objective approach for feature extraction: distribution analysis and statistical descriptors for scale choice and channel network identification. Hydrol. Earth Syst. Sci. 15, 1387–1402. Sofia, G., Pirotti, F., Tarolli, P., 2013. Variations in multiscale curvature distribution and signatures of lidar DTM errors. Earth Surf. Process. Landf. 38, 1116–1134. Sofia, G., Dalla Fontana, G., Tarolli, P., 2014. High-resolution topography and anthropogenic feature extraction: testing geomorphometric parameters in floodplains. Hydrol. Process. 28, 2046–2061. Stock, J., Dietrich, W.E., 2003. Valley incision by debris flows: evidence of a topographic signature. Water Resour. Res. 39, 1089. http://dx.doi.org/10.1029/2001WR001057. Tarboton, D.G., 1989. The analysis of river basins and channel networks using digital terrain data. Sc.D. Thesis M.I.T., Cambridge, MA (also available as Tarboton D. G., R. L. Bras and I. Rodriguez-Iturbe, (same title)), Technical Report no 326, Ralph M. Parsons Laboratory for Water Resources and Hydrodynamics. Department of Civil Engineering, M.I.T. (September 1989). Tarboton, D.G., 1997. A new method for the determination of flow directions and contributing areas in grid digital elevation models. Water Resour. Res. 33, 309–319. Tarboton, D.G., 2003. Terrain analysis using digital elevation models in hydrology. ESRI Users Conference, San Diego, 7–11 July, 2003. Tarboton, D.G., Bras, R.L., Rodriguez-Iturbe, I., 1991. On the extraction of channel network from digital elevation data. Hydrol. Process. 5, 81–100. Tarolli, P., Dalla Fontana, G., 2009. Hillslope to valley transition morphology: new opportunities from high resolution DTMs. Geomorphology 113, 47–56. Tarolli, P., Tarboton, D.G., 2006. A new method for determination of most likely landslide initiation points and the evaluation of digital terrain model scale in terrain stability mapping. Hydrol. Earth Syst. Sci. 10, 663–677. Tarolli, P., Arrowsmith, J.R., Vivoni, E.R., 2009. Understanding earth surface processes from remotely sensed digital terrain models. Geomorphology 113, 1–3. Tarolli, P., Sofia, G., Dalla Fontana, G., 2012. Geomorphic features extraction from highresolution topography: landslide crowns and bank erosion. Nat. Hazards 61, 65–83. Tarolli, P., Calligaro, S., Cazorzi, F., Dalla Fontana, G., 2013. Recognition of surface flow processes influenced by roads and trails in mountain areas using high-resolution topography. Eur. J. Remote Sens. 46, 176–197. Tarolli, P., Preti, F., Romano, N., 2014. Terraced landscapes: from an old best practice to a potential hazard for soil degradation due to land abandonment. Anthropocene. http://dx.doi.org/10.1016/j.ancene.2014.03.002. Tarquini, S., Favalli, M., 2011. Mapping and DOWNFLOW simulation of recent lava flow fields at Mount Etna. J. Volcanol. Geoth. Res. 204, 27–39. Tarquini, S., Favalli, M., Mazzarini, F., Isola, I., Fornaciai, A., 2012. Morphometric analysis of lava flow units: case study of lidar-derived topography at Mount Etna, Italy. J. Volcanol. Geotherm. Res. 235–236, 11–22. Thoma, D.P., Gupta, S.C., Bauerc, M.E., Kirchoff, C.E., 2005. Airborne laser scanning for riverbank erosion assessment. Remote Sens. Environ. 95, 493–501. Thommeret, N., Bailly, J.S., Puech, C., 2010. Robust extraction of thalwegs network from DTM: application on badlands. Hydrol. Earth Syst. Sci. 14, 1527–1536. Trevisani, S., Cavalli, M., Marchi, L., 2010. Reading the bed morphology of a mountain stream: a geomorphometric study on high-resolution topographic data. Hydrol. Earth Syst. Sci. 14, 393–405. Tseng, C.-M., Lin, C.-W., Stark, C.P., Liu, J.-K., Fei, L.-Y., Hsieh, Y.-C., 2013. Application of a multi-temporal, lidar-derived, digital terrain model in a landslide-volume estimation. Earth Surf. Process. Landf. 38, 1587–1601.

Tucker, G.E., Bras, R.L., 1998. Hillslope processes, drainage density and landscape morphology. Water Resour. Res. 34, 2751–2764. Tucker, G.E., Catani, F., Rinaldo, A., 2001. Statistical analysis of drainage density from digital terrain data. Geomorphology 36, 187–202. Van Den Eeckhaut, M., Poesen, J., Verstraeten, G., Vanacker, V., Moeyersons, J., Nyssen, J., Van Beek, L.P.H., 2005. The effectiveness of hillshade maps and expert knowledge in mapping old deep-seated landslides. Geomorphology 67, 351–363. Van Den Eeckhaut, M., Poesen, J., Verstraeten, G., Vanacker, V., Nyssen, J., Moeyersons, J., Van Beek, L.P.H., Vandekerckhove, L., 2007. Use of lidar-derived images for mapping old landslides under forest. Earth Surf. Process. Landf. 32, 754–769. Van Den Eeckhaut, M., Kerle, N., Poesen, J., Hervás, J., 2012. Object-oriented identification of forested landslides with derivatives of single pulse lidar data. Geomorphology 173–174, 30–42. Ventura, G., Vilardo, G., 2008. Emplacement mechanism of gravity flows inferred from high resolution lidar data: the 1944 Somma-Vesuvius lava flow (Italy). Geomorphology 95, 223–235. Ventura, G., Vilardo, G., Terranova, C., Sessa, E.B., 2011. Tracking and evolution of complex active landslides by multi-temporal airborne lidar data: the Montaguto landslide (Southern Italy). Remote Sens. Environ. 115, 3237–3248. Vianello, A., D'Agostino, V., 2007. Bankfull width and morphological units in an alpine stream of the dolomites (Northern Italy). Geomorphology 83, 266–281. Vianello, A., Cavalli, M., Tarolli, P., 2009. Lidar-derived slopes for headwater channel network analysis. Catena 76, 97–106. Wang, J., González-Jorge, H., Lindenbergh, R., Arias-Sánchez, P., Menenti, M., 2013. Automatic estimation of excavation volume from laser mobile mapping data for mountain road widening. Remote Sens. 5, 4629–4651. Wechsler, N., Rockwell, T.K., Ben-Zion, Y., 2009. Application of high resolution DEM data to detect rock damage from geomorphic signals along the central San Jacinto Fault. Geomorphology 113, 82–96. Wemple, B.C., Swanson, F.J., Jones, J.A., 2001. Forest roads and geomorphic process interactions, Cascade Range, Oregon. Earth Surf. Proc. Land. 26, 191–204. Westoby, M.J., Brasington, J., Glasser, N.F., Hambrey, M.J., Reynolds, J.M., 2012. ‘Structurefrom-Motion’ photogrammetry: a low-cost, effective tool for geoscience applications. Geomorphology 179, 300–314. Wheaton, J.M., Brasington, J., Darby, S.E., Sear, D., 2010. Accounting for uncertainty in DEMs from repeat topographic surveys: improved sediment budgets. Earth Surf. Process. Landf. 35, 136–156. White, R.A., Dietterick, B.C., Mastin, T., Strohman, R., 2010. Forest roads mapped using LiDAR in steep forested terrain. Remote Sens. 2, 1120–1141. Wood, J.D., 1996. The geomorphological characterisation of digital elevation models. Ph.D. Thesis University of Leicester. Yokoyama, R., Sirasawa, M., Pike, R., 2002. Visualizing topography by openness: a new application of image processing to digital elevation models. Photogramm. Eng. Remote Sens. 68, 257–265. Zalasiewicz, J., 2013. The epoch of humans. Nat. Geosci. 6, 8–9. Zalasiewicz, J., Williams, M., Smith, A.G., Barry, T.L., Coe, A.L., Bown, P.R., Brenchley, P., Cantrill, D., Gale, A., Gibbard, P., Gregory, F.J., Hounslow, M.W., Kerr, A.C., Pearson, P., Knox, R., Powell, J., Waters, C., Marshall, J., Oates, M., Rawson, P., Stone, P., 2008. Are we now living in the Anthropocene? GSA Today 18, 4–8. Zalasiewicz, J., Williams, M., Haywood, A., Ellis, M., 2011. The Anthropocene: a new epoch of geological time? Philos. Trans. R. Soc. A369, 835–841. Zhang, W., Montgomery, D., 1994. Digital elevation model grid size, landscape representation and hydrologic simulations. Water Resour. Res. 30, 1019–1028. Zielke, O., Arrowsmith, J.R., Ludwig, L.G., Akçiz, S.O., 2010. Slip in the 1857 and earlier large earthquakes along the Carrizo Plain, San Andreas Fault. Science 327, 1119. Zielke, O., Arrowsmith, J.R., Grant Ludwig, L., Akciz, S.O., 2012. High resolution topography derived offsets along the 1857 Fort Tejon earthquake rupture trace, San Andreas Fault. Bull. Seismol. Soc. Am. 102, 1135–1154.