Landslide inventories and their statistical properties

Landslides are generally associated with a trigger, such as an earthquake, a rapid snowmelt or a large storm. The landslide event can include a single landslide or many thousands. The frequency–area (or volume) distribution of a landslide event quanti?es the number of landslides that occur at different sizes. We examine three well‐documented landslide events, from Italy, Guatemala and the USA, each with a different triggering mechanism, and ?nd that the landslide areas for all three are well approximated by the same three‐parameter inverse‐gamma distribution. For small landslide areas this distribution has an exponential ‘roll‐over’ and for medium and large landslide areas decays as a power‐law with exponent ‐2·40. One implication of this landslide distribution is that the mean area of landslides in the distribution is independent of the size of the event. We also introduce a landslide‐event magnitude scale m_L = log(N_LT), with N_LT the total number of landslides associated with a trigger. If a landslide‐event inventory is incomplete (i.e. smaller landslides are not included), the partial inventory can be compared with our landslide probability distribution, and the corresponding landslide‐event magnitude inferred. This technique can be applied to inventories of historical landslides, inferring the total number of landslides that occurred over geologic time, and how many of these have been erased by erosion, vegetation, and human activity. We have also considered three rockfall‐dominated inventories, and ?nd that the frequency–size distributions differ substantially from those associated with other landslide types. We suggest that our proposed frequency–size distribution for landslides (excluding rockfalls) will be useful in quantifying the severity of landslide events and the contribution of landslides to erosion. Copyright © 2004 John Wiley & Sons, Ltd.     

Landslides triggered by the 23 November 2000 rainfall event in the Imperia Province, Western Liguria, Italy

From mid-October to 22 November 2000, the western Liguria Region of Italy experienced prolonged and intense rainfall, with cumulative values exceeding 1000 mm in 45 days. The severe rainfall sequence ended on November 23 with a high-intensity storm that dumped more than 180 mm of rain in 24 h. The high-intensity event caused flooding and triggered more than 1000 soils slips and debris flows and a few large, complex landslides. Slope failures caused three fatalities and severe damage to roads, private homes, and agriculture. Large (1:13,000) and very large (1:5000) scale colour aerial photographs were taken 45 days after the event over the areas most affected by the landslides. Through the interpretation of the 334 photographs covering an area of 500 km², we prepared a landslide inventory map that shows 1204 landslides, for a total landslide area of 1.6 km². We identified the rainfall conditions that triggered landslides in the Armea valley using cumulative- and continuous-rainfall data, combined with detailed information on the time of landslide occurrence. Landslide activity initiated 8 to 10 h after the beginning of the storm, and the most abundant activity occurred in response to rainfall intensities of 8 to 10 mm per hour. For the Ceriana Municipality, an area where the landslides were numerous in November 2000, we also collected information about a historical event that occurred on 8–11 December 1910 and triggered abundant landslides resulting in severe economic damage. A comparison of the damage caused by the historical and the recent landslide events indicated that damage caused by the 1910 historical event was more diffused but less costly than the damage caused by the 2000 event.     

The role of mineralogy,geochemistry and grain size in badland development in Pisticci (Basilicata,southern Italy)

Mineralogical, geochemical and grain‐size composition of soil and pore‐water chemistry parameters were characterized on both eroded (south‐facing) and non‐eroded (north‐facing) clayey‐silt slopes in the Basilicata region (southern Italy). Only a few grain‐size parameters and clay mineralogy discriminate eroded from non‐eroded substrates. Compared with the latter, the former have fractions of over 63 µm and 1–4 µm lower and fractions 4–63 µm higher. Grain‐size characters of crusts did not discriminate with respect to substrate. Bulk rock mineralogy was not distinctive, but the clay mineral assemblage shows that the eroded slope is enriched in kaolinite, mixed layers (illite–smectite) and chlorite, whereas illite decreases, although overlaps are common. Chemical data enable discrimination between eroded and non‐eroded slopes. pH, SAR (sodium adsorption ratio), TDS (total dissolved salts) and PS (percentage of sodium) are distinctive parameters for both eroded and non‐eroded slopes. TDS increases in depth in the non‐eroded slope, whereas the maximum TDS is just below the crust in the eroded one. On average, eroded substrates are higher in pH, SAR and PS than non‐eroded ones. The ESP (exchangeable sodium percentage) of the eroded slope has a higher value than the non‐eroded one. Crusts are less dispersive than eroded substrates, and non‐eroded substrates behave as crusts. This suggests that the portion of the slope most severely exposed to weathering tends to stabilize, due to strong decreases in SAR, PS and ESP. Several diagrams reported in the literature show similarly anomalous crust samples on eroded slopes, compared with other samples coming from greater depths on eroded slopes. In the present case study, the exchangeable form of Na characterizes crusts more than the soluble form. This study describes the erosional mechanism, which involves morphological and geographic exposure and climatic elements, as well as grain size, mineralogy, chemistry and exchangeable processes of soils. Copyright © 2006 John Wiley & Sons, Ltd.     

Oscillatory Pumping Test to Estimate Aquifer Hydraulic Parameters in a Bayesian Geostatistical Framework

Comprehensive information about the spatial distribution of the subsurface hydraulic properties is crucial to model groundwater flow, to predict solute transport in aquifers and to design remediation actions. In this work, a Bayesian Geostatistical approach, as implemented in bgaPEST, was adopted to estimate the hydraulic properties of a well field located at the Campus of Science and Technology of the University of Parma (Northern Italy), in a contest of a highly parameterized inversion. Head data, collected by means of multi frequency oscillatory pumping tests, were used to both estimate the hydraulic parameters and validate the results. The groundwater flow processes were modelled by means of MODFLOW 2005 and an adjoint-state formulation of the same software was used to efficiently calculate the sensitivity matrix, required by the inverse procedure. The Bayesian Geostatistical approach estimated the hydraulic conductivity and specific storage fields, handling a large number of parameters. The results of the inversion are consistent with the alluvial nature of the investigated aquifer and the preliminary traditional pumping tests carried out at the site.     

Analysis of Ground Deformation Detected Using the SBAS-DInSAR Technique in Umbria, Central Italy

Ground deformation affecting the Umbria region (central Italy) in the 9-year period from 1992 to 2000 was investigated through multi-temporal Differential Synthetic Aperture Radar Interferometry (DInSAR). For the purpose, the Small BAseline Subset (SBAS) technique was adopted, which allows studying the temporal evolution of the detected deformation at two spatial scales: a low-resolution (regional) scale, and a full-resolution (local) scale. For the analysis, SAR data acquired by the European Remote Sensing (ERS-1/2) satellites along ascending and descending orbits were used. The detected deformation was analysed to investigate its relevance to geophysical, geomorphologic, and human-induced processes that may result in hazardous conditions to the population of Umbria. Low-resolution deformation data were used to: (i) determine the amount of displacement caused by the Umbria-Marche earthquake sequence from September 1997 to April 1998 in the Foligno area, (ii) determine the number and percentage of the known landslides that can be monitored by the DInSAR technology in the investigated area, and (iii) identify and measure subsidence induced by exploitation of a confined aquifer in the Valle Umbra. Results indicate that earthquakes moved through the Foligno area westwards up to 3.9 cm and with an uplift reaching 1.7 cm. Intersection in a GIS of the low-resolution deformation maps with a detailed landslide inventory map allowed the determination that the portion of landslides that can be monitored by the SBAS-DInSAR technique in Umbria ranges from 2.7% to 3.4%, and the percentage of the total landslide area ranges from 10.4% to 12.8%. In the Valle Umbra, a dependency was found between the time and the amount of detected ground deformation, and the record of water withdrawal. The full-resolution deformation data were used to investigate the movement of the Ivancich landslide, in the Assisi Municipality. Joint analysis of the spatial and the temporal characteristics of the ground displacement allowed the formulation of a hypothesis on the landslide geometry and deformation pattern.     

The stability of the pristine magnetopause

A MHD theory of combined Kelvin-Helmholtz (KH) and Rayleigh-Taylor (RT) instabilities for a transition layer with two different scale lengths (Δ and δ for the variation of velocity/magnetic fields and density, respectively) is presented. The study is motivated by reports of magnetopauses with no low latitude boundary layer, in which a sharp density drop over a distance δ?Δ is observed (“pristine” magnetopauses (J. Geophys. Res. 101 (1996) 49). The theory ignores compressibility effects and applies to subsonic regions of the dayside magnetopause. The RT effect is included to account for temporary periods of acceleration of the magnetopause, caused by sudden changes of the solar wind dynamic pressure. For small wavelengths λ, such that δ?λ?Δ, a WKB solution shows that the velocity gradient operates, together with magnetic tensions, to attenuate or even stabilize the Rayleigh-Taylor instability within a certain wavelength range. An exact dispersion relation for flute modes, valid for all λ, in the form of a fourth order polynomial for the complex frequency ω, is obtained from a model with a constant velocity gradient, dV/dy within Δ, and with δ→0. Flute modes are possible because of the existence of bands of very small magnetic shear on the dayside magnetopause (J. Geophys. Res. 103 (1998) 6703). The exact solution allows for a study of the change of the action of the velocity gradient with λ from the long-λ range where dV/dy is KH destabilizing to the short-λ range where dV/dy produces a stabilizing effect. Both, the WKB approximation and the well known tangential discontinuity model (Δ→0) are recovered as limiting cases of the exact solution. Properties of the KH and RT instabilities, for different density ratios on either side of the magnetopause, are described. For flute modes, at very small λ the RT instability grows faster and becomes the dominant effect. However, it is shown that the growth rate remains bounded at a finite value as λ→0, when a theory with a finite δ model is considered. To study configurations with finite, arbitrary, δ/Δ ratios, the MHD perturbation equations are solved numerically, using hyperbolic tangent functions for both the density and velocity transitions across the magnetopause. To examine the influence of different δ/Δ ratios on the growth rates of KH and RT, calculations are performed for different δ/Δ, with and without acceleration, and for two different density ratios. It is found that the general features exhibited by the constant dV/dy model, are confirmed by these numerical solutions. The stability of pristine magnetopauses, and the possibility of observing some theoretical predictions during magnetopause crossings in ongoing missions, are discussed.     

Assessing the influence of morpho-structural setting on landslide abundance

Knowing the factors that influence landslide abundance and distribution is important to evaluate landslide susceptibility and hazard. Visual interpretation of aerial photographs (API) can be used to collect spatially distributed information on bedding attitude (BA), in an area. Where a map of the location of bedding traces (BTs), i.e. lines showing the intersection of bedding planes with the local topography, is available, the map can be used to obtain BA point data and to prepare maps showing morpho-structural domains. The possibility of using BA maps to investigate the influence of morpho-structural settings on landslide abundance is hampered by the lack of understanding of the influence of the length of the BTs, and of the parameters used to interpolate the BA data on the structural zonation. To investigate the problem, we used information on 207 BTs obtained through API in the Collazzone area, Central Italy, and we prepared 150 maps showing BA information. This was accomplished using 15 different values for the segmentation length of the BTs (S), and 10 different values for the tension parameter (T) used for the interpolation. We compare the results against previous results obtained for the same area adopting a heuristic approach to the segmentation of the same set of BTs. Next, we compare the geographical distribution of old deep-seated, deep-seated and shallow landslides in five morpho-structural domains in the study area, and we analyse the influence of the structural settings on the abundance of the different types of landslides.     

Review of Modeling Approaches to Groundwater Flow in Deformed Carbonate Aquifers

We discuss techniques to represent groundwater flow in carbonate aquifers using the three existing modeling approaches: equivalent porous medium, conduit network, and discrete fracture network. Fractures in faulted stratigraphic successions are characterized by dominant sets of sub-vertical joints. Grid rotation is recommended using the equivalent porous medium to match higher hydraulic conductivity with the dominant orientation of the joints. Modeling carbonate faults with throws greater than approximately 100 m is more challenging. Such faults are characterized by combined conduit-barrier behavior. The barrier behavior can be modeled using the Horizontal Flow Barrier Package with a low-permeability vertical barrier inserted to represent the impediment of horizontal flow in faults characterized by sharp drops of the piezometric surface. Cavities can occur parallel to the strike of normal faults generating channels for the groundwater. In this case, flow models need to account for turbulence using a conduit network approach. Channels need to be embedded in an equivalent porous medium due to cavities a few centimeters large, which are present in carbonate aquifers even in areas characterized by low hydraulic gradients. Discrete fracture network modeling enables representation of individual rock discontinuities in three dimensions. This approach is used in non-heavily karstified aquifers at industrial sites and was recently combined with the equivalent porous medium to simulate diffusivity in the matrix. Following this review, we recommend that the future research combines three practiced modeling approaches: equivalent porous medium, discrete fracture network, and conduit network, in order to capture structural and flow aspects in the modeling of groundwater in carbonate rocks.     

Moisture source in the Hyblean Mountains region (south-eastern Sicily, Italy): Evidence from stable isotopes signature

Here the authors present results of an isotope study on precipitation collected during a 2-a period from a rain-gauge network consisting of 6 stations located at different elevations in the Hyblean Mountains (HM) region, in south-eastern Sicily. The slope of the local meteoric water line (δD = 6.50 δ¹⁸O + 9.87) obtained for the region suggests that precipitation is affected by evaporation during rainfall events. The main variations in rainwater isotope composition are due to seasonal effects and elevation. An average ²H excess value of +21.2‰ was found for precipitation events less affected by evaporation (i.e. when the rainfall was >65 mm/month). The spatial distribution of O isotope composition of precipitation shows a negative gradient from east and south to the inner areas. The depositional rate of Cl, used as a tracer of the origin of air masses, is highest at the coastal rain-gauges (SR and MRG stations) and lowest on the northern flank of the HM region (SC station). Based on these findings, a model is proposed for the origin of precipitation in the HM region, which assumes that a Mediterranean-derived component is the main source of moisture in the studied area. D/H and ¹⁸O/¹⁶O ratios of inferred meteoric recharge waters were also compared with the isotope composition of waters collected from the main local springs and wells. The best linear fit of the δ¹⁸O vs δD relationship for Hyblean groundwater is δD = 4.85 δ¹⁸O–2.01. The enrichment of heavy isotopes in Hyblean groundwater is probably due to evaporation occurring after precipitation events or to a recharging contribution from surface waters (lakes or rivers) enriched in heavy isotopes.