Alessandro Cesare Mondini

Landslide inventory map for the Briga and the Giampilieri catchments, NE Sicily, Italy

On 1 October 2009, a high intensity storm hit the Ionian coast of Sicily, SW of Messina, Italy. The Santo Stefano di Briga rain gauge, located 2 km W of the Ionian coast, recorded 225 mm of rain in seven hours. The intense rainfall event triggered abundant slope failures, and resulted in widespread erosion and deposition of debris along ephemeral drainage channels, extensive inundation, and local modifications of the coastline. Landslides occurred in a territory prone to slope failures, due to the local geological and geomorphological settings. Many landslides were related to the presence of roads lacking adequate drainage. Abandoned terraced slopes lacking proper drainage, and unmaintained dry walls were also related to slope failures. Damage was particularly severe in small villages and at several sites along the transportation network. The shallow landslides and the inundation resulted in 37 fatalities, including 31 deaths and six missing persons, and innumerable injured people. After the event, an accurate landslide inventory map was prepared for the Briga and the Giampilieri catchments. The map shows: (i) the distribution of the event landslides triggered by the 1 October 2009 rainfall event; (ii) the distribution of the pre-existing slope failures; and (iii) other geomorphological features related to fluvial processes and slope movements. The landslide inventory map was prepared at 1:10,000 scale through a combination of field surveys and photo-interpretation of pre-event and post-event, stereoscopic and pseudo-stereoscopic, aerial photography. Different types of aerial photographs were analysed visually to prepare the landslide inventory map. The event landslides were mapped through the interpretation of pseudo-stereoscopic colour photographs taken shortly after the event at 1:3500 scale, combined with digital stereoscopic photographs at approximately 1:4500 scale, taken in November 2009. The pre-event landslides and the associated geomorphological features were mapped using 1:33,000 scale aerial photographs flown in 1954, 1955, and 2005. The event and pre-existing landslides were checked in the field in the period October–November 2009.

Very-High Resolution Stereoscopic Satellite Images for Landslide Mapping

Landslide inventory maps are essential for geomorphological studies, and to evaluate landslide hazard, vulnerability, and risk. Landslide maps, including geomorphological, event, seasonal, and multi-temporal inventory maps, are prepared using different techniques. We present the results of an experiment aimed a testing the possibility of using very high resolution, stereoscopic satellite images to map rainfall induced shallow landslides. Three landslide inventory maps were prepared for the Collazzone study area, Umbria, Italy. Two of the maps were prepared through the visual interpretation of stereoscopic satellite images and cover the periods January to March 2010, and March to May 2010. The third inventory map shows landslides occurred in the period January to May 2010, and was obtained through reconnaissance field surveys. We describe the statistics of landslide area for the three inventories, and compare quantitatively two of the landslide maps.

TRMM satellite rainfall estimates for landslide early warning in Italy: preliminary results

Early warning systems can predict rainfall-induced landslides by comparing rainfall data with landslide rainfall thresholds. These systems are based on empirical rainfall thresholds defined using rain gauges data. Despite quantitative satellite rainfall estimates are currently available, limited research has compared satellite estimates and rain gauge measurements for the forecasting of possible landslide occurrence. In this work, we validate satellite estimates obtained for Italy by the NASA Tropical Rainfall Measuring Mission (TRMM) against rainfall measurements from the Italian rain gauge network (< 1950 rain gauges), in the period from 1 September 2009 to 31 August 2010. Using cumulative rainfall measurements/estimates, we: (i) evaluate the correlation between the rain gauge measurements and the satellite estimates in different morpho-climatological domains, (ii) analyse the distributions of the ground-based measurements and the satellite estimates using different statistical approaches, and (iii) compare rainfall events derived automatically from satellite and rain gauge rainfall series. We observe differences between satellite estimates and rain gauge measurements in different morpho-climatological domains. The differences are larger in mountain areas, and collectively reveal a complex relationship between the ground-based measurements and the satellite estimates. We find that a power law correlation model is appropriate to describe the relation between the two rainfall data series. We conclude that specific rainfall thresholds must be defined to exploit satellite rainfall estimates in existing landslide early warning systems.

Measures of Spatial Autocorrelation Changes in Multitemporal SAR Images for Event Landslides Detection

Landslides cause damages and affect victims worldwide, but landslide information is lacking. Even large events may not leave records when they happen in remote areas or simply do not impact with vulnerable elements. This paper proposes a procedure to measure spatial autocorrelation changes induced by event landslides in a multi-temporal series of synthetic aperture radar (SAR) intensity Sentinel-1 images. The procedure first measures pixel-based changes between consecutive couples of SAR intensity images using the Log-Ratio index, then it follows the temporal evolution of the spatial autocorrelation inside the Log-Ratio layers using the Moran’s I index and the semivariance. When an event occurs, the Moran’s I index and the semivariance increase compared to the values measured before and after the event. The spatial autocorrelation growth is due to the local homogenization of the soil response caused by the event landslide. The emerging clusters of autocorrelated pixels generated by the event are localized by a process of optimal segmentation of the log-ratio layers. The procedure was used to intercept an event that occurred in August 2015 in Myanmar, Tozang area, when strong rainfall precipitations triggered a number of landslides. A prognostic use of the method promises to increase the availability of information about the number of events at the regional scale, and to facilitate the production of inventory maps, yielding useful results to study the phenomenon for model tuning, landslide forecast model validation, and the relationship between triggering factors and number of occurred events.

Preliminary analysis of a correlation between ground deformations and rainfall: the Ivancich landslide, central Italy

We exploited Differential Synthetic Aperture Radar Interferometry (DInSAR) to investigate the geographical and the temporal pattern of ground deformations in the Ivancich landslide area, Assisi, Italy, in the 18.4-year period April 1992 - September 2010. We used SAR data obtained by the European Remote Sensing (ERS-1/2) satellites in the period April 1992 - July 2007, and SAR data captured by the ASAR sensor on board the Envisat satellite in the period October 2003 - September 2010. We used the Small Baseline Subset (SBAS) technique to process the SAR data, obtaining full resolution measurements for multiple radar targets inside and outside the landslide area, and the history of deformation of the individual targets. The geographical pattern of the ground deformation was found consistent with independent topographic information. The deformation time series of the individual targets were compared to the rainfall history in the area. Results revealed the lack of an immediate effect of rainfall on the ground deformation, and confirmed the existence of a complex temporal interaction between the rainfall and the ground deformation histories in the landslide area. Availability of very long, spatially distributed time series of surface deformation has provided an unprecedented opportunity to investigate the history of the active landslide area.

Land Use Change Scenarios and Landslide Susceptibility Zonation: The Briga Catchment Test Area (Messina, Italy)

Landslides spatial distribution and frequency are the consequence of different meteorological and environmental conditions including morphological, hydrological, lithology the land use settings. In this work we have attempted to evaluate the influence of land use change on landslide spatial distribution occurrence (susceptibility) for a portion of the Briga catchment, located along the Ionian coast of Sicily, Italy. On 1 October 2009, the area was hit by a high intensity rainfall event that triggered abundant slope failures, and resulted in widespread erosion and deposition of debris along ephemeral drainage channels. After the storm, an accurate event landslide inventory map was made for the catchment and a pre-event landslide map was prepared using aerial photographs. Moreover two different land use maps were realized, the first was obtained through a semi-automatic classification of digitized aerial photographs acquired in 1954, the second through the combination of supervised classifications of two QuickBird images acquired in 2006 and 2009. Using the available thematic data, we have prepared susceptibility zonation through multivariate statistical models exploiting the 2009 event landslides as grouping variable and simple morphological and 2009 land use data as explanatory variables. To evaluate the influence of land use change on the susceptibility zonation, the same discriminant models were applied to different land use distribution including the 1954 land use map. Differences in the landslide susceptibility maps were analysed to understand how land use change affects the landslide occurrence.

Probabilistic Prediction of Landslides Induced by Rainfall

Landslides are widespread and cause every year casualties and extensive damages. Predicting their spatial and temporal occurrence of landslides is a problem of scientific and societal interest. Empirical threshold model approaches proposed in the literature have limitations related to the heuristic identification of rainfall conditions triggering landslide, to the subjective choice of threshold model, to the biased probability estimation related to the classical empirical threshold model, and to limited use of rainfall events not associated to landslides. A new probabilistic empirical prediction schema is proposed to overcome these limitations. The model was applied successfully in Umbria region considering rain gauge measures and satellite rainfall estimate.

Modelling Landslides’ Susceptibility by Fuzzy Emerging Patterns

This contribution proposes an approach to model regional landslide susceptibility, based on a supervised learning technique that mines fuzzy emerging patterns on a set of classified data. In our approach the training set contains positive and negative examples of areas, (i.e., slope units), affected or not affected by landslides. The fuzzy emerging patterns characterise the positive and the negative areas exploiting their ability to discriminate between the two classes. The approach consists first, in inducing a set of fuzzy rules, and then in reducing them by retaining those that identify fuzzy emerging patterns for the given training set. The fuzzy rules define the main characteristics of the slope units that are affected or not affected by landslides and are used to classify other slope units in the same region. The classification technique provides an estimate of the hesitation of the decision process, which is a measure of its ability to uniquely associate a slope unit to the susceptible or not susceptible class. In the paper we describe the approach and discuss the preliminary results.

Slope Dynamics and Climatic Change Through Indirect Interactions

The rapid variation of climate can cause direct changes in slope dynamics due to a modified rainfall regime. Variations in evapotranspiration regime determines changes in soil moisture, modifies shrinking-swelling cycles, creeping, surface mass movement, and soil erosion, including gully erosion. All these effects can be considered as direct consequences of any climate modification. Besides them, other indirect effects should be considered to fully determine climate change impact on slope dynamics. This is the case of the effects of climate change on vegetation, that strongly controls slope instability phenomena. Here we will concentrate on the effect of increased danger due to forest fire, and in particular we discuss the changes in the hydrogeological hazard linked to the effect of drought on wild fires in a case-study in Umbria (Italy), mainly considering field observations and simulations with LANDPLANER (LANDscape, Plant, LANdslide and ERosion) model. This study shows that when discussing of climate changes particular emphasis must be put on side effects that can influence slope dynamics and basin behavior. In particular the understanding of where threats can come, requires the identification of complex framework describing the dynamic interaction of all the elements coexisting in a slope.

TXT-tool 1.039-1.1: Very-High Resolution Stereo Satellite Images for Landslide Mapping

Landslides are common phenomena in mountainous countries, and play an important role in the evolution of landscapes. They also represent a serious hazard in many areas of the world. Acquiring systematic information on the type, abundance, and distribution of landslides, and preparing landslide inventory maps is of fundamental importance to mitigate landslide risk. Landslide inventory maps are essential for evaluating landslide hazard, vulnerability and risk, and for studying the evolution of landscapes dominated by mass-wasting processes. Landslide maps, including geomorphological, event, seasonal, and multi-temporal inventory maps, can be prepared using different techniques. We present the results of an experiment aiming at testing the possibility of using very high resolution, stereoscopic satellite images to map rainfall-induced shallow landslides. Three landslide inventory maps were prepared for the Collazzone study area, Umbria, Italy. Two of the maps were prepared through the visual interpretation of stereoscopic satellite images, and cover the periods January–March 2010, and March–May 2010. The third inventory map shows landslides occurred in the period January–March 2010, and was obtained through reconnaissance field surveys. We describe the statistics of landslide area for the three inventories, and compare quantitatively two of the landslide maps.