A landslide risk index for municipal land use planning in Portugal

Learning Ground Displacement Signals Directly from InSAR-Wrapped Interferograms

Monitoring ground displacements identifies potential geohazard risks early before they cause critical damage. Interferometric synthetic aperture radar (InSAR) is one of the techniques that can monitor these displacements with sub-millimeter accuracy. However, using the InSAR technique is challenging due to the need for high expertise, large data volumes, and other complexities. Accordingly, the development of an automated system to indicate ground displacements directly from the wrapped interferograms and coherence maps could be highly advantageous. Here, we compare different machine learning algorithms to evaluate the feasibility of achieving this objective. The inputs for the implemented machine learning models were pixels selected from the filtered-wrapped interferograms of Sentinel-1, using a coherence threshold. The outputs were the same pixels labeled as fast positive, positive, fast negative, negative, and undefined movements. These labels were assigned based on the velocity values of the measurement points located within the pixels. We used the Parallel Small Baseline Subset service of the European Space Agency's GeoHazards Exploitation Platform to create the necessary interferograms, coherence, and deformation velocity maps. Subsequently, we applied a high-pass filter to the wrapped interferograms to separate the displacement signal from the atmospheric errors. We successfully identified the patterns associated with slow and fast movements by discerning the unique distributions within the matrices representing each movement class. The experiments included three case studies (from Italy, Portugal, and the United States), noted for their high sensitivity to landslides. We found that the Cosine K-nearest neighbor model achieved the best test accuracy. It is important to note that the test sets were not merely hidden parts of the training set within the same region but also included adjacent areas. We further improved the performance with pseudo-labeling, an approach aimed at evaluating the generalizability and robustness of the trained model beyond its immediate training environment. The lowest test accuracy achieved by the implemented algorithm was 80.1%. Furthermore, we used ArcGIS Pro 3.3 to compare the ground truth with the predictions to visualize the results better. The comparison aimed to explore indications of displacements affecting the main roads in the studied area.

Evaluation of potential changes in landslide susceptibility and landslide occurrence frequency in China under climate change

Climate change can alter the frequency and intensity of extreme rainfall across the globe, leading to changes in hazards posed by rainfall-induced landslides. In recent decades, China suffered great human and economic losses due to rainfall-induced landslides. However, how the landslide hazard situation will evolve in the future is still unclear, also because of sparse comprehensive evaluations of potential changes in landslide susceptibility and landslide occurrence frequency under climate change. This study builds upon observed and modelled rainfall data from 24 bias-corrected Coupled Model Intercomparison Project Phase 6 (CMIP6) Global Climate Models (GCMs), a statistical landslide susceptibility model, and empirical rainfall thresholds for landslide initiation, to evaluate changes in landslide susceptibility and landslide occurrence frequency at national-scale. Based on four Shared Socioeconomic Pathways (SSP) scenarios, changes in the rainfall regime are projected and used to evaluate subsequent alterations in landslide susceptibility and in the frequency of rainfall events exceeding empirical rainfall thresholds. In general, the results indicate that the extend of landslide susceptible terrain and the frequency of landslide-triggering rainfall will increase under climate change. Nevertheless, a closer inspection provides a spatially heterogeneous picture on how these landslide occurrence indicators may evolve across China. Until the late 21st century (2080-2099) and depending on the SSP scenarios, the mean annual precipitation is projected to increase by 13.4 % to 28.6 %, inducing an 1.3 % to 2.7 % increase in the modelled areal extent of moderately to very highly susceptible terrain. Different SSP scenarios were associated with an increase in the frequency of landslide-triggering rainfall events by 10.3 % to 19.8 % with respect to historical baseline. Spatially, the southeastern Tibetan Plateau and the Tianshan Mountains in Northwestern Basins are projected to experience the largest increase in landslide susceptibility and frequency of landslide-triggering rainfall, especially under the high emission scenarios. Adaptation and mitigation methods should be prioritized for these future landslide hotspots. This work provides a better understanding of potential impacts of climate change on landslide hazard across China and represents a first step towards national-scale quantitative landslide exposure and risk assessment under climate change.

A methodological approach of QRA for slow-moving landslides at a regional scale

Landslides represent a serious worldwide hazard, especially in Italy, where exposure to hydrogeological risk is very high; for this reason, a landslide quantitative risk assessment (QRA) is crucial for risk management and for planning mitigation measures. In this study, we present and describe a novel methodological approach of QRA for slow-moving landslides, aiming at national replicability. This procedure has been applied at the basin scale in the Arno River basin (9100 km², Central Italy), where most landslides are slow-moving. QRA is based on the application of the equation risk = hazard (H) × vulnerability (V) × exposure (E) and on the use of open data with uniform characteristics at the national scale. The study area was divided into a grid with a 1 km² cell size, and for each cell, the parameters necessary for the risk assessment were calculated. The obtained results show that the total risk of the study area amounts to approximately 7 billion €. The proposed methodology presents several novelties in the risk assessment for the regional/national scale of the analysis, mainly concerning the identification of the datasets and the development of new methodologies that could be applicable over such large areas. The present work demonstrates the feasibility of the methodology and discusses the obtained results.

Environment diagnosis for land-use planning based on a tectonic and multidimensional methodology

In this research, emphasis is placed on the information and diagnostic phase of the physical environment for land-use planning (LUP). Our work is mainly focused on a land-planning case study of a tectonic depression, the Tulum Valley, which extends into the Pampean flat-slab segment. We propose the use of tectonic structures to define Environment Units (EUs) as necessary boundaries for the LUP. For this purpose, we have studied tectonic structures using geophysical methods and, subsequently, subjected multiple dimensions of the physical environment in the territory to an exhaustive analysis. Moreover, we have examined the influence of structural geology on water, soils, processes, materials and forms in the landscape. The study revealed the close and significant relationship between the different elements of the physical environment observable on the surface (shape, distribution, appearance, degree of development) and the tectonic structures, which supported the use of this criterion to define EUs. In order to test it, we applied the same methodology in another area of South America, the city of Bucaramanga, where it was possible to define EUs based on tectonics and to also establish comparisons. The methodology proposed for the diagnostic phase based mainly on the tectonic factor represents a challenge as regards its application in other active tectonic zones. Some limitations could arise such as fragmented environmental information from different institutions or the small to non-existent number of tectonic studies available. As a strong point, we find that the method allows achieving a comprehensive study of the environmental setting and thus to propose activities and land uses in each EU according to the real reception capacity of the land. This exhaustive analysis of the physical environment will also help decision-makers to understand and manage the socio-natural risks of the territory where communities develop.

Comparison of Tree-Structured Parzen Estimator Optimization in Three Typical Neural Network Models for Landslide Susceptibility Assessment

Landslides pose a constant threat to the lives and property of mountain people and may also cause geomorphological destruction such as soil and water loss, vegetation destruction, and land cover change. Landslide susceptibility assessment (LSA) is a key component of landslide risk evaluation. There are many related studies, but few analyses and comparisons of models for optimization. This paper aims to introduce the Tree-structured Parzen Estimator (TPE) algorithm for hyperparameter optimization of three typical neural network models for LSA in Shuicheng County, China, as an example, and to compare the differences of predictive ability among the models in order to achieve higher application performance. First, 17 influencing factors of landslide multiple data sources were selected for spatial prediction, hybrid ensemble oversampling and undersampling techniques were used to address the imbalanced sample and small sample size problem, and the samples were randomly divided into a training set and validation set. Second, deep neural network (DNN), recurrent neural network (RNN), and convolutional neural network (CNN) models were adopted to predict the regional landslides susceptibility, and the TPE algorithm was used to optimize the hyperparameters respectively to improve the assessment capacity. Finally, to compare the differences and optimization effects of these models, several objective measures were applied for validation. The results show that the high-susceptibility regions mostly distributed in bands along fault zones, where the lithology is mostly claystone, sandstone, and basalt. The DNN, RNN, and CNN models all perform well in LSA, especially the RNN model. The TPE optimization significantly improves the accuracy of the DNN and CNN (3.92% and 1.52%, respectively), but does not improve the performance of the RNN. In summary, our proposed RNN model and TPE-optimized DNN and CNN model have robust predictive capability for landslide susceptibility in the study area and can also be applied to other areas containing similar geological conditions.

Definition of Environmental Indicators for a Fast Estimation of Landslide Risk at National Scale

The purpose of this paper is to propose a new set of environmental indicators for the fast estimation of landslide risk over very wide areas. Using Italy (301,340 km2) as a test case, landslide susceptibility maps and soil sealing/land consumption maps were combined to derive a spatially distributed indicator (LRI—landslide risk index), then an aggregation was performed using Italian municipalities as basic spatial units. Two indicators were defined, namely ALR (averaged landslide risk) and TLR (total landslide risk). All data were processed using GIS programs. Conceptually, landslide susceptibility maps account for landslide hazard while soil sealing maps account for the spatial distribution of anthropic elements exposed to risk (including buildings, infrastructure, and services). The indexes quantify how much the two issues overlap, producing a relevant risk and can be used to evaluate how each municipality has been prudent in planning sustainable urban growth to cope with landslide risk. The proposed indexes are indicators that are simple to understand, can be adapted to various contexts and at various scales, and could be periodically updated, with very low effort, making use of the products of ongoing governmental monitoring programs of Italian environment. Of course, the indicators represent an oversimplification of the complexity of landslide risk, but this is the first time that a landslide risk indicator has been defined in Italy at the national scale, starting from landslide susceptibility maps (although Italy is one of the European countries most affected by hydro-geological hazards) and, more in general, the first time that land consumption maps are integrated into a landslide risk assessment.

Reassessing wildfire susceptibility and hazard for mainland Portugal

Recent large wildfires have caused serious environmental and social impacts in different areas of the world, such as Portugal, Greece and Australia. Sociodemographic changes in rural areas and climate change issues create unprecedented and challenging circumstances that call for an adjustment of fire management strategies and planning tools. In this research, we revise the procedure to assess wildfire susceptibility and hazard for mainland Portugal in a structural perspective. Based on a dataset of burned areas covering 44 years and a set of predisposing factors related to topography and landcover, we tested several susceptibility models using the Likelihood Ratio method, to calculate favourability scores. Landcover classes were analysed using national detailed sources and considering different timeframes. Hazard levels were obtained by combining the susceptibility model with the best accuracy and the probability to burn estimated from fire history (1975-2018). The resulting hazard values were classified in 5 levels based on the breaks of the success curve, instead of quantiles as in previous maps, to ensure territorial continuity at different spatial scales. The results show where the combination of terrain features is more favorable to fire propagation. The higher favourability scores were found for shrubland-type vegetation, whereas agricultural areas, cork and holm forests show lower scores. Eucalyptus and maritime pine forests show similar intermediate scores and have increased since 2007. The two highest hazard classes classify correctly 90% of the burned area over 44 years, demonstrating the high accuracy of the model. If integrated in spatial planning instruments, in conjunction with municipal plans, the two most hazardous classes may hinder the expansion of built-up areas. About 2% of the municipalities have more than 90% of their territory classified as hazardous, whereas 32% of the municipalities have, instead, less than 10% of hazardous area. These structural maps are a useful baseline for a long-term approach and can be complemented by estimations regarding the behavior and severity of wildfires, which should be further explored.