Correlation does not imply geomorphic causation in data-driven landslide susceptibility modelling - Benefits of exploring landslide data collection effects.

Data-driven landslide susceptibility models formally integrate spatial landslide information with explanatory environmental variables that describe predisposing factors of slope instability. Well-performing models are commonly utilized to identify landslide-prone terrain or to understand the causes of slope instability. In most cases, however, the available landslide data is affected by spatial biases (e.g. underrepresentation of landslides far from infrastructure or in forests) and does therefore not perfectly represent the spatial distribution of past slope instabilities. Literature shows that implications of such data flaws are frequently ignored. This study was built upon landslide information that systematically relates to damage-causing and infrastructure-threatening events in South Tyrol, Italy (7400 km2). The created models represent three conceptually different strategies to deal with biased landslide information. The aims were to demonstrate why an inference of geomorphic causation from apparently well-performing models is invalid under common landslide data bias conditions (Model 1), to test a novel bias-adjustment approach (Model 2) and to exploit the underlying data bias to model areas likely affected by potentially damaging landslides (Model 3; intervention index), instead of landslide susceptibility. The study offers a novel perspective on how biases in landslide data can be considered within data-driven models by focusing not only on the process under investigation (landsliding), but also on the circumstances that led to the registration of landslide information (data collection effects). The results were evaluated in terms of statistical relationships, variable importance, predictive performance, and geomorphic plausibility. The results revealed that none of the models reflected landslide susceptibility. Despite partly high predictive performances, the models were unable to create geomorphically plausible spatial predictions. The impact-oriented intervention index, however, enabled to identify damage-causing landslides with high accuracy. We conclude that the frequent practice of inferring geomorphic causation from well-performing models without accounting for data limitations is invalid.

Author Correction: Changes in climate patterns and their association to natural hazard distribution in South Tyrol (Eastern Italian Alps).

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Changes in climate patterns and their association to natural hazard distribution in South Tyrol (Eastern Italian Alps).

In Alpine regions changes in seasonal climatic parameters, such as temperature, rainfall, and snow amount have already been observed. Specifically, in the South Tyrol area, meteorological observations indicate that temperatures are increasing and the number of snow days has generally diminished over time with perennial snow line now observed at higher elevations. Changes in rainfall have also been observed with more events associated with higher temperatures in the summer season. Natural hazards - mainly debris and mud flows, landslides, avalanches, rock falls, and (flash) floods - that affect this area every year, damaging population and infrastructures, are either weather or cryosphere-related. While these events have been recorded sporadically since the beginning of the 20th century, a systematic approach of their inventory has been done by local authorities since the 1990s. So far, Earth observation data has not been exploited to complete or complement existing inventories nor have they been used to investigate the influence of climate perturbation on potentially dangerous natural phenomena. The research presented here thus has three objectives: (i) analyse long time series of climate data and hazard occurrence in South Tyrol to examine if these records exhibit a coherent response of hazards to changes in climate; (ii) measure the spatio-temporal evolution of climatic and natural hazard events recorded, and (iii) explore potential relations between meteorological conditions and the hazard occurrence. In this context, in-situ and satellite-based climate data are exploited to study natural hazard triggers while the potential of Earth observation data is evaluated as a complement to the existing historical records of natural hazards. Specifically, Copernicus Sentinel-1 images are used to detect the spatio-temporal distribution of slow earth surface deformations and the results used for checking the completeness of the actual slow-moving landslide inventories. Hazard-related changes in the South Tyrolian landscape have also been analysed in relation to particular meteorological events at a regional scale, assessing trends and anomalies. Results show that: (i) satellite data are very useful to complement the existing natural hazard inventories; (ii) in-situ and satellite-based climate records show similar patterns but differ due to regional versus local variability; (iii) even in a data-rich region such as the analysed area, the overall response of natural hazard occurrence, magnitude, and frequency to change in climate variables is difficult to decipher due to the presence of multiple triggers and locally driven ground responses. However, an increase in the average annual duration of rainfall events and debris flow occurrence can be observed.