Influence of uncertain identification of triggering rainfall on the assessment of landslide early warning thresholds

Abstract

Uncertainty in rainfall datasets and landslide inventories is known to have negative impacts on the assessment of landslidetriggering thresholds. In this paper, we perform a quantitative analysis of the impacts that the uncertain knowledge of landslide initiation instants have on the assessment of landslide intensity-duration early warning thresholds. The analysis is based on an ideal synthetic database of rainfall and landslide data, generated by coupling a stochastic rainfall generator and a physically 15 based hydrological and slope stability model. This dataset is then perturbed according to hypothetical "reporting scenarios", that allow to simulate possible errors in landslide triggering instants, as derived from historical archives. The impact of these errors is analysed by combining different criteria to single-out rainfall events from a continuous series and different temporal aggregations of rainfall (hourly and daily). The analysis shows that the impacts of the above uncertainty sources can be significant. Errors influence thresholds in a way that they are generally underestimated. Potentially, the amount of the 20 underestimation can be enough to induce an excessive number of false positives, hence limiting possible landslide mitigation benefits. Moreover, the uncertain knowledge of triggering rainfall, limits the possibility to set up links between thresholds and physio-geographical factors.

Landslide precipitation thresholds in Rwanda

“…Those thresholds are mainly inferred from empirical methods which are based on statistical analysis of historical rainfall characteristics and landslides inventories to distinguish the landslide conditions from no-landslide conditions. However, many limitations, constraints and uncertainties associated with empirical thresholds have been highlighted (Peres et al 2017;Prenner et al 2018;Bogaard and Greco 2018). Some limitations are due to the fact that empirical thresholds are mainly based on the rainfall event during which a landslide occurred, which is in reality the actual landslide trigger.…”

Examples of Application of GASAKe for Predicting the Occurrence of Rainfall-Induced Landslides in Southern Italy

“…On the other hand, model performances are known to be hampered by quality or completeness of input data (e.g., missing dates of activation; dates related to other triggering factors; unsuitability of the rain gauge network) [65]. Furthermore, available rain gauges are usually far from the sites of interest, and located at either different elevations or aspects.…”

Probabilistic landslide ensemble prediction systems: Lessons to be learned from hydrology

“…All used variables, deterministic model outputs (the FoS maps) and the probabilistic model output (the PoF map) are parsed through to R (R Core Team, 2017). In R, all piped arguments from the python script are used for producing ready-to-use maps (packages: rgdal , sp (Pebesma and Bivand, 2005)) or to visualize performance measures such as ROC plots (package: ROCR (Sing et al, 2005)). The entire procedure from 15 importing raw data to producing usable maps is fully automated within an executable file that may be initiated every hour.…”

“…Guzzetti et al (2007) give an overview of rainfall and climate variables used in the literature for the definition of rainfall thresholds for the initiation of landslides, however, such empirical-statistical approaches only pose a simplification between 15 rainfall occurrence and the physical mechanisms leading to landslides, neglecting local environmental conditions and the role of the hydrological processes occurring along slopes (Reichenbach et al, 1998, Bogaard and. Attempts to relate landslide-triggering thresholds to weather and other physically based characteristics can be very challenging given the quality of currently available data (Peres et al, 2017). Another reason for the negligence of physically based forecasting initiatives used to be the lacking spatial resolution and computational power for considering such convective-scale phenomena which are 20 of particular interest for modelling small scale related phenomena with a rapid onset such as shallow landslides and flash floods.…”

“…Empirical approaches, such as the commonly used rainfall thresholds in landslide early warning applications, started to incorporate estimates of uncertainty only recently by defining rainfall thresholds at different 25 exceedance probabilities (e.g. Melillo et al, 2016, Piciullo et al, 2017), yet they rely on very good landslide event catalogues and thus purely on past reportings, which adds a tremendously large source of error (Peres et al, 2017). Gariano et al (2015) found that an underestimation of only 1% in the number of considered landslides can result in a significant decrease in the performance of a threshold based landslide EWS.…”