HYDRO-METEOROLOGICAL THRESHOLDS FOR SHALLOW LANDSLIDE INITIATION IN PUERTO RICO

Landslides are a prominent geologic hazard in Puerto Rico (PR). The tropical setting of PR makes it susceptible to high intensity rainfall events like hurricanes; being exposed to these atmospheric phenomena promotes slope instability, especially in sites disrupted by human activities. In an effort to mitigate this hazard, a quantitative preliminary analysis was executed to optimize new potential shallow landslide forecasting and “nowcasting” systems by evaluating hydro-meteorological thresholds of five stations across the island’s mountainous topography.

Traditional studies have proposed simple rainfall intensity duration thresholds for global and PR-specific landslide initiation; this approach is limited because rainfall events are considered independently of antecedent soil moisture conditions, which significantly influence landslide occurrence. HydroMet is a Python program developed by Conrad et al. (2021) that also considers antecedent soil saturation and creates bi-linear thresholds using landslide times to predict the conditions necessary to trigger slope instability. Positive pore-water pressure and piezometer response were selected as shallow landslide proxies in different stations. The program yielded statistically robust threshold models that suggest antecedent soil moisture conditions are not always negligible parameters in landslide hazard potential, implying that not all soils in the tropics have constant humid conditions (similar to Thomas et al., 2020).

Our in situ hydro-meteorological observations and threshold analysis indicate that using piezometer measurements as a proxy is preferable to solely using positive pore-pressure conditions for inferred “landslide times”. This approach is necessary because exact landslide times are unlikely to be available for any given event. Further data collection during high intensity rainfall events and the addition of more monitoring sites could improve the statistical performance of the forecasting efforts.