HYLAND · Hydrometeorological Controls and Warning Procedures for Shallow Landslides in an Alpine Region

Hydrometeorological Controls and Warning Procedures for Shallow Landslides in an Alpine Region

Rainfall-triggered landslides are among the most widespread geological hazards on Earth. Shallow land sliding in steep, soil-mantled landscapes can generate debris flows that scour low-order channels, deposit large quantities of sediment in higher-order channels and pose a significant hazard. As a consequence of climatic changes and potential global warming, an increase of landslide activity is expected in the future, due to increased rainfall, changes of hydrological cycle, more extreme weather and concentrated rain within shorter periods of time. Thus enhancing societal response to these events is mandatory to reduce vulnerability. The effective management of landslide hazard requires understanding of the rainfall conditions that result in slope instabilities. Accurate derivation of critical rainfall thresholds is of high importance because they can be used as the basis for establishing landslide warning systems. In addition to the accurate rainfall estimates required to correctly establish the critical rainfall thresholds, knowledge on the dynamics of soil water content has also proven to be important for understanding and predicting landslides. The objective of HYLAND is to advance understanding of hydrometeorological controls on shallow landslides at a regional scale with the ultimate goal to advance early warning procedures for this type of hazards. HYLAND involves observational and modeling studies of rainfall-triggered landslides occurred in the Upper Adige region (~7000 km2) in Central Alps. The region is characterized by rugged topography, with elevations ranging from 225 m a.s.l. to 3900 m a.s.l. The core of the methodological approach followed in this project involves the combination of detailed observations (in-situ and remote sensing) with information from a hydrological model applied to the whole area, including sites that were impacted by landslides and sites that were not. The four principal components available over a period of eleven years (2000-2010) are: 1) observations of landslide/debris flow locations 2) rainfall data from raingauges and weather radar 3) streamflow data at a number of catchments 4) a semi-distributed hydrologic model calibrated and validated over the area. Specific objectives addressed within this project include: i) Developing and testing a coherent set of methodologies to support radar-based estimation of landslide/debris flow triggering rainfall. ii) Investigating the relationships between initial soil wetness conditions and rainfall properties for landslide/debris flow events. iii) Developing and testing a set of procedures to predict shallow landslides/debris flows in space and time for an alpine region. To address these objectives we utilized the available dataset and developed and validated a methodology for creating radar-rainfall fields that provided accurate space/time representation of triggering rainfall for a number of landslides/debris flows in the study area. This allowed us to gain very important insight on the structure of rainfall fields at and around debris flow initiation points. As a consequence we were able to a) derive more accurate thresholds for critical rainfall (i.e. responsible for triggering the hazard) and b) assess the impact of rainfall sampling uncertainty on the derivation of critical rainfall thresholds. The later lead to important findings regarding the significance of rainfall sampling uncertainty on the estimation of rainfall thresholds and highlighted the great potential of using remote sensing observations (relative to in-situ measurements) for warning applications. At a second step, having accurate rainfall information at locations where landslide/debris flows initiated, allowed us to assess with a higher degree of certainty the dependence between rainfall thresholds and antecedent moisture conditions (derived from the hydrologic model applied). Results from this analysis revealed that antecedent soil moisture conditions did not have a strong control on the rainfall thresholds derived from the cases available. A potential explanation is related to the fact that, as it was discovered, the majority of debris flow cases analyzed were triggered due to channel bed mobilization and much less due to hillslope failure, which is know to be highly dependent to antecedent soil moisture conditions. According to these findings, an extra research avenue was opened in which we seek to identify the critical discharge responsible for the channel bed mobilization. This effort lead to the development of a framework that combined hydrological and geomorphological modeling in order to provide spatially distributed estimates of critical discharge along the channel path. Results from this framework were demonstrated for a number of available cases. Overall, results and findings within the context of HYLAND project highlighted a number of important aspects regarding the uncertainty in estimation of critical rainfall thresholds and their dependence to hydrologic conditions. The impact of these findings is expected to be important from both scientific and societal perspective. From the scientific point of view, the research carried out introduced several novel elements to existing literature and set the foundation for numerous other research questions to be explored. It is therefore expected that scientists working on similar topics will be benefited from the work and results obtained from HYLAND as well as from the dataset produced. The core dataset used within HYLAND can be found online at the project’s website (http://intra.tesaf.unipd.it/cms/hyland/default.asp) and it has been made available, for scientific purposes, to everyone. From a societal point of view, improvements on the operational use of critical rainfall thresholds, following the work in HYLAND, can potentially have a great impact in mitigating the risk during landslide/debris flow hazards and therefore reduce the impact of these hazards on human lives and infrastructure. National or regional authorities dealing with the operational early warning systems (e.g. Civil Protection Agencies) can certainly benefit from the research finding within HYLAND. HYLAND website: http://intra.tesaf.unipd.it/cms/hyland/default.asp

Data: CORDIS, © European Union