SubProject #4

Developing a Rule-based Probabilistic FF & DF Forecasting System


Objective: Develop a rule-based probabilistic FF & DF forecasting system adapted to provide real time warnings using rainfall as the main triggering factor in combination with additional features, such as terrain properties, natural and anthropogenic debris availability, soil saturation and socio-economical characteristics.

Task 4.1: Identification of the governing variables to be included in the rule-based system. The most important are related to the rainfall input, terrain morphology and natural / anthropogenic debris availability. Provide a translation between these variables used in the equations of the models able to simulate FF & DF and the measurable input variables to be used in the forecasting system. Explore all the range of these governing variables and determine the range of values associated to FF & DF generating events [UPC-GITS, CUGRI, WSL, ULANC, WU].


Task 4.2: Develop a methodology to perform the hydrological-hydraulic simulations for the different selected combinations as input variables of the system in Task 4.1. The simulation methodology will include a high resolution version of the hydrological models used in WP3 as well as simplified 1D and 2D hydraulic models able to deal with sediment and flows, to assess DF occurrences or even slope failures in a probabilistic way (self developed codes, e.g. FLATMODEL, Medina et al. 2007 or commercial codes e.g. FLO-2D, DAN-W; O’Brien et al. 1993; Hungr 1995 will be used) [UPC-GITS, CUGRI, WSL, ULANC, WU].

Task 4.3: Simulate a huge number of hypothetical cases covering the entire range of values of the input variables representing the triggering factors identified in Task 4.1. The simulation will be performed at high resolution, but in virtual places described by the combination of the values of the input variables. The goal of all these simulation runs is to assess the potential hazard (relation between flow energy and probability) and to provide knowledge base to develop the rule system. [UPC-GITS, CUGRI].

Task 4.4: Develop the rule-based system identifying links between the values of the input variables assessing the hazard and the model outputs, or consequences, obtained by the simulation methodology. The use of additional input variables describing socio-economic aspects, infrastructure proximity as well as information about damage assessment, will allow to translate the model outputs into risk and vulnerability assessments [UPC-GITS, CUGRI, WSL, ULANC, WU].

Task 4.5: Develop an operationally oriented fuzzy-logical methodology based on the rule-based knowledge established in Task 4.4 able to assign in real-time consequences, even in terms of risk and vulnerability (including damage assessment), from the values of the input variables related to a given rainfall forecast over a given hillslope [UPC- CRAHI, ULANC].