Application of Sensitivity Analysis for Process Model Calibration of Natural Hazards

Abstract

Sensitivity analysis (SA) describes how varying inputs to a model subsequently varies its outputs. Its inclusion can support the systematic calibration of numerical models to back-calculate intensity properties of past torrent events that would otherwise be difficult or impossible to collect during their occurrence. Sensitivity analysis for model calibration is assessed with the back-calculation of a known torrent event. In particular, FLO-2D, a cell-based numerical model, is used to simulate the 2005 debris flow event that occurred in Brienz, Switzerland. Under 4000 simulations were completed with ranges of physically reasonable parameter values. Model results were compared in 3-dimensions with both sediment deposition extents (x, y) and estimated deposition heights (z) from available post-event images. The comparisons highlighted that more accurate input and validation data, namely the flow behavior of hazardous processes and post-event deposition heights, are required to produce stronger agreements between simulated and observed results. Furthermore, the application of SA for model calibration supports systematic exploration of large parameter spaces characteristic of complex phenomena like natural hazard events. These findings demonstrated how important model input factors can be identified, which provide guidance for future data collection efforts to capture both the rheology and the spatial distribution of hazards more accurately.

Evaluation of Debris Flows for Flood Plain Estimation in a Small Ungauged Tropical Watershed for Hurricane Otto

“…The rheological parameters found in the literature [62] and the ones estimated by density curves from soil samples generated the opportunity to choose suitable rheological values in the model. However, Chow et al [4] considered that rheology calculated from soil samples generates much higher agreement in modeled results than the reported in the literature.…”

“…The same behavior was found in HE simulations with a tendency to increase the flood area (8.28 ha), as shown in Figures 4i and 5i. In addition, the sensitivity analysis demonstrated that the Cv is the most important parameter for debris flow computation [4]. Therefore, debris mapping and real-time floodplain estimates would be utilized in the emergency management and flood alert systems to minimize risk using possible Cv based on historical estimations.…”

“…The impact of debris flow is aggravated if there is not a preliminary identification of floodplains, making local risk management and early warnings impossible to apply, especially at ungauged small watersheds. Several methodologies have been utilized to perform debris flood hazard assessments for gauged watersheds and focus on calibrated hydrological and two-dimensional hydraulic models and rheological parameters quantified from field measurements during the event [3][4][5][6][7]. In ungauged watersheds, some authors have applied regionalization for peak flow event estimation [8] and continuous flow modeling [9,10].…”

Sensitivity and identifiability of rheological parameters in debris flow modeling

“…The maximum flood area and deposited volume are relatively easy to measures, and are probably the most used post-event measurements for calibrating debris flows models (Chow et al, 2018;Cesca and D'Agostino, 2008;D'Agostino and Tecca, 2006;Sosio et al, 2007;Frey et al, 2016;Quan Luna et al, 2011).…”

“…Boniello et al (2010) compared FLO-2D model results from a set of 12 back-calculated rheological parameters selected from previous studies, with another set of parameter values obtained from laboratory rheological analyses, finding a better representation of debris flow behavior with back-calculated parameters. Chow et al (2018) conducted simulations with FLO 2D using 26 different sets of rheological parameters obtained from previous studies, combined with different values of volumetric sediment concentration C v , specific gravity G s and the surface detention SD, a parameter used in FLO2D to represent flow detention. They found that the most important parameters were C v , SD and β 1 , which characterizes fluid viscosity.…”