Assessment of flood frequency after forest fires in small ungauged basins based on uncertain measurements

Abstract

A methodology is proposed for constructing flood frequency curves before and after forest fires in small ungauged basins. It follows the rainfall-based derived distribution approach using Monte Carlo simulation and allowing for uncertainty in some of the variables involved. The generalized extreme value distribution for annual maximum rainfall is used, combined with the SCS-CN method for estimating rainfall excess, and the rational method for peak flow estimation. For pre-fire conditions, uncertainty in the spatially-averaged curve number is assumed. For post-fire conditions, measurements of peak flood water level at the basin outlet are exploited, while uncertainty is considered both in these measurements and in the hydraulic roughness. Multi-variate Latin Hypercube sampling for correlated variables is used to assess flood quantiles. The methodology is applied to a small basin in Attica, Greece, which was burnt in 2007. Sensitivity tests are carried out regarding the effect of sample size and of the amount of uncertainty involved.Key words flood frequency; forest fires; uncertainty analysis; derived distribution approach; Monte Carlo simulation; Latin Hypercube sampling; generalized extreme value distribution; SCS-CN method; rational method; Torrent Giannoula; Attica, Greece Evaluation de la fréquence des crues après les incendies de forêt basée sur des mesures incertaines dans de petits bassins non jaugés Résumé Une méthodologie est proposée pour construire des courbes de fréquence des crues, avant et après incendies de forêts dans des petits bassins non jaugés. Elle suit l'approche de la distribution dérivée basée sur les pluies en utilisant une simulation Monte Carlo et en autorisant de l'incertitude dans certaines des variables impliquées. La distribution des valeurs extrêmes généralisée est utilisée pour les précipitations maximales annuelles. Elle est combinée avec la méthode SCS-CN pour l'estimation des précipitations en excès, et avec la méthode rationnelle pour l'estimation du débit de pointe. Pour les conditions pré-incendie, on suppose une incertitude dans le "curve number" spatialement moyenné. Pour les conditions post-incendie, les mesures du niveau d'eau de la pointe de crue à l'exutoire du bassin sont exploitées, tandis que l'incertitude est considérée à la fois dans ces mesures et dans la rugosité hydraulique. Un échantillonnage par hypercube latin multi-varié pour variables corrélées est utilisé pour évaluer les quantiles de crues. La méthodologie est appliquée à un petit bassin de l'Attique, en Grèce, qui a brûlé en 2007. Des tests de sensibilité sont réalisés sur l'effet de taille d'échantillon et le niveau d'incertitude en jeu.Mots clefs fréquence des crues; feux de forêt; analyse d'incertitude; approche de la distribution dérivée; simulation de Monte

Evaluation of Hydrological and Hydraulic Models Applied in Typical Mediterranean Ungauged Watersheds Using Post-Flash-Flood Measurements

“…Additionally, Segura-Beltrán et al [13] analyzed the flash flood of the Girona River (Spain) in 2007 and found that some human-induced factors, such as artificially narrow stream width, walls, and roads, caused the overbank flow. Similar findings from other researches [12,36,[60][61][62][63] in watersheds very close to the study area (Chalkidiki region) showed the significant influence of…”

“…The lag time was defined by calculating the difference between the time of the precipitation peak and the time of the observed peak discharge at the outlet point of the watershed. To estimate the hydrographs of the two rainfalls (October 2006 and February 2010) in the Vatonias watershed, the SCS-CN model was used, a well-known and widely applied model in many countries [29][30][31][32][33] and in Greece [22,[34][35][36][37][38][39][40][41][42][43]. The CN is a dimensionless empirical parameter for predicting runoff and infiltration from rainfall excess [39].…”

Flood Depth-Damage Functions for Built Environment

“…The accuracy of flood hazard assessments is limited by the fact that most drainage basins, including urban ones, are poorly gauged (Nalbantis 1995) or even ungauged and affected by changes such as forest fires (Nalbantis and Lymperopoulos 2012;Batelis and Nalbantis 2014). When advancing a step further, specifically to flood damage assessments, the data scarcity problem is accentuated and damage estimation methods inevitably become crude Merz et al 2010).…”

Effect of uncertainty in Digital Surface Models on the extent of inundated areas

“…In cases with erroneous DSMs or rainfall, we use quantiles with 90% probability of nonexceedance, that is, the 90% quantile of A , d i , V i , and W i , respectively denoted as A90, d90 i , V90 i , and W90 i for each cross‐section i . The selection of probability is dictated by two conflicting requirements (Candela, Aronica, & Santoro, ; Nalbantis & Lymperopoulos, ): the desire for a high degree of safety and an acceptable accuracy that is achievable using reasonable sample sizes in Monte Carlo simulations.…”