摘要：AbstractBackgroundThis study tested a first-order perturbation method based on Karhunen-Loevè expansion (FP-KLE), to analyze flood inundation modeling under uncertainty. The floodplain roughness over a 2-dimensional domain was assumed to be a statistically heterogeneous field with log-normal distributions. Firstly, we attempted to use KLE to decompose the random field of log-transferred floodplain roughnessN(x), which was based on the eigenvalues and eigenfunctions of the covariance function ofN(x), and a set of orthogonal normal random variables. Secondly, the maximum flow depths were expanded by the first-order perturbation method by using the same set of random variables as used in the KLE decomposition. Then, a flood inundation model, named FLO-2D, was adopted to numerically solve the corresponding perturbation expansions.ResultsTo illustrate the methodology, a one-in-five-years flood event was chosen as the study case. The results indicated that the mean of the maximum flow-depth field obtained from the proposed method was fairly close to that from Monte Carlo Simulation (MCS), but the standard deviation was somewhat higher. However, the FP-KLE method was computationally more efficient than MCS.ConclusionsThe study verified the applicability of FP-KLE in handling uncertainties of flood modeling in a more efficient manner; further test with multiple inputs of random fields is desired.
关键词：KeywordsEnKarhunen-Loevè expansionRoughness coefficientFlood inundation modellingMonte Carlo simulations
