Estimating a-priori kinematic wave model parameters based on regionalization for flash flood forecasting in the Conterminous United States

JOURNAL OF HYDROLOGY

卷:541

Article

Vergara, Humberto^1,2,3,4  Kirstetter, Pierre-Emmanuel^2,3  Gourley, Jonathan J.²  Flamig, Zachary L.^1,2,3  Hong, Yang^3,4  Arthur, Ami^1,2  Kolar, Randall⁴ 

[1] Univ Oklahoma, CIMMS, Norman, OK 73019 USA

[2] NOAA, Natl Severe Storms Lab, Norman, OK 73069 USA

[3] Univ Oklahoma, ARRC, Norman, OK 73019 USA

[4] Univ Oklahoma, Sch Civil Engn & Environm Sci, Norman, OK 73019 USA

关键词: Kinematic wave routing;    Ungauged prediction;    Regionalization;    Fluvial hydraulics;    Multi-dimensional analysis;    Large sample hydrology;   

DOI  :  10.1016/j.jhydrol.2016.06.011

来源: Elsevier

【 摘 要 】

This study presents a methodology for the estimation of a-priori parameters of the widely used kinematic wave approximation to the unsteady, 1-D Saint-Venant equations for hydrologic flow routing. The approach is based on a multi-dimensional statistical modeling of the macro scale spatial variability of rating curve parameters using a set of geophysical factors including geomorphology, hydro-climatology and land cover/land use over the Conterminous United States. The main goal of this study was to enable prediction at ungauged locations through regionalization of model parameters. The results highlight the importance of regional and local geophysical factors in uniquely defining characteristics of each stream reach conforming to physical theory of fluvial hydraulics. The application of the estimates is demonstrated through a hydrologic modeling evaluation of a deterministic forecasting system performed on 1672 gauged basins and 47,563 events extracted from a 10-year simulation. Considering the mean concentration time of the basins of the study and the target application on flash flood forecasting, the skill of the flow routing simulations is significantly high for peakflow and timing of peakflow estimation, and shows consistency as indicated by the large sample verification. The resulting a-priori estimates can be used in any hydrologic model that employs the kinematic wave model for flow routing. Furthermore, probabilistic estimates of kinematic wave parameters are enabled based on uncertainty information that is generated during the multi-dimensional statistical modeling. More importantly, the methodology presented in this study enables the estimation of the kinematic wave model parameters anywhere over the globe, thus allowing flood modeling in ungauged basins at regional to global scales. (C) 2016 Elsevier B.V. All rights reserved.

【 授权许可】

Free