New Derivation Method of Region Specific Curve Number for Urban Runoff Prediction at Melana Watershed in Johor, Malaysia

2nd International Conference on Architecture and Civil Engineering

土木建筑工程

Tan, W.J.^1 ; Ling, L.^1 ; Yusop, Z.^2 ; Huang, Y.F.^1

Centre of Disaster Risk Reduction, Lee Kong Chian Faculty of Engineering and Science, Civil Engineering Department of Universiti Tunku Abdul Rahman, Jalan Sungai Long, Cheras, Bandar Sungai Long

43000, Malaysia^1

Centre for Environmental Sustainability and Water Security, Faculty of Civil Engineering Department, Universiti Teknologi Malaysia, Skudai, Johor

81310, Malaysia^2

关键词: Bootstrapping;    Calibration methodologies;    Curve numbers;    Inferential statistics;    Rainfall;    Runoff modelling;    Rainfall-runoff characteristics;    Runoff prediction;    Soil conservation services;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/401/1/012008/pdf
DOI  :  10.1088/1757-899X/401/1/012008

学科分类：土木及结构工程学

来源: IOP

【 摘 要 】

The Soil Conservation Services (SCS) rainfall-runoff model has been incorporated into many types of software and adopted by Malaysian government agencies for design use. However, hydrologists argue the accuracy of the predicted runoff results from this model in recent decades. Malaysia does not have a regional specific curve numbers available for the use in rainfall-runoff modelling, and therefore SCS-CN practitioner has no option but to adopt its guideline and handbook values which are specific to US region. Unlike the conventional SCS-CN technique, the proposed calibration methodology in this paper discarded the use of CN as input to the SCS model, rearranged the model equation and derived statistically significant CN value(s) of a specific region through rainfall-runoff events directly through the use of inferential statistics. The SCS rainfall-runoff framework can now be calibrated quickly to address urban runoff prediction challenge under rapid land use and land cover or climate changes with the proposed methodology. Between July and October of 2004, the derived λ was 0.0005 while λ = 0.20 was found to be statistically insignificant at alpha = 0.01 level at Melana watershed in Johor, Malaysia. Optimum CN of 90.45 was derived to represent the rainfall-runoff characteristic of Melana watershed. Runoff prediction error was reduced by 90% and achieved the Nash-Sutcliffe (NS) index of 0.86. The new method can provide CN adjustment guidelines for SCS practitioners and any software which incorporated the model to predict urban runoff.