Azmi Ab Rahman

Uncertainty Analysis of HEC-HMS Model Parameters Using Monte Carlo Simulation

Abstract Rainfall-runoff modelling was carried out to determine Hydrologic Engineering Centers Hydrologic Modeling System model parameters for catchments in Johor, Malaysia. Uncertainty analysis, using Monte Carlo simulation, with a sample trial of 5,000 is then applied to the model parameters based on uniformly distributed samples. Sixteen sub-catchments drawn from Sungai Johor, Sungai Tebrau, Sungai Skudai and Sungai Segamat were selected. The rainfall-runoff modelling gives a mean correlation coefficient of 0.9038 and mean efficiency index of 81.53%, indicating a good prediction. Results from the uncertainty analysis give values in the range of 1.25–14.99mm for initial loss, with a mean of 8.09 mm. The constant loss rate takes a range of 7.98–299.87 mm/h with mean value of 153.55 mm/h. The range of values for imperviousness is between of 0.017% and 44.99% with mean value of 22.47%. Simulation result for time of concentration is in the range of 0.44–14.53h with mean value of 7.62h. The storage coefficient takes values from 0.22h to 5.16h with mean value of 2.71h. The outcome of this work quantifies the uncertainty associated with these parameters and the result can be used to narrow down the range at which initial estimates are made during future modelling processes.

Sediment phosphorus loadings uncertainty using fuzzy-logic in detention pond at Southern Malaysia

This study vitalized the uncertainty and fuzzy rules consideration in the detention pond sediment phosphorus loadings estimation .Sediment phosphorus is pollutant of primary concern as it may result in adverse environmental effect of the hydrological systems.Detention pond establishmentbecome a practical control measures to improve the downstream water quality.The purpose of this study is to estimate the sediment phosphorus loadingsuncertainty for the detention pond at UniversitiTeknologi Malaysia (UTM) using Fuzzy Logic MUSLE-IHACRES model.The IHACRES model was used to predict the streamflow into the detention pond that later be used in the MUSLE model.The streamflow modelling using IHACRES model showed that the largest bias was -1.2157 mm/day and the smallest bias is -0.0265 mm/day. The highest r2 value was 0.9204 while the lowest was 0.7813. Both statistics showed that the model was able to perform well. The resultsfrom MUSLE-IHACRES model showed that the maximum value ofsediment yield was 254.88 tons/ha /yr and minimum value was 84.40 tons/ha/yr.This study also forecast accumulated sediment loads and depth using MUSLE-IHACRES over 10-100 years. The highestfrequencyor most likely range of sediment phosphorus wasless than 350 ton/year (257tons/ha/year). The average phosphorus content in sediment obtained was 34.0 mg/g and the sediment phosphorus loadings estimated using fuzzy logic was 28.72ton/year.The sediment phosphorus distribution investigated using Fuzzy Logic was presented. Suggested control measures need to be established as the sediment deposition is degrading the pondand approaching the Eutrophic Status.