Lloyd Hock Chye Chua

A novel application of a neuro-fuzzy computational technique in event-based rainfall-runoff modeling

Intelligent computing tools based on fuzzy logic and Artificial Neural Networks (ANN) have been successfully applied in various problems with superior performances. A new approach of combining these two powerful AI tools, known as neuro-fuzzy systems, has increasingly attracted scientists in different fields. Although many studies have been carried out using this approach in pattern recognition and signal processing, few studies have been undertaken to evaluate their performances in hydrologic modeling, specifically rainfall-runoff (R-R) modeling. This study presents an application of an Adaptive Network-based Fuzzy Inference System (ANFIS), as a neuro-fuzzy-computational technique, in event-based R-R modeling in order to evaluate the capabilities of this method for a sub-catchment of Kranji basin in Singapore. Approximately two years of rainfall and runoff data which from 66 separate rainfall events were analyzed in this study. Two different approaches in the selection criteria for calibration events were adopted and the performance of an ANFIS R-R model was compared against an established physically-based model called Storm Water Management Model (SWMM) in R-R modeling. The results of this study show that the selected neuro-fuzzy-computational technique (ANFIS) is comparable to SWMM in event-based R-R modeling. In addition, ANFIS is found to be better at peak flow estimation compared to SWMM. This study demonstrates the promising potential of neuro-fuzzy-computationally inspired hybrid tools in R-R modeling and analysis.

Effect of externally generated turbulence on wave boundary layer

Abstract This experimental study deals with the effect of externally generated turbulence on the oscillatory boundary layer to simulate the turbulence in the wave boundary layer (WBL) under broken waves in the swash zone. The subject has been investigated experimentally in a U-shaped, oscillating water tunnel with a smooth bottom. Turbulence was generated ‘externally’ as the flow in the oscillator was passed through a series of grids that extended from the cover of the water tunnel to about mid-depth. Two different types of grid porosities were used. Direct measurements of the bed shear stress and velocity measurements were carried out. For the velocity measurements, mean and turbulence properties were measured in both the streamwise direction and in the direction perpendicular to the bed. A supplementary measurement for the undisturbed (without grids) case was also carried out, for comparison with the grid results. The mean and turbulence quantities in the outer flow region are increased substantially with the introduction of the grids. It is shown that the externally generated turbulence is able to penetrate the bed boundary layer, resulting in an increase in the bed shear stress, and therefore the friction coefficient. Other features related to the bed shear stress, such as transition, the friction factor and phase lead, are discussed. The range of the Reynolds number studied is Re=1×104–2×106.

Nutrients and suspended solids in dry weather and storm flows from a tropical catchment with various proportions of rural and urban land use

The results of an investigation characterizing the nutrients and suspended solids contained in stormwater from Kranji Catchment in Singapore are reported in this paper. Stormwater samples were collected from 4 locations and analyzed for the following eleven analytes: TOC, DOC, TN, TDN, NH(4)(+), NO(2)(-)+NO(3)(-) (NO(x)), TP, TDP, OP, SiO(2) and TSS. Stormwater was sampled from catchments with various proportions of rural and urban land use, including forested areas, grassed areas, agricultural and residential and commercial areas. The event mean concentrations (EMCs) of nutrients and TSS from sampling stations which have agricultural land use activities upstream were found to be higher. Comparison of site EMCs (SMCs) with published data showed that the SMCs of the nutrients and TSS are generally higher than SMCs reported for forested areas but lower than published SMCs for urban areas. Positive correlations (p<5%) were found between loading and peak flow at locations most impacted by ubanisation or agricultural activities. Correlation between loading and rainfall variables was less distinct. EMC was found to correlate less with rainfall and flow variables compared to pollutant loading. Unlike loading, no consistent pattern exists linking EMC to any particular storm or flow variable in any of the catchments. Lastly, positive correlations were obtained between the particulate forms of nitrogen and phosphorus and TSS.

Distribution and abundance of human‐specific Bacteroides and relation to traditional indicators in an urban tropical catchment

The study goals were to determine the relationship between faecal indicator bacteria (FIB), the HF183 marker and land use, and the phylogenetic diversity of HF183 marker sequences in a tropical urban watershed.

The Four National Taps of Singapore: A Holistic Approach to Water Resources Management from Drainage to Drinking Water

Water resources in Singapore are managed following the principles of a closed loop hydrologic cycle by one agency, the Public Utility Board (PUB), which promot…

Associations of chemical tracers and faecal indicator bacteria in a tropical urban catchment.

Surface water contamination by human faecal wastes is a widespread hazard for human health. Faecal indicator bacteria (FIB) are the most widely used indicators to assess surface water quality but are less-human-specific and have the potential to survive longer and/or occur naturally in tropical areas. In this study, 13 wastewater chemicals (chloride, boron, orthosphophate, detergents as methylene blue active substances, cholesterol, cholestanol, coprostanol, diethylhexyl phthalate, caffeine, acetaminophen, ibuprofen, sucralose and saccharin) were investigated in order to evaluate tracers for human faecal and sewage contamination in tropical urban catchments. Surface water samples were collected at an hourly interval from sampling locations with distinct major land uses: high-density residential, low-density residential, commercial and industrial. Measured concentrations were analysed to investigate the association among indicators and tracers for each land-use category. Better correlations were found between different indicators and tracers in each land-use dataset than in the dataset for all land uses, which shows that land use is an important determinant of drain water quality. Data were further segregated based on the hourly FIB concentrations. There were better correlations between FIB and chemical tracers when FIB concentrations were higher. Therefore, sampling programs must be designed carefully to take the time of sampling and land use into account in order to effectively assess human faecal and sewage contamination in urban catchments. FIB is recommended as the first tier in assessment of surface water quality impairment and chemical tracers as the second tier. Acetaminophen and coprostanol are recommended as chemical tracers for high-density residential areas, while chloride, coprostanol and caffeine are recommended for low-density residential areas.

Sensor Placement and Measurement of Wind for Water Quality Studies in Urban Reservoirs

We collaborate with environmental scientists to study the hydrodynamics and water quality in an urban district, where the surface wind distribution is an essential input but undergoes high spatial and temporal variations due to the complex urban landform created by surrounding buildings. In this work, we study an optimal sensor placement scheme to measure the wind distribution over a large urban reservoir with a limited number of wind sensors. Unlike existing sensor placement solutions that assume Gaussian process of target phenomena, this study measures the wind which inherently exhibits strong non-Gaussian yearly distribution. By leveraging the local monsoon characteristics of wind, we segment a year into different monsoon seasons which follow a unique distribution respectively. We also use computational fluid dynamics to learn the spatial correlation of wind in the presence of surrounding buildings. The output of sensor placement is a set of the most informative locations to deploy the wind sensors, based on the readings of which we can accurately predict the wind over the entire reservoir surface in real time. 10 wind sensors are finally deployed around or on the water surface of an urban reservoir. The in-field measurement results of more than 3 months suggest that the proposed sensor placement and spatial prediction approach provides accurate wind measurement which outperforms the state-of-the-art Gaussian model based or interpolation based approaches.

Information recovery from measured data by linear artificial neural networks-An example from rainfall-runoff modeling

The results of a study using linear artificial neural networks (ANNs) to determine the physical parameters in event-based rainfall-runoff modeling are presented in this paper. The input structure of the ANN was determined based on an analysis of the discretized form of the kinematic wave equations, and the physical parameters were obtained through a back calculation of the weights and biases of the ANN. Two cases were considered; using ANNs trained on datasets derived from: (1) effective rain of entire events and (2) total rain of wet portion of events only. For Case (1), the parameters @Dt and @a(=S/n) were determined with excellent accuracy. For Case (2), in addition to @Dt and @a, the constant loss rate, @F, was also determined with excellent accuracy. Further, the use of total rain of entire events (a common practice in ANN applications in rainfall-runoff modeling) was also investigated, and it was found that the results from this analysis are less realistic compared to those of Case (2).

Controlled field studies on soil aquifer treatment in a constructed coastal sandfill

A controlled artificial recharge experiment was conducted to investigate the effect of soil aquifer treatment during percolation of secondary and tertiary (ultrafiltered) treated wastewater through the shallow vadoze zone of a newly constructed coastal sandfill. The sandfill is a reclaimed land constructed from marine sand dredged from the seabed. To obtain 1-D flow, a stainless steel column was driven to a depth of 2.5 m, penetrating the phreatic surface. Wastewater was percolated through the column under fully-saturated and unsaturated conditions. Infiltration rates, dissolved organic carbon (DOC) and ultra-violet absorption (UVA) were monitored. The wastewaters were recharged at similar infiltration rates of approximately 5.5 m/day and 3.5 m/day under fully-saturated and unsaturated conditions, respectively. In both cases, clogging occurred 40 days after the start of recharge, under saturated conditions. For secondary treated wastewater, DOC concentration (mg/l) reduced by 28% and 13% under unsaturated and saturated conditions, respectively. The corresponding UVA reduction was 19.4% and 14.1%. Similar reductions in DOC were observed for the tertiary treated wastewater; however, the reduction in UVA was higher; 28% and 22% under unsaturated and saturated conditions, respectively. On an mass removal (mg/m(2) DOC) basis, DOC reduction appeared to be more significant under unsaturated conditions. This is attributed to the presence of interstitial oxygen.

Online Ensemble Modeling for Real Time Water Level Forecasts

Accurate and reliable flood forecasting is essential to mitigate the threats brought by floods. Ensemble approaches have been used in limited studies to improve the forecasts of component models. In this paper an ensemble model based on neural-fuzzy inference system (NFIS) and three real time updating approaches were used to synthesize the water level forecasts from a Adaptive-Network-based Fuzzy Inference System (ANFIS) model and the Unified River Basin Simulator (URBS) model for three stations in Lower Mekong. The NFIS ensemble model results are compared with the simple average model (SAM) which is adopted as a benchmark ensemble model. The ensemble model of offline learning without real time updating (EN-OFF), ensemble model with real time updating using offline learning (EN-RTOFF), ensemble model with real time updating using online learning (EN-RTON1) and ensemble model with real time updating using online learning and sub-models (EN-RTON2) were studied in this paper. Statistical analysis of the models for all the three stations indicated the superiority of the EN-RTON2 model over EN-RTOFF, EN-RTON1 models, SAM and the EN-OFF model. Not only the spikes in the URBS model were eliminated, but also the time shift problems in the ANFIS model results were decreased.