Search

Droughts are constantly threatening the global water availability and food securities worldwide. This study aims to evaluate the short- and long-term (1-, 6- and 12-month) drought conditions in Peninsular Malaysia during 1989-2018 using Standardized Precipitation Index and Evaporative Demand Drought Index. Historical trends of drought conditions were analyzed using modified Mann-Kendall test. Spearman’s ρ approach was also applied to examine the spatial patterns of correlations between these drought indices. Based on the findings, Evaporative Demand Drought Index shows increasing tendency towards drier conditions in the northern half of Peninsular Malaysia, but opposite trends are observed for Standardized Precipitation Index. The time series of Evaporative Demand Drought Index are generally well-correlated to that of Standardized Precipitation Index at all three timescales for the whole study area, except for the northern region. The evidence presented suggests Evaporative Demand Drought Index is a great alternative for drought monitoring applications in Peninsular Malaysia.

The core objective of this study is to carry out rainfall forecasting over the Langat River Basin through the integration of wavelet transform (WT) and convolutional neural network (CNN). The proposed method involves using CNN for feature extraction to efficiently learn from the raw rainfall dataset. With the aid of deep architecture, a highly abstracted representation of the inputs time series with a high level of interpretation is formed at each subsequent CNN layer. The use of WT in forecasting the rainfall time series is by preprocessing the raw rainfall dataset into a set of decomposed wavelet components as inputs for the CNN model using discrete wavelet transform (DWT). The conditions for discretizing the raw input through DWT are discussed, along with the criteria to be used. Daily datasets, ranging from January 2002 to December 2017, were used. The results showed that the proposed model could satisfactorily capture patterns of the rainfall time series, for both monthly rainfalls forecasting or daily rainfall forecasting. Three performance indices were used to evaluate the model accuracy: RMSE, RSR, and MAE. These statistical indices have a range of value from 0 to a finite value that depends on the scale of the number used. In general, a lower value is better than a higher one.

Even though satellite precipitation products have received an increasing amount of attention in hydrology and meteorology, their estimations are prone to bias. This study investigates the three approaches of bias correction, i.e., linear scaling (LS), local intensity scaling (LOCI) and power transformation (PT), on the three advanced satellite precipitation products (SPPs), i.e., CMORPH, TRMM and PERSIANN over the Langat river basin, Malaysia by focusing on five selected extreme floods due to northeast monsoon season. Results found the LS scheme was able to match the mean precipitation of every SPP but does not correct standard deviation (SD) or coefficient of variation (CV) of the estimations regardless of extreme floods selected. For LOCI scheme, only TRMM and CMORPH estimations in certain floods have showed some improvement in their results. This might be due to the rainfall threshold set in correcting process. PT scheme was found to be the best method as it improved most of the statistical performances as well as the rainfall distribution of the floods. Sensitivity of the parameters used in the bias correction is also investigated. PT scheme is found to be least sensitive in correcting the daily SPPs compared to the other two schemes. However, careful consideration should be given for correcting the CMORPH and PERSIANN estimations.

Although satellite precipitation products (SPPs) increasingly provide an alternative means to ground-based observations, these estimations exhibit large systematic and random errors which may cause large uncertainties in hydrologic modeling. Three approaches of bias correction (BC), i.e. linear scaling (LS), local intensity scaling (LOCI), and power transformation (PT), were applied on four SPPs (TRMM, IMERG, CMORPH, and PERSIANN) during 2014/2015 extreme floods in Langat river basin, and the performance in terms of rainfall and streamflow were investigated. The results show that the original TRMM had a potential to predict the peak streamflow although CMORPH show the best performance in general. After performing BC, it is found that the LS-IMERG and LOCI-TRMM show the best performance at both rainfall and streamflow analysis. Generally, it is indicated that the current SPP estimations are still imperfect for any hydrological applications. Cross validation of different datasets is required to avoid the calibration effects of datasets.

Wall slip can be defined as a phenomenon in the flow of suspensions due to the movement of particles away from the wall boundary, leaving a thin liquid rich layer adjacent to the wall. It should be taken into consideration in material designing, manufacturing and transportation as it may cause inaccurate rheological measurements such as shear rate and viscosity. The apparent (measured) shear rate is higher than the actual shear rate. The traditional method for actual shear rate determination is not efficient from the perspective of time and cost consumption. Therefore, there is a need to develop a mathematical model that is able to detect the complex pattern of the actual shear rate and accurately predict its value based on the available measured input variables. Volumetric concentration, particle size, temperature and shear stress are selected as the input, while the actual shear rate is kept as output while designing the architecture of ANFIS model. Sixteen ANFIS models with different architecture were designed and evaluated using different statistical indices. Model XVI with number of membership function equal to 3, input product of two sigmoid membership function, output linear membership function type integrated with hybrid optimization method appears to be the most suitable model architecture for predicting the actual shear rate.

The most important and indispensable substance for life on earth is water. Regrettably, the growths of civilization, population, and industrssialization have led to a steady deterioration in the quality of available water resources. Heavy metals are a particular problem because of their high toxicity. This work highlights up-to-date methods for the removal of heavy metals from water using the technique of adsorption. After a comprehensive review of the range of materials used as adsorbents to remove heavy metals, the paper focuses on one particular technique, involving carbon nanotubes (CNTs). Tests have confirmed that CNTs are highly efficient, but have also revealed some drawbacks. Further adjustments and refinements are therefore required to improve their ability to remove heavy metals. This study shows that using deep eutectic solvents (DESs) as a functionalization agent to replace ionic liquids (ILs) has a number of advantages, including lower costs, easier availability of materials and environmentally friendliness. In sum, it suggests that the most efficient adsorption process is to use functionalized CNTs with DESs as an adsorbent. The critical analysis and conclusions emerging from this study should be of benefit to engineers, environmental scientists and chemists interested in the use of carbon nanotubes for use in environmental remediation.

Wall slip is a phenomenon in which particles migrate from solid boundaries, leaving a thin liquid rich layer adjacent to a wall, which can affect the measurement of the rheological properties. Curr...

The presence of wall slip in concentrated suspensions affect the rheological measurements such as shear stress, shear rate, and viscosity. The measured shear rate will have a value lower than the actual shear rate. Therefore, it is important to have a study on such scenario as the concentrated suspensions have a wide industrial application. The current method for actual shear rate prediction is a challenging task and quite time-consuming as several experimental works are required. Therefore, the development of a mathematical model with an acceptable accuracy is required. Multi-layer perceptron neural network (MLP-NN) is employed to develop the prediction model by applying shear stress, volumetric concentration, particle size, and temperature as the input parameters while the actual shear rate is kept as the output variable. MLP-NN model with 9 hidden neurons is the model that has achieved the best performance in term of statistical analyses. Besides, MLP-NN models for five different temperature ranges (i.e., 11–20 °C, 21–30 °C, 31–40 °C, 41–50 °C, and 51–60 °C) are also proposed to investigate the potential of the developed model to achieve a better accuracy and ease the user while dealing with the specified temperature range. It is found that the MLP-NN models with the best performance for each temperature range are the models with 8 hidden neurons, 10 hidden neurons, 8 hidden neurons, 9 hidden neurons, and 9 hidden neurons respectively. The novelty of this research study is the application of artificial intelligence method to mimic rheological actual shear rate under wall slip condition.

The effect of northeast monsoon rainfall to soil erosion on hillslope adjacent the Guthrie Corridor Expressway, Malaysia, is investigated by examining a relationship between vegetation covers and slope steepness. Five plot-scale study areas [i.e., NBNM, PDM, PDNM, NBM and natural dense microbe (NDM)] are marked on the hillslope with different percentage and type of vegetation densities. Experimental equipment is set up at those plots to collect rainwater and soil loss starting from November 2014 until March 2015 with twice collection per month. Based on the data collected, the effect of different vegetation densities associated to soil loss is examined. These are the conclusions that can be made based on this investigation: (1) the observed rainfall pattern shows that mostly the recorded rainfall depth is about 5 mm with 20 min duration of rainfall; (2) plot without vegetation cover (i.e., NBNM) yields greater soil loss with the maximum runoff, whereas the least is exhibited by the plot having NDM. Meanwhile, the rate of soil loss is found greater for the planted vegetation compared to the natural plant; (3) a positive relationship is found between runoff and slope steepness regardless of the vegetation cover types.

Kelantan is a flood-prone area where in the past years flood had occurred quite frequently. Determining a hydrological model that can represent Kelantan River basin by giving plausible simulated runoff according to the observed runoff is essential as this will allow appropriate prediction of future flood by using forecasted rainfall and other data. In this study, the IHACRES model was used to simulate runoff and the calibrated simulated runoff by daily scale and seasonal flood events were compared with observed runoff. In general, the IHACRES model performed better in seasonal scale as compared with annual scale in terms of calibration. However, performance of IHACRES degraded during validation stage, whereby the model tends to underestimate the high peak flows but estimate rather more accurate when no peak flows were present. In terms of annual scale, the best model was obtained by calibrating the streamflow in 2012 – 2013 (2 years), the validation results were not satisfactory with NSE = 0.473 and PBIAS = 27.7%. On the other hand, for seasonal analysis, the best model was obtained by calibrating the data of NEM 6 (November 2017 – March 2018). 3 out of 5 of the validation periods show unsatisfactory results (NSE ≤ 0.50). NEM 1 (November 2012 – March 2013) show the best validation results with NSE = 0.853. Further calibration is required in order to enhance the accuracy of the model.

Sungai Selangor is very important from the viewpoint of water supply and multipurpose water use in Malaysia. The water quality of this river is degrading due to point and non-point sources of pollution. This study, focus on the water quality assessment and simulation the effect of the pollution sources from urbanization to the Sungai Selangor basin. Water quality Index (WQI) is used to define the status of river water quality and the QUAL2K was used as a simulation model. Water quality parameters DO, BOD and NH3-N have been chosen for modeling. In addition, five different model scenarios were simulated to observe the impacts of pollution sources on the Sungai Selangor water quality. WQI results showed that most of the stations in this river basin recorded water inferior to Class III. The water quality model presented different scenarios for changes of Sungai Selangor water quality. Simulation results for different scenarios showed that reduced levels of BOD and NH3-N at 51.10% and 66.18%, respectively, can be obtained if Scenario-5 is employed. The river water quality issue in the Rawang sub- basin within the study area is considered crucial to create significant improvement within the sub basin and in the downstream area of Sungai Selangor basin.

Heavy metal concentration has become a major concern for water quality in rivers. Rapid urbanization and industrialization contribute to heavy metal concentrations in river water. This study aims to investigate the distribution, source, and environmental risk of heavy metals in the Selangor River basin in Malaysia. A total of 132 water samples were collected from 11 sampling stations on a monthly basis over a one-year period. Thirteen heavy metals were analyzed using inductively coupled plasma optical emission spectroscopy (ICP-OES). In order to identify the sources of heavy metals along the river basin, multivariate statistical techniques like principal component analysis (PCA) and cluster analysis (CA) were performed. It was found that As, Mn, and Fe exceeded the admissible limits of the Malaysian National Standard Water Quality (NSDWQ) at several of the sampling stations. Heavy metal pollution index (HPI) was below the critical pollution index value of 100. Statistical analyses showed that potential sources of heavy metals are land-based, thereby implying that former tin mining and industries in the surrounding area are the most likely sources. Anthropogenic metal concentrations were found to be low in the Selangor River, indicating that it has yet to be burdened by pollution of heavy metals.

This study aimed at evaluating the three advanced satellite precipitation products (SPPs), i.e. CMORPH, TRMM 3B42V7 and PERSIANN, against the ground observation to evaluate their performance in detecting rain, capturing storms and rainfall pattern during 2014–2015 extreme flood events at three different river basins in Peninsular Malaysia (Kelantan, Langat and Johor river basins). Several spatial interpolation methods, including Arithmetic Mean, Thiessen Polygon, Inverse Distance Weighting, Ordinary Kriging and Spline were applied on the ground observations to transform the point-based precipitation into areal precipitation. Slight variations in the interpolated values were found, but overall it was comparable. Based on the daily rainfall data for the duration of 62 days, this study found that all SPPs performed with acceptable accuracy, as shown by the Kelantan river basin; however, these SPPs did not estimate accurately for Langat and Johor river basins. Overall, TRMM and CMORPH outperformed PERSIANN for the Langat and Johor river basins. In conclusion, all SPPs were capable of predicting heavy rainfall during the northeast monsoon and the level of accuracy is promising for the northern part of Peninsular Malaysia. However, as for the rest of the region, careful consideration should be given when applying the SPPs.

The main focus of this study is to find a better approach for the improvement of rainfall intensity–duration–frequency (IDF) for Peninsular Malaysia by fitting the best statistical model to each data series. This is because IDF curve is generally used to determine the design rainfall in designing water resource structures. Data are collected from 60 selected rainfall stations in Peninsular Malaysia based on their completeness of data and coverage area. Two models are chosen and compared to derive the IDF relationship, which are fitting PDS to generalized Pareto (GPA) distribution (GPA/PDS) and fitting AMS to generalized extreme value (GEV) distribution (GEV/AMS). Based on the L-moment ratio diagram, it is found that the model based on fitting PDS to GPA distribution is more suitable in derivation of rainfall IDF relationship in Peninsular Malaysia. A software package containing Rain Event Mining Tool (RainEMT), Rain IDF and RainMap is developed to aid the study.

The production and use of energy from hydropower generation play a vital role in the economy. Besides, the presence of uncertainty further increases the complexity in optimizing the reservoir operation. A synthetic streamflow generation based on historical inflow records was employed using the Thomas–Fiering model for handling the uncertainty and variability of reservoir inflows. However, under the circumstances of water deficiency, the hydropower output is significantly reduced. In this study, an investigation of a parameter free Jaya algorithm as an optimization method for reservoir operation was carried out. When deriving the optimal operational rule a hedging strategy is introduced to attenuate the impact of reduced water supply. This strategy can effectively counterbalance the lack of water supply with reservoir storage requirements. The higher amount of hydropower generated by the proposed algorithm than the other algorithms used in this study, such as genetic algorithm (GA), the ant colony algorithm (ACO), the bat algorithm (BA), the particle swarm optimization (PSO) algorithm, chicken swarm optimization (CSO) algorithm, grasshopper optimization algorithm (GOA), equilibrium optimizer (EO) and firefly algorithm (FA), has shown its efficiency in the reservoir system. Several reservoir performance indices, such as total hydropower generation, reliability, and resilience, were used to access the proposed algorithm and other algorithms efficiency

The minimum inter-event time (MIT) between two wet periods has to be set in order to separate rainfall events. The objective of this study is to analyse the effect of MIT on rainfall events in Peninsular Malaysia. Rainfall event data from 16 rainfall stations, containing the annual number of rainfall events together with the number of events that exceed a certain range of duration values, have been extracted using RainEMT. It is found that the average total number of rainfall events will be higher with smaller MIT values.

In this study the data collected by 16 automatic rainfall stations distributed uniformly across Peninsular Malaysia are analysed using a Rainfall Event Mining Tool (RainEMT) developed by the authors in Microsoft Access, using Visual Basic for Applications. RainEMT allows the easy extraction of large-scale rainfall event data such as the total number of yearly or monthly rainfall events, annual maximum rainfall or annual maxima series, and event total rainfall, with threshold selection from a large time series rainfall dataset. The rainfall events are separated by a user-defined minimum inter-event time (MIT).

A new method for obtaining the C factor (i.e., vegetation cover and management factor) of the RUSLE model is proposed. The method focuses on the derivation of the C factor based on the vegetation density to obtain a more reliable erosion prediction. Soil erosion that occurs on the hillslope along the highway is one of the major problems in Malaysia, which is exposed to a relatively high amount of annual rainfall due to the two different monsoon seasons. As vegetation cover is one of the important factors in the RUSLE model, a new method that accounts for a vegetation density is proposed in this study. A hillslope near the Guthrie Corridor Expressway (GCE), Malaysia, is chosen as an experimental site whereby eight square plots with the size of $$8\times 8$$ and $$5\times 5$$ m are set up. A vegetation density available on these plots is measured by analyzing the taken image followed by linking the C factor with the measured vegetation density using several established formulas. Finally, erosion prediction is computed based on the RUSLE model in the Geographical Information System (GIS) platform. The C factor obtained by the proposed method is compared with that of the soil erosion guideline Malaysia, thereby predicted erosion is determined by both the C values. Result shows that the C value from the proposed method varies from 0.0162 to 0.125, which is lower compared to the C value from the soil erosion guideline, i.e., 0.8. Meanwhile predicted erosion computed from the proposed C value is between 0.410 and $$3.925\, \hbox {t ha}^{-1 }\,\hbox {yr}^{-1}$$ compared to 9.367 to $$34.496\, \hbox {t ha}^{-1}\,\hbox {yr}^{-1 }$$ range based on the C value of 0.8. It can be concluded that the proposed method of obtaining a reasonable C value is acceptable as the computed predicted erosion is found to be classified as a very low zone, i.e. less than $$10\, \hbox {t ha}^{-1 }\,\hbox {yr}^{-1}$$ whereas the predicted erosion based on the guideline has classified the study area as a low zone of erosion, i.e., between 10 and $$50\, \hbox {t ha}^{-1 }\,\hbox {yr}^{-1}$$ .

This study explores capability of adaptive neuro-fuzzy interface system (ANFIS) model for soil erosion assessment of which various parameters are involved and the parameters interaction is highly nonlinear. Soft computing technique has been applied widely in various fields, but its application in soil science is limited. A number of models have been established in predicting soil erosion such as a revised universal soil loss equation (RUSLE) model whereby parameters of the RUSLE model are employed in testing the ANFIS model in this study. Besides, a new parameter of biomass microbe (Mf) is introduced and its effectiveness in controlling erosion is demonstrated. A hilly area near the Guthrie Corridor Expressway, Malaysia, is chosen as a study area whereby five experimental plots size of 8 × 8 m and 5 × 5 m are developed. These five plots are characterized by either natural or planted vegetation or either with or without microbe bio-fertilizer. Input parameters of RUSLE, i.e. rainfall erosivity, soil erodibility, slope length and steepness, vegetation and support practice, are measured, and observed erosion is monitored for 1 year. The microbe factor presented as Mf is examined by pairing the plot with and without microbe while keeping other similar characteristic for those plots. Result shows a satisfactory achievement between observed and predicted model by ANFIS with R2, RMSE and MAE being 0.8275, 1.4276 and − 0.0165, respectively. In terms of a relationship between the Mf factor and soil erosion, it is noticed that the soil erosion is inversely proportional to the Mf when the entire erosion conservation practices are unchanged for a given period. Therefore, the application of microbes could be an alternative or additional erosion conservation technique in soil conservation works.

This article investigates the potential for using machine learning algorithms to delineate stratiform/convective (S/C) rain regimes for passive microwave imager taking calibrated brightness temperatures as only spectral parameters. The algorithms have been implemented for the Tropical Rainfall Measuring Mission (TRMM) microwave imager (TMI), and calibrated as well as validated taking the Precipitation Radar (PR) S/C information as the target class variables. Two different algorithms are particularly explored for the delineation. The first one is metaheuristic adaptive boosting algorithm that includes the real, gentle, and modest versions of the AdaBoost. The second one is the classical linear discriminant analysis that includes the Fisher's and penalized versions of the linear discriminant analysis. Furthermore, prior to the development of the delineation algorithms, a feature selection analysis has been conducted for a total of 85 features, which contains the combinations of brightness temperatures from 10 GHz to 85 GHz and some derived indexes, such as scattering index, polarization corrected temperature, and polarization difference with the help of mutual information aided minimal redundancy maximal relevance criterion (mRMR). It has been found that the polarization corrected temperature at 85 GHz and the features derived from the “addition” operator associated with the 85 GHz channels have good statistical dependency to the S/C target class variables. Further, it has been shown how the mRMR feature selection technique helps to reduce the number of features without deteriorating the results when applying through the machine learning algorithms. The proposed scheme is able to delineate the S/C rain regimes with reasonable accuracy. Based on the statistical validation experience from the validation period, the Matthews correlation coefficients are in the range of ~0.60–0.70. Since, the proposed method does not rely on any a priori information, this makes it very suitable for other microwave sensors having similar channels to the TMI. The method could possibly benefit the constellation sensors in the Global Precipitation Measurement (GPM) mission era.

Prediction of the groundwater dynamics via models can help better manage the groundwater resources and guarantee their sustainable use. Three types of data-driven models are built for groundwater depth prediction in the plain of Shijiazhuang, the capital of Hebei Province in North China. The data-driven models include the Power Function Model (PFM), Back-Propagation Artificial Neural Network (BPANN) and Support Vector Machines (SVM) with two kernel functions of linear kernel function (LKF) and radial basis function (RBF). Five classes of factors (including 12 indices) are considered as potential model input variables. The Gamma Test (GT) is adopted in this study to help identify the relative importance of the input indices and tackle the tricky issue of the optimal input combinations for the data-driven models. The established models are evaluated in both fitting and testing procedures based on the root mean squared error (RMSE) and Nash-Sutcliffe efficiency (E) for different input combination schemes. The results show that SVM (RBF) performs the best. It is interesting to find that the natural factors (i.e., precipitation and evaporation) are less relevant to the groundwater depth variations. The methods used in this study have much significance for groundwater depth prediction in areas lacking hydrogeological data.

Soil moisture acts as a key variable in the exchanges of water and energy from land surface to atmosphere. The work presented here showed the first-time comprehensive evaluation of land surface temperature (LST) derived from moderate resolution imaging spectroradiometer (MODIS) and soil moisture from soil moisture and ocean salinity (SMOS) for prediction hydrological soil moisture deficit (SMD). The main focus of the work relies on the application of adaptive neuro fuzzy inference system (ANFIS) for evaluating its capabilities towards SMD estimation. The performances between estimated and ANFIS simulated SMD are assessed in terms of Nash-Sutcliffe Efficiency, root mean square error, and the percentage of bias. The performance statistics revealed a good agreement between benchmark and ANFIS estimated SMD from MODIS as compared to the SMOS. This present study has strong supportive evidence that there are SMOS soil moisture and MODIS LST products for different hydrometeorological studies. The assessment of the product with respect to this peculiar evidence is an important step for successful development of hydrometeorological model and forecasting system.

Wan Jaafar, Wan Zurina, and Dawei Han, 2012. Calibration Catchment Selection for Flood Regionalization Modeling. Journal of the American Water Resources Association (JAWRA) 48(4): 698-706. DOI: 10.1111/j.1752-1688.2012.00648.x Abstract: There are two unsolved problems in flood regionalization model development related to the quantity and quality of calibration catchments: (1) how many calibration catchments should be used? and (2) how to select the calibration catchments? This study explores these two questions through a case study on the median annual maximum flood (QMED) model in the United Kingdom. It has been found that the chance of developing a good QMED model decreases significantly when the number of calibration catchments drops below a critical number (e.g., 60 in the case study). However, no significant improvement is achieved if the number of calibration catchments is above it. This number could be used as a benchmark for choosing randomly selected calibration catchments. Across a broad range of calibration catchment numbers, there are good and poor calibrated models regardless of calibration catchment numbers. High quality models could be developed from a small number of calibration catchments and also poor models from a large number of calibration catchments. This indicates that the number of calibration catchments may not be the dominating factor for developing a high quality regionalization model. Instead, the information content could be more important. The study has demonstrated that the standard deviation values between the best and poorest groups are distinctive and could be used in choosing appropriate calibration catchments.

Flood frequency regionalisation for ungauged catchments depends on relevant catchment characteristics. In this study, alternative catchment characteristics for south-west England were derived using the data freely downloadable from the internet. It was found that the alternative catchment characteristics were more heterogeneous than the Flood Estimation Handbook (FEH) catchment descriptors. Therefore, they suffered less in information redundancy and had a potential to produce more reliable regression models. The results have demonstrated that the model from the alternative catchment characteristics had a better performance than the FEH model and FEH local model. It is also useful to note that a recalibrated FEH local model had a poorer performance than the original FEH model. This indicates that a national model should not simply be recalibrated using a regional data set. The methodology and catchment characteristics used in the present study are more applicable to other parts of the world than the catchment descriptors in the FEH due to the easier data access used in the present study.

Regional frequency analysis is an important tool in estimating design flood for ungauged catchments. Index flood is an important component in regionalized flood formulas. In the past, many formulas have been developed based on various numbers of calibration catchments (e.g. from less than 20 to several hundred). However, there is a lack of systematic research on the model uncertainties caused by the number of calibration catchments (i.e. what is the minimum number of calibration catchment? and how should we choose the calibration catchments?). This study uses the statistical resampling technique to explore the impact of calibration catchment numbers on the index flood estimation. The study is based on 182 catchments in England and an index flood formula has been developed using the input variable selection technique in the data mining field. The formula has been used to explore the model uncertainty due to a range of calibration catchment numbers (from 15 to 130). It is found that (1) as expected, the more catchments are used in the calibration, the more reliable of the models developed are (i.e. with a narrower band of uncertainty); (2) however, poor models are still possible with a large number of calibration catchments (e.g. 130). In contrast, good models with a small number of calibration catchments are also achievable (with as low as 15 calibration catchments). This indicates that the number of calibration catchments is only one of the factors influencing the model performance. The hydrological community should explore why a smaller calibration data set could produce a better model than a large calibration data set. It is clear from this study that the information content in the calibration data set is equally if not more important than the number of calibration data. Copyright © 2011 John Wiley & Sons, Ltd.

Vegetation cover and management factor (C) of the Universal Soil Loss Equation (USLE) is usually can be found from guideline. However, the value would actually bias for small scale area. This study was therefore aimed to use digital photograph to obtain the vegetation cover and management factor (C) at small scale. A total of 8 photographs were taken at the centre of the 8 experimental plots location. These photographs were processed using ArcGIS 10.1 whereby the vegetation cover and bared soil were identified using the maximum likelihood algorithm method. For validation purposes, the results obtained were compared with that of the previous studies. We found that the newly derived value of C is of comparable with that of the previous studies. This indicates the simple technique is beneficial for quantification of the vegetation ground cover for small scale and sloping topography as we applied to hilly area beside highway. The proposed technique is also applicable for large area but it is suggested to use unmanned aerial vehicle (UAV) in order to obtain overall picture of the area.

Although the nonlinear power form model structure is widely accepted by practitioners in the flood regionalization modelling, there is a lack of studies on whether there is a room for further improvement, and if the answer is yes, what should be done to explore alternative model structures. A framework is proposed in this study towards investigating this issue by the following steps: (i) a universal data-driven model is utilized to see if there is a room for improvement compared with the conventional model, and (ii) if improvement is achieved, this means that there should exist more effective model structures than the current form. However, because the universal data-driven models are usually opaque, more explicit model structures should be explored, which are convenient for practical usage. In this study, the proposed framework is applied in a case study using the catchment characteristics from the Flood Estimation Handbook in conjunction with the gamma test, support vector machine (SVM) and genetic programming (GP). First, the gamma test is used for the purpose of input variables selection where no model structure needs to be defined as a priori, and therefore, the result can be applied to any model structures for model building. Second, an SVM, which is a powerful data-driven nonlinear model capable of modelling a variety of nonlinear systems, is applied to the index flood model for the first time. Once the best model is determined using those two data-driven tools, GP is employed to find an alternative model structure. As the SVM is not formulated for producing explicit model functional form, the GP offers an advantage at this point where it can infer an explicit mathematical model functional form. Copyright © 2012 John Wiley & Sons, Ltd.