Stations In Peninsular Malaysia Research Articles

A Framework to Spatially Cluster Air Quality Monitoring Stations in Peninsular Malaysia using the Hybrid Clustering Method

Multiple variables must be analyzed in order to assess air quality trends. It turns into a multidimensional issue that calls for dynamic methods. In order to provide an improved spatial cluster distribution with distinct validation, this study set out to illustrate the hybrid cluster method in air quality monitoring stations in Peninsular Malaysia. The Department of Environment, Malaysia (DOE), provided the data set, which covered the two-year period from 2018 to 2019. This study included six air quality pollutants: PM10, PM2.5, SO2, NO2, O3, and CO. Principal component analysis (PCA), a multivariate technique, was used to condense the information found in enormous data tables in order to better comprehend the variables (to reduce dimensionality) prior to grouping the data. The PCA factor scores were then used to produce the AHC. The clusters were validated using discriminant analysis (DA). 36 of 47 stations required additional analysis using AHC, according to the PCA factor scores. Low Polluted Region (LPR = seven stations), Moderate Polluted Region (MPR = 20 stations), and High Polluted Region (HPR = nine stations) were created from AHC and share the same characteristics. The DA results showed 84 % correct classification rate for the clusters. With regard to identifying and categorizing stations according to air quality characteristics, the framework presented here offers an improved method. This illustrates that the hybrid cluster method utilized in this work can produce a new method of pollutant distributions that is helpful in air pollution investigations.

Tweedie models for Malaysia rainfall simulations with seasonal variabilities

Abstract This study aims to evaluate the suitability of the Tweedie generalised linear model for characterising monthly rainfall patterns across 18 meteorological stations in Peninsular Malaysia. It incorporates harmonic functions consisting of sine and cosine functions as seasonal predictors and El Niño Southern Oscillation (ENSO) indices as climatic predictors. Results indicate that three harmonic functions are essential to accurately portray rainfall dynamics in the southwestern and northwestern regions, while two suffice for the inland and western regions. However, incorporating four harmonic functions is the most optimal representation of the eastern region. An additional 1-month lag in ENSO indices is introduced to the optimal seasonal predictor model. Based on the findings, the southern oscillation index notably impacts monthly rainfall significantly in eastern and inland areas, while meteorological stations in the western and northwestern areas fit better with the multivariate ENSO index. Strikingly, no substantial impact of climate predictors is observed on the monthly rainfall within the southwestern region. Thus, the influence of climate indices is very much influenced by the geographical locations of the regions. Importantly, generating simulated data through the Tweedie model contributes to a more accurate representation of the statistical properties inherent in rainfall analysis.

Analysis of trends in extreme temperature indices over selected areas in Peninsular Malaysia from 1991-2020

This study investigates the temporal variations of extreme temperature trends across three stations in Peninsular Malaysia with different settings, Petaling Jaya represents urban station, while Kluang and Batu Embun represent rural areas from 1991-2020. Eleven extreme warm and cold indices trends were analysed using the Mann-Kendall test, with Sen’s slope as a magnitude estimator. The findings revealed that the annual daily mean temperature, daily mean maximum temperature, and daily mean minimum temperature increased significantly at all stations by 0.04°C yr–1, 0.02°C yr–1, and 0.05°C yr–1, respectively. The station in the rural area shows a significant increment of an average of 0.03°C yr–1 for mean and warm extreme indices. Despite that, there were no changes in the total number of warm days and nights and cold days and nights for all settings. For the cold temperature indices, there were significant decreasing trends for the annual minimum of maximum temperature in rural areas, while urban station did not exhibit changes. All stations also exhibit similar annual patterns when compared although significantly higher readings were observed during El Niño events.

Exploring rainfall variabilities using statistical functional data analysis

Functional data analysis (FDA) has been widely applied in various scientific fields, including climatological, hydrological, environmental, and biomedical. The flexibility of the FDA in incorporating temporal elements into the statistical analysis makes the method highly demanded compared to the conventional statistical approach. This study introduces FDA methods to investigate the variations and patterns of rainfall throughout Peninsular Malaysia, which includes 16 rain gauge stations in Peninsular Malaysia from 1999 to 2019. A descriptive statistic of the functional data depicted the mean and variation of the rainfall curve over time, while the functional principal component analysis measured the temporal variability of the rainfall curve. According to the findings, the first and second principal components accounted for 87.4% of all variations. The first principal component was highly characterised by the stations over the eastern region during the northeast monsoon since the highest variability was observed from November to January. On the other hand, the stations impacted by the inter-monsoon season were best described by the second principal component. Based on the factor scores derived from the functional principal component, those rain gauge stations with comparable features were then clustered. Overall, the results showed that the rainfall pattern is strongly influenced by their geographical and topographical features and the seasonal monsoon effect.

Validation of Hershfield probable maximum precipitation estimation using homogeneous region in Malaysia

Probable maximum precipitation (PMP) is used to design major hydraulic structures, such as dams and spillways, flood protection works, and nuclear power plants. Long observation data are required for PMP estimates using the Hershfield statistical approach. In Malaysia, the PMP calculated using state boundary is commonly used, but the estimated PMP values are close to historical maximums. Homogeneous regions are regions containing stations with similar rainfall characteristics. Therefore, homogeneous region boundaries could be used for PMP estimations. In this paper, 1-day rainfall data of more than 25 years for 161 stations in Peninsular Malaysia; were analysed to estimate the PMP for 1-day duration based on an appropriate frequency factor by considering homogeneous regions. Using PMP estimates, a generalised graph was prepared to show the spatial distribution of 1-day PMP. The 1-day PMP estimates in Peninsular Malaysia by considering state boundary, varied from 146 to 1364.7 mm and by considering the homogeneous regions, the 1-day PMP estimates varied from 208.5 to 2094 mm. Hence, the PMP considering homogeneous regions is considered more accurate and safe to be used in designs.

Statistical Analysis of Rainfall Data in Peninsular Malaysia using Strong Transformation of Cubic Spline Analysis

The constant rainfall is important for equatorial regions. The extreme rainfall shows our nature is facing climate change significantly and this needs to be paid attention by researchers hence the effects of disasters can be avoided as early as possible. Therefore, it is important to understand the rain pattern for the existing data so that forecasting can be done. This article discussed how to transform abnormal data into normal data and then analyze the data using the spline cubic analysis method. The study found that by using the transformation algorithm, the best r value is obtained. Cubic spline analysis shows that there are 30 rainfall stations in peninsular Malaysia where there are significant changes in rainfall patterns.

DIFFERENCES BETWEEN 2009 AND 2016 REVISIONS BASED ON COORDINATES IN GDM2000

The Geocentric Datum of Malaysia (GDM200) is realised with respect to International Terrestrial Reference Frame (ITRF) 2000 at epoch 2nd January 2000. In comparison with the 2000 frame, ITRF2014 has significant improvement in terms of its definition and realisation. Moreover, several great earthquakes that struck the Indonesian region for the past decades have deformed the tectonic plate, resulting in a shifted GDM2000. These earthquakes, followed by post-seismic activities, has caused GDM2000 to become obsolete. Following that, the Department of Survey and Mapping Malaysia (DSMM) has taken the initiative to revise the coordinate of Malaysia Real-Time Kinematic Global Navigation Satellite Systems (GNSS) Network (MyRTKnet) stations in GDM2000 into a new set of coordinates. Therefore, this paper presents an effort to analyse the differences between coordinates in GDM2000 based on 2009 and 2016 revisions. In order to measure the discrepancy, forty-seven (47) MyRTKnet stations in Peninsular Malaysia were chosen to estimate the differences between the two (2) revisions. The coordinates obtained from MyRTKnet stations were then projected into Rectified Skewed Orthomorphic (RSO) coordinate system to compute the differences in horizontal position and ellipsoidal height. The finding showed that the discrepancy ranges from 0.8 to 11.8 cm, with the smallest values at SETI station and the biggest value at KRAI station. Meanwhile, for the differences in ellipsoidal height, LIPI station has the biggest value of 8.1 cm, followed by the smallest value of 0.4 cm at SETI station. In conclusion, as the differences in revision gave impact on the changes of coordinates of MyRTKnet stations in Peninsular Malaysia, the frequent revision of GDM2000 should also consider the latest frame to give better positional accuracy, and a proper datum transformation (ITRF2014 to ITRF2000) need to be implemented for mapping purposes.

Performance of potential evapotranspiration models in Peninsular Malaysia

Abstract Potential evapotranspiration (PET) is an important parameter for the operation of irrigation projects and water resources management. The globally recognized PET estimation model, the FAO-56 Penman–Monteith (FAO-56 PM) model, had been criticized for its requirement of many detailed meteorological variables, but nevertheless has been accepted as the baseline model in many worldwide studies. The performances of different PET models can be found to be excellent for a specific location but may not be representative in other regions. The aim of this study is to select the most suitable PET model to estimate PET in Malaysia. Three radiation-based models and four temperature-based models were compared with the FAO-56 PM model at seven selected meteorological stations in Peninsular Malaysia. The mean bias error, relative error (Re) and normalized root-mean-square error (NRMSE) and coefficient of determination (R2) were used to evaluate the performances of the PET models. The Re values of Turc models were below 0.2 at all stations, while Priestly–Taylor, Thornthwaite, Thornthwaite-corrected and Blaney–Criddle models were above 0.2. The Makkink and Hargreaves–Samani models were below 0.2 at most of the stations. Thus, the Turc model was recommended as the best model to estimate PET in Peninsular Malaysia.

Stationary and Non-Stationary Models of Extreme Ground-Level Ozone in Peninsular Malaysia

High ground-level ozone (GLO) concentrations will adversely affect human health, vegetations as well as the ecosystem. Therefore, continuous monitoring for GLO trends is a good practice to address issues related to air quality based on high concentrations of GLO. The purpose of this study is to introduce stationary and non-stationary model of extreme GLO. The method is applied to 25 selected stations in Peninsular Malaysia. The maximum daily GLO concentration data over 8 hours from year 2000 to 2016 are used. The factors of this distribution are anticipated using maximum likelihood estimation. A comparison between stationary (constant model) and non-stationary (linear and cyclic model) is performed using the likelihood ratio test (LRT). The LRT is based on the larger value of deviance statistics compared to a chi-square distribution providing the significance evidence to non-stationary model either there is linear trend or cyclic trend. The best fit model between selected models is tested by Akaike's Information Criterion. The results show that 25 stations conform to the non-stationary model either linear or cyclic model, with 14 stations showing significant improvement over the linear model in location parameter while 11 stations follow the cyclic model. This study is important to identify the trends of ozone phenomenon for better quality risk management.

Statistical modelling of extreme rainfall in Peninsular Malaysia

Flash floods are known as one of the common natural disasters that cost over billions of Ringgit Malaysia throughout history. Academically, an extreme rainfall model is effective in modelling to predict and prevent the occurrence of flash floods. This paper compares four probability distributions, namely, exponential distribution, generalized extreme value distribution, gamma distribution, and Weibull distribution, with the rainfall data of 10 stations in peninsular Malaysia. The period of the data is from 1975 to 2008. The comparison is based on the descriptive and predictive analytics of the models. The determination of the most effective model is through Kolmogorov-Smirnov, Anderson-Darling, and chi-square test. The result shows that generalized extreme value is the most preferred extreme rainfall model for the rainfall cases in Peninsular Malaysia.

Modeling the spatio-temporal dynamics of air pollution index based on spatial Markov chain model.

An environmental problem which is of concern across the globe nowadays is air pollution. The extent of air pollution is often studied based on data on the observed level of air pollution. Although the analysis of air pollution data that is available in the literature is numerous, studies on the dynamics of air pollution with the allowance for spatial interaction effects through the use of the Markov chain model are very limited. Accordingly, this study aims to explore the potential impact of spatial dependence over time and space on the distribution of air pollution based on the spatial Markov chain (SMC) model using the longitudinal air pollution index (API) data. This SMC model is pertinent to be applied since the daily data of API from 2012 to 2014 that have been gathered from 37 different air quality stations in Peninsular Malaysia is found to exhibit the property of spatial autocorrelation. Based on the spatial transition probability matrices found from the SMC model, specific characteristics of air pollution are studied in the regional context. These characteristics are the long-run proportion and the mean first passage time for each state of air pollution. It is found that the probability for a particular station's state to remain good is 0.814 if its neighbors are in a good state of air pollution and 0.7082 if its neighbors are in a moderate state. For a particular station having neighbors in a good state of air pollution, the proportion of time for it to continue being in a good state is 0.6. This proportion reduces to 0.4, 0.01, and 0 for the cell of moderate, unhealthy, and very unhealthy states, respectively. In addition, there exists a significant spatial dependence of API, indicating that air pollution for a particular station is dependent on the states of the neighboring stations.

Existence of Long Memory Phenomenon in Air Pollutant Concentrations using Surrogate Data

This study investigated the existence of long memory phenomenon in air pollutant concentrations specifically the ozone concentration obtained at six monitoring stations in Peninsular Malaysia. The main objective was to select the best method in detecting the long memory. Four methods used in this study were Rescaled Range (R/S) method, Aggregated Variance (V/S) method, Aggregated Absolute Value (A/S) method and Peng’s (P/S) method. Surrogate data testing was used to verify the existence of long memory. The average estimated Hurst value obtained by using VS and PS method are found to be near to the actual Hurst value compared to RS and AS method. The biased and MSE value also showed that the VS and PS method is the most appropriate method in estimating the H value. Based on the result obtained, it can be concluded that long memory exists in ozone concentration data used in the study. The VS and PS method are the best method in detecting the long memory phenomenon.

The used of the Boosted Regression Tree Optimization Technique to Analyse an Air Pollution data.

The stochastic boosted regression trees (BRT) technique has the capability to quantify and explain the relationships between explanatory variables. We applied this machine learning modelling technique to derive the relationships between the gases air pollutants, meteorological conditions and time system variables of particulate matter (PM10) concentrations. In order to get lowest prediction error and to avoid over-fitting, the parameters of the BRT model need to be tuned. In this experiment, 25 BRT models were generated from 14 years’ worth of hourly data (122,736 a one hour averaged data from January 2000 to December 2013 gathered from four Continuous Automated Air Quality Monitoring Stations in peninsular Malaysia (located in Klang, Selangor (CA0011), Perai, Penang (CA0003), Kota Bharu, Kelantan (CA0022) and Kemaman, Terengganu (CA0002)). Seventy percent of the data were used for training and 30 percent for validation of the models. An experiment was conducted to determine the best iteration that could model hourly PM10 concentrations by optimizing the BRT parameter which are learning rate (lr), tree complexity (tc) and number of trees (nt). Five different lr (0.001, 0.005, 0.01, 0.05 and 0.1) were tested with different tree complexities (1 to 20) in the BRT model development process. From the experiment, the combination of lr = 0.05 and tc = 5 for the training set for the BRT model achieved the lowest root mean squared error (RMSE) compared to the other tested combinations. It was also found that the number of trees increased with the increment in the number of samples. A high coefficient of determinant (R2) value (0.90) for the linear relationship between the number of samples and nt was found for all the four stations. The optimum number of trees for the model was estimated by using 10-fold cross-validation. It was found that the best number of iterations for Klang, Perai, Kota Bahru and Kemaman were 12,327, 32,987, 16,370 and 57,634, respectively. The prediction accuracy of the model was tested by using the fraction of prediction namely a factor of two (FAC2), mean bias, mean gross error, RMSE, correlation coefficient (R), and index of agreement (IOA). The prediction performance of the final BRT model based on the R value was 0.81, 0.78, 0.85 and 0.81 for for Perai, Kemaman, Klang and Kota Bahru, respectively, which indicates that the BRT model developed and applicability of this can be used in other atmospheric environment data.

Markov chain modeling for air pollution index based on maximum a posteriori method

Air pollution is a major environmental problem, which brings about a threat to human health and the natural environment. Thus, determination and assessment of the level of air pollution is an important component in monitoring of the air quality. This study involves estimating the transition probability matrix of the Markov chain model based on maximum a posteriori (MAP) method using hourly data of air pollution index (API). The API data has been collected from seven air-monitoring stations in peninsular Malaysia. The estimated transition probability matrix is used to determine the characteristics of air pollution such as the steady-state probability and the expected return time. The transition probabilities of the Markov chain are fitted based on maximum a posteriori method under three different priors, which are Dirichlet, Jeffreys, and uniform. The results found show that the maximum a posteriori method under the Dirichlet prior produced the most precise estimates as compared with the other priors. In addition, for the areas considered in the study, the moderate state is more persistent as opposed to unhealthy states indicating that the problem of air quality is not very serious. In general, this study could provide important implications for developing proper strategies for managing air quality and for improving public health.

TEMPORAL RAINFALL DISAGGREGATION BY A SIMPLE RANDOM CASCADE MODEL

This study evaluates the utility and suitability of a simple discrete multiplicative random cascade model for temporal rainfall disaggregation. Two of a simple random cascade model, namely log-Poisson and log-Normal models are applied to simulate hourly rainfall from daily rainfall at seven rain gauge stations in Peninsular Malaysia. The cascade models are evaluated based on the capability to simulate data that preserve three important properties of observed rainfall: rainfall variability, intermittency and extreme events. The results show that both cascade models are able to simulate reasonably well the commonly used statistical measures for rainfall variability (e.g. mean and standard deviation) of hourly rainfall. With respect to rainfall intermittency, even though both models are underestimated, the observed dry proportion, log-Normal model is likely to simulate number of dry spells better than log-Poisson model. In terms of rainfall extremes, it is demonstrated that log-Poisson and log-Normal models gave a satisfactory performance for most of the studied stations herein, except for Dungun and Kuala Krai stations, which both located in the east part of Peninsula.

Surfactants in the sea surface microlayer, subsurface water and fine marine aerosols in different background coastal areas

This study aims to determine the concentrations of surfactants in the surface microlayer (SML), subsurface water (SSW) and fine mode aerosol (diameter size < 1.5μm) at different coastal stations in Peninsular Malaysia. The concentrations of anionic and cationic surfactants were determined through colorimetric methods as methylene blue active substances (MBAS) and disulphine blue active substances (DBAS), respectively. Water-soluble ions, for the determination of fine mode aerosol sources, were determined using ion chromatography (IC) for anions (SO42-, NO3-, Cl- and F-) and cations (Na+, K+, Ca2+ and Mg2+). Principal component analysis (PCA), combined with multiple linear regression (MLR), was used to identify the possible sources of surfactants in fine aerosol. The results showed the concentrations of surfactants as MBAS and DBAS in the SML ranged between 0.23 ± 0.03 and 0.35 ± 0.01μmolL-1 and between 0.21 ± 0.02 and 0.29 ± 0.01μmolL-1, respectively. The enrichment factors (Efs) ratios between MBAS and DBAS in the SML and SSW ranged between 1.04 ± 0.01 and 1.32 ± 0.04, respectively. The station that is located near to tourism and industrial activities recorded the highest concentrations of surfactants in SML and SSW. The concentrations of surfactants in fine aerosol ranged between 62.29 and 106.57pmolm-3. The three possible sources of fine aerosol during the northeast monsoon were aged sea spray/biomass burning (which accounted for 69% of the atmospheric aerosol), nitrate/mineral dust (23%) and sulphate/fresh sea salt (8%). During the southwest monsoon, the three main sources of atmospheric aerosol were biomass burning (71%), secondary inorganic aerosol (23%) and sea spray (6%). This study suggests anthropogenic sources are main contributors to the concentrations of surfactants in SML, SSW and fine aerosols.

Bayesian autoregressive spatiotemporal model of PM10 concentrations across Peninsular Malaysia

Rapid industrialization and haze episodes in Malaysia ensure pollution remains a public health challenge. Atmospheric pollutants such as PM10 are typically variable in space and time. The increased vigilance of policy makers in monitoring pollutant levels has led to vast amounts of spatiotemporal data available for modelling and inference. The aim of this study is to model and predict the spatiotemporal daily PM10 levels across Peninsular Malaysia. A hierarchical autoregressive spatiotemporal model is applied to daily PM10 concentration levels from thirty-four monitoring stations in Peninsular Malaysia during January to December 2011. The model set in a three stage Bayesian hierarchical structure comprises data, process and parameter levels. The posterior estimates suggest moderate spatial correlation with effective range 157 km and a short term persistence of PM10 in atmosphere with temporal correlation parameter 0.78. Spatial predictions and temporal forecasts of the PM10 concentrations follow from the posterior and predictive distributions of the model parameters. Spatial predictions at the hold-out sites and one-step ahead PM10 forecasts are obtained. The predictions and forecasts are validated by computing the RMSE, MAE, R2 and MASE. For the spatial predictions and temporal forecasting, our results indicate a reasonable RMSE of 10.71 and 7.56, respectively for the spatiotemporal model compared to RMSE of 15.18 and 12.96, respectively from a simple linear regression model. Furthermore, the coverage probability of the 95% forecast intervals is 92.4% implying reasonable forecast results. We also present prediction maps of the one-step ahead forecasts for selected day at fine spatial scale.

Comparison of Clustering Algorithms on Air Quality Substances in Peninsular Malaysia

Air quality is one of the most popular environmental problems in this globalization era. Air pollution is the poisonous air that comes from car emissions, smog, open burning, chemicals from factories and other particles and gases. This harmful air can give adverse effects to human health and the environment. In order to provide information which areas are better for the residents in Malaysia, cluster analysis is used to determine the areas that can be clustering together based on their a ir quality through several air quality substances. Monthly data from 37 monitoring stations in Peninsular Malaysia from the year 2013 to 2015 were used in this study. K - Means (KM) clustering algorithm, Expectation Maximization (EM) clustering algorithm and Density Based (DB) clustering algorithm have been chosen as the techniques to analyze the cluster analysis by utilizing the Waikato Environment for Knowledge Analysis (WEKA) tools. Results show that K - means clustering algorithm is the best method among ot her algorithms due to its simplicity and time taken to build the model. The output of K - means clustering algorithm shows that it can cluster the area into two clusters, namely as cluster 0 and cluster 1. Clusters 0 consist of 16 monitoring stations and clu ster 1 consists of 36 monitoring stations in Peninsular Malaysia.

Modelling non-stationary extreme streamflow in Peninsular Malaysia

Global change has raised concerns among hydrologists about the use of the stationary assumption (independent and identically distributed flood series) in infrastructure-designed methods. This confirms the necessity of evaluating the stationary or non-stationary behaviour of hydrological variables before deriving flood plans for infrastructure projects and flood mitigation. Trends were evaluated in the annual maximal streamflow of 49 stations in Peninsular Malaysia, using the Mann Kendall and Spearman Rho trend tests. Three models, a stationary model (GEV0) and two non-stationary models, with: 1) location parameter; 2) location and log-transformed scale parameters as a linear function of time (GEV2), were considered for stations with significant trend. It was found that a quarter of the analysed stations show statistically significant trends in their annual maximal streamflow. These results indicate the importance of taking into consideration the non-stationary behaviour of the flood series in order to improve the quality of flood estimation.

Statistical Analysis Towards Improvement of Design Rainfall: Case Study of Peninsular Malaysia

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.