Environmental Engineering
M.I. Rumaling; F.P. Chee; H.W.J. Chang; C.M. Payus; S.K. Kong; J. Dayou; J. Sentian
Abstract
BACKGROUND AND OBJECTIVES: Air quality in some developing countries is dominated by particulate matter, especially those with size 10 micrometers and smaller or PM10. They can be inhaled and sometimes can get deep into lungs; some may even get into bloodstream and cause serious health problems. Therefore, ...
Read More
BACKGROUND AND OBJECTIVES: Air quality in some developing countries is dominated by particulate matter, especially those with size 10 micrometers and smaller or PM10. They can be inhaled and sometimes can get deep into lungs; some may even get into bloodstream and cause serious health problems. Therefore, future PM10 concentration forecasting is important for early prevention and in urban development planning, which is crucial for developing cities. This paper presents the development of PM10 forecasting model using nonlinear autoregressive with exogenous input model.METHODS: To improve performance of nonlinear autoregressive with exogenous input model, principal component analysis is used prior to the model for variable selection. The first stage of principal component analysis involves Scree plot, which determines the number of principal components based on explained variance. This is then followed by selecting variables using a rotated component matrix, based on their strength of contribution towards variation of PM10 concentration. To test the model, PM10 data in Kota Kinabalu from 2003 – 2010 was used. Neural network models are developed using this data by varying number of input variables with the inclusion of temporal variables. The developed forecasting models are evaluated using data PM10 in the city from 2011 to 2012. Four performance indicators, namely root mean square error, mean absolute error, index of agreement and fractional bias are reported.FINDINGS: Results from principal component analysis show that five variables including wind direction index, relative humidity, ambient temperature, concentration of nitrogen dioxide and concentration of ozone strongly contribute to the variation of PM10 concentration. By using these variables together with temporal variables as input in the nonlinear autoregressive with exogenous input models, the resultant model shows good forecasting performance, with root mean square error of 7.086±0.873 µg/m3. The selection of significant variables helps in reducing input variables inside the forecast model without degrading its forecast performance.CONCLUSION: This model shows very promising performance in forecasting PM10 concentration in Kota Kinabalu as it requires fewer input variables and does not require variable transformation.
Environmental Management
C. Loukrakpam; B. Oinam
Abstract
BACKGROUND AND OBJECTIVE: Soil erosion is considered one of the major indicators of soil degradation in our environment. Extensive soil erosion process leads to erosion of nutrients in the topsoil and decreases in fertility and hence productivity. Moreover, creeping erosion leads to landslides in the ...
Read More
BACKGROUND AND OBJECTIVE: Soil erosion is considered one of the major indicators of soil degradation in our environment. Extensive soil erosion process leads to erosion of nutrients in the topsoil and decreases in fertility and hence productivity. Moreover, creeping erosion leads to landslides in the hilly regions of the study area that affects the socio-economics of the inhabitants. The current study focuses on the estimation of soil erosion rate for the year 2011 to 2019 and projection for the years 2021, 2023 and 2025.METHODS: In this study, the Revised Universal Soil Loss Equation is used for estimation of soil erosion in the study area for the year 2011 to 2019. Using Artificial Neural Network-based Cellular Automata simulation, the Land Use Land Cover is projected for the future years 2021, 2023 and 2025. Using the projected layer as one of the spatial variables and applying the same model, Soil Erosion based on Revised Universal soil loss equation is projected for a corresponding years.FINDINGS: For both cases of projection, simulated layers of 2019 (land use land cover and soil erosion) are correlated with the estimated layer of 2019 using actual variables and validated. The agreement and accuracy of the model used in the case land use are 0.92 and 96.21% for the year 2019. The coefficient of determination of the model for both simulations is also observed to be 0.875 and 0.838. The simulated future soil erosion rate ranges from minimum of 0 t/ha/y to maximum of 524.271 t/ha/y, 1160.212 t/ha/y and 783.135 t/ha/y in the year 2021, 2023 and 2025, respectively.CONCLUSION: The study has emphasized the use of artificial neural network-based Cellular automata model for simulation of land use and land cover and subsequently estimation of soil erosion rate. With the simulation of future soil erosion rate, the study describes the trend in the erosion rate from past to future, passing through present scenario. With the scarcity of data, the methodology is found to be accurate and reliable for the region under study.
Environmental Management
S.K. Tamang; P.D. Singh; B. Datta
Abstract
Artificial neural network is considered one of the most efficient methods in processing huge data sets that can be analyzed computationally to reveal patterns, trends, prediction, forecasting etc. It has a great prospective in engineering as well as in medical applications. The present work employs artificial ...
Read More
Artificial neural network is considered one of the most efficient methods in processing huge data sets that can be analyzed computationally to reveal patterns, trends, prediction, forecasting etc. It has a great prospective in engineering as well as in medical applications. The present work employs artificial neural network-based curve fitting techniques in prediction and forecasting of the Covid-19 number of rising cases and death cases in India, USA, France, and UK, considering the progressive trends of China and South Korea. In this paper, three cases are considered to analyze the outbreak of Covid-19 pandemic viz., (i) forecasting as per the present trend of rising cases of different countries (ii) forecasting of one week following up with the improvement trends as per China and South Korea, and (iii) forecasting if followed up the progressive trends as per China and South Korea before a week. The results have shown that ANN can efficiently forecast the future cases of COVID 19 outbreak of any country. The study shows that the confirmed cases of India, USA, France and UK could be about 50,000 to 1,60,000, 12,00,000 to 17,00,000, 1,40,000 to 1,50,000 and 2,40,000 to 2,50,000 respectively and may take about 2 to 10 months based on progressive trends of China and South Korea. Similarly, the death toll for these countries just before controlling could be about 1600 to 4000 for India, 1,35,000 to 1,00,000 for USA, 40,000 to 55,000 for France, 35,000 to 47,000 for UK during the same period of study.
M. Memarianfard; A.M. Hatami; M. Memarianfard
Abstract
Most parts of the urban areas are faced with the problem of floating fine particulate matter. Therefore, it is crucial to estimate the amounts of fine particulate matter concentrations through the urban atmosphere. In this research, an artificial neural network technique was utilized to model the PM2.5 ...
Read More
Most parts of the urban areas are faced with the problem of floating fine particulate matter. Therefore, it is crucial to estimate the amounts of fine particulate matter concentrations through the urban atmosphere. In this research, an artificial neural network technique was utilized to model the PM2.5 dispersion in Tehran City. Factors which are influencing the predicted value consist of weather-related and air pollution-related data, i.e. wind speed, humidity, temperature, SO2, CO, NO2, and PM2.5 as target values. These factors have been considered in 19 measuring stations (zones) over urban area across Tehran City during four years, from March 2011 to March 2015. The results indicate that the network with hidden layer including six neurons at training epoch 113, has the best performance with the lowest error value (MSE=0.049438) on considering PM2.5 concentrations across metropolitan areas in Tehran. Furthermore, the “R” value for regression analysis of training, validation, test, and all data are 0.65898, 0.6419, 0.54027, and 0.62331, respectively. This study also represents the artificial neural networks have satisfactory implemented for resolving complex patterns in the field of air pollution.