Comparison of hybrid machine learning methods for the prediction of short-term meteorological droughts of Sakarya Meteorological Station in Turkey

Drought is a harmful natural disaster with various negative effects on many aspects of life. In this research, short-term meteorological droughts were predicted with hybrid machine learning models using monthly precipitation data (1960–2020 period) of Sakarya Meteorological Station, located in the northwest of Turkey. Standardized precipitation index (SPI), depending only on precipitation data, was used as the drought index, and 1-, 3-, and 6-month time scales for short-term droughts were considered. In the prediction models, drought index was predicted at t?+?1 output variable by using t, t???1, t???2, and t???3 input variables. Artificial neural networks (ANNs), adaptive neuro-fuzzy inference system (ANFIS), Gaussian process regression (GPR), support vector machine regression (SVMR), k-nearest neighbors (KNN) algorithms were employed as stand-alone machine learning methods. Variation mode decomposition (VMD), discrete wavelet transform (DWT), and empirical mode decomposition (EMD) were utilized as pre-processing techniques to create hybrid models. Six different performance criteria were used to assess model performance. The hybrid models used together with the pre-processing techniques were found to be more successful than the stand-alone models. Hybrid VMD-GPR model yielded the best results (NSE?=?0.9345, OI?=?0.9438, R²?=?0.9367) for 1-month time scale, hybrid VMD-GPR model (NSE?=?0.9528, OI?=?0.9559, R²?=?0.9565) for 3-month time scale, and hybrid DWT-ANN model (NSE?=?0.9398, OI?=?0.9483, R²?=?0.9450) for 6-month time scale. Considering the entire performance criteria, it was determined that the decomposition success of VMD was higher than DWT and EMD.

     

A review of recent developments in the application of machine learning in solar thermal collector modelling

Environmental Science and Pollution Research - Over the past few decades, the popularity of solar thermal collectors has increased dramatically because of many significant advantages like being a...     

Road grade estimation for on-road vehicle emissions modeling using light detection and ranging data

Vehicle-specific power (VSP) is useful for explaining a substantial portion of variability in real-world vehicle emissions, such as those measured with portable emissions monitoring systems (PEMS). VSP is a function of vehicle speed, acceleration, and road grade. Road grade is shown to significantly affect estimates of both VSP and of real-world emissions via sensitivity analysis and analysis of empirical data. However, road grade is difficult to measure reliably using PEMS. Therefore, alternative methods for estimating road grade were identified and compared. A preferred method for estimating road grade was explored in more detail based on light detection and ranging (LIDAR) data. The method includes buffering LIDAR data onto roadway maps using a geographic information system tool, defining segments of roadway based on criteria pertaining to vertical curvature, quantification of roadway elevations within the buffered segments, and estimation of road grade and banking by fitting a plane to each segment. Factors influencing errors in road grade estimates are discussed. The method was evaluated by application to selected interstate highways and comparison to design drawing data. The development and application of LIDAR-based road grade data are demonstrated via a case study using PEMS data collected in the Research Triangle Park, NC, area. LIDAR data are shown to be reliable and accurate for road grade estimation for vehicle emissions modeling.     

A review about COVID-19 in the MENA region: environmental concerns and machine learning applications

Environmental Science and Pollution Research - Coronavirus disease 2019 (COVID-19) has delayed global economic growth, which has affected the economic life globally. On the one hand, numerous...     

Agricultural Irrigation Recommendation and Alert (AIRA) system using optimization and machine learning in Hadoop for sustainable agriculture

Environmental Science and Pollution Research - Internet of Things (IoT) in the field of agriculture promises to continuously provide global access to the farming information. The smart agriculture...     

Inland harmful cyanobacterial bloom prediction in the eutrophic Tri An Reservoir using satellite band ratio and machine learning approaches

Environmental Science and Pollution Research - In recent years, Tri An, a drinking water reservoir for millions of people in southern Vietnam, has been affected by harmful cyanobacterial blooms...     

Pollution,economic growth,and COVID-19 deaths in India: a machine learning evidence

Environmental Science and Pollution Research - This study uses two different approaches to explore the relationship between pollution emissions, economic growth, and COVID-19 deaths in India. Using...     

GIS-based air quality modelling: spatial prediction of PM10 for Selangor State,Malaysia using machine learning algorithms

Environmental Science and Pollution Research - Rapid urbanization has caused severe deterioration of air quality globally, leading to increased hospitalization and premature deaths. Therefore,...     

In silico prediction of Tetrahymena pyriformis toxicity for diverse industrial chemicals with substructure pattern recognition and machine learning methods

There is an increasing need for the rapid safety assessment of chemicals by both industries and regulatory agencies throughout the world. In silico techniques are practical alternatives in the environmental hazard assessment. It is especially true to address the persistence, bioaccumulative and toxicity potentials of organic chemicals. Tetrahymena pyriformis toxicity is often used as a toxic endpoint. In this study, 1571 diverse unique chemicals were collected from the literature and composed of the largest diverse data set for T. pyriformis toxicity. Classification predictive models of T. pyriformis toxicity were developed by substructure pattern recognition and different machine learning methods, including support vector machine (SVM), C4.5 decision tree, k-nearest neighbors and random forest. The results of a 5-fold cross-validation showed that the SVM method performed better than other algorithms. The overall predictive accuracies of the SVM classification model with radial basis functions kernel was 92.2% for the 5-fold cross-validation and 92.6% for the external validation set, respectively. Furthermore, several representative substructure patterns for characterizing T. pyriformis toxicity were also identified via the information gain analysis methods.     

Application of ensemble machine learning model in downscaling and projecting climate variables over different climate regions in Iran

Environmental Science and Pollution Research - This study evaluates the future climate fluctuations in Iran’s eight major climate regions (G1–G8). Synoptic data for the period...     

Influencing factors of carbon emissions and their trends in China and India: a machine learning method

China and India are the largest coal consumers and the most populated countries in the world. With industrial and population growth, the need for energy has increased, which has inevitably led to an increase in carbon dioxide (CO₂) emissions because both countries depend on fossil fuel consumption. This paper investigates the impact of energy consumption, financial development (FD), gross domestic product (GDP), population, and renewable energy on CO₂ emissions. The study applies the long short-term memory (LSTM) method, a novel machine learning (ML) approach, to examine which influencing driver has the greatest and smallest impact on CO₂ emissions; correspondingly, this study builds a model for CO₂ emission reduction. Data collected between 1990 and 2014 were analyzed, and the results indicated that energy consumption had the greatest effect and renewable energy had the smallest impact on CO₂ emissions in both countries. Subsequently, we increased the renewable energy coefficient by one and decreased the energy consumption coefficient by one while keeping all other factors constant, and the results predicted with the LSTM model confirmed the significant reduction in CO₂ emissions. Finally, this study forecasted a CO₂ emission trend, with a slowdown predicted in China by 2022; however, CO₂ emission’s reduction is not possible in India until 2023. These results suggest that shifting from nonrenewable to renewable sources and lowering coal consumption can reduce CO₂ emissions without harming economic development.

     

Spatio-temporal classification and prediction of land use and land cover change for the Vembanad Lake system,Kerala: a machine learning approach

Environmental Science and Pollution Research - Land use and land cover (LULC) change has become a critical issue for decision planners and conservationists due to inappropriate growth and its...     

Interpretable machine learning approach to analyze the effects of landscape and meteorological factors on mosquito occurrences in Seoul,South Korea

Environmental Science and Pollution Research - Mosquitoes are the underlying cause of various public health and economic problems. In this study, patterns of mosquito occurrence were analyzed based...     

Development of new computational machine learning models for longitudinal dispersion coefficient determination: case study of natural streams,United States

Environmental Science and Pollution Research - Natural streams longitudinal dispersion coefficient (Kx) is an essential indicator for pollutants transport and its determination is very important....     

Estimation and uncertainty analysis of groundwater quality parameters in a coastal aquifer under seawater intrusion: a comparative study of deep learning and classic machine learning methods

Environmental Science and Pollution Research - Excessive withdrawal of groundwater for agricultural irrigation can cause seawater intrusion into coastal aquifers. Such a case will in turn results...     

Hybrid-based Bayesian algorithm and hydrologic indices for flash flood vulnerability assessment in coastal regions: machine learning,risk prediction,and environmental impact

Environmental Science and Pollution Research - Natural hazards and severe weather events are a matter of serious threat to humans, economic activities, and the environment. Flash floods are one of...     

Evolution of vehicle exhaust particles in the atmosphere

Aerosol mass spectrometer (AMS) measurements are used to characterize the evolution of exhaust particulate matter (PM) properties near and downwind of vehicle sources. The AMS provides time-resolved chemically speciated mass loadings and mass-weighted size distributions of nonrefractory PM smaller than 1 microm (NRPM1). Source measurements of aircraft PM show that black carbon particles inhibit nucleation by serving as condensation sinks for the volatile and semi-volatile exhaust gases. Real-world source measurements of ground vehicle PM are obtained by deploying an AMS aboard a mobile laboratory. Characteristic features of the exhaust PM chemical composition and size distribution are discussed. PM mass and number concentrations are used with above-background gas-phase carbon dioxide (CO2) concentrations to calculate on-road emission factors for individual vehicles. Highly variable ratios between particle number and mass concentrations are observed for individual vehicles. NRPM1 mass emission factors measured for on-road diesel vehicles are approximately 50% lower than those from dynamometer studies. Factor analysis of AMS data (FA-AMS) is applied for the first time to map variations in exhaust PM mass downwind of a highway. In this study, above-background vehicle PM concentrations are highest close to the highway and decrease by a factor of 2 by 200 m away from the highway. Comparison with the gas-phase CO2 concentrations indicates that these vehicle PM mass gradients are largely driven by dilution. Secondary aerosol species do not show a similar gradient in absolute mass concentrations; thus, their relative contribution to total ambient PM mass concentrations increases as a function of distance from the highway. FA-AMS of single particle and ensemble data at an urban receptor site shows that condensation of these secondary aerosol species onto vehicle exhaust particles results in spatial and temporal evolution of the size and composition of vehicle exhaust PM on urban and regional scales.     

Modeling of energy and emissions from animal manure using machine learning methods: the case of the Western Mediterranean Region,Turkey

Environmental Science and Pollution Research - In Turkey, facilities for the use of biomass resources in energy production are increasing, and new conversion facilities are commissioned every year...     

A new inventory for two-wheel vehicle emissions in West Africa for 2002

Rather surprisingly, urban atmospheric particulate levels in West Africa compare with measured concentrations in Europe and Asia megacities (Liousse, C., Galy-Lacaux, C., Assamoi, E.-M., Ndiaye, A., Diop, B., Cachier, H., Doumbia, T., Gueye, P., Yoboue, V., Lacaux, J.-P., Guinot, B., Guillaume, B., Rosset, R., Castera, P., Gardrat, E., Zouiten, C., Jambert, C., Diouf, A., Koita, O., Baeza, A., Annesi-Maesano, I., Didier, A., Audry, S., Konare, A., 2009. Integrated Focus on West African Cities (Cotonou, Bamako, Dakar, Ouagadougou, Abidjan, Niamey): Emissions, Air Quality and Health Impacts of Gases and Aerosols. Third International AMMA Conference on Predictability of the West African Moosoon Weather, Climate and Impacts. Ouagadougou, Burkina Faso. July 20–24). This pollution mainly derives from road traffic emissions with, in some capitals (e.g. Cotonou), the strong contribution of two-wheel vehicles. Two key questions arise: are presently available emission inventories (e.g. Junker, C., Liousse, C., 2008. A global emission inventory of carbonaceous aerosol from historic records of fossil fuel and biofuel consumption for the period 1860–1997. Atmospheric Chemistry Physics, 8, 1–13; Bond, T.C., Streets, D.G., Yarber, K.F., Nelson, S.M., Woo, J.H., Klimont, Z., 2004. A technology-based global inventory of black and organic carbon emissions from combustion. Journal of Geophysical Research, 1009, D14203, DOI:10.1029/2003JD003697) able to account for these emissions? And, if not, how can we remedy this? The aim of this paper is to develop a methodology to estimate emissions produced by two-wheel vehicles in West Africa for 2002 in a context where reliable information is hardly available. Fuel consumption ratios between two-wheel engines (in this work) and all vehicles issued from UN database (http://data.un.org/Data.aspx?d=EDATA&f=cmID%3aMO%3btrID%3a1221) are as high as 169%, 264% and 628%, for Burkina Faso, Mali and Chad respectively, indicating that this global database does not properly account for regional specificities. Moreover, emission factors for black carbon (BC) and primary organic carbon (OCp) have been measured for two-stroke engines in Benin (Guinot, B., Liousse, C., Cachier, H., Guillaume, B., et al. New emission factor estimates for biofuels and mobile sources. Atmospheric Environment, in press.), giving significantly higher values than in Europe. This is particularly true for OCp, and consequently the calculated emissions for two-stroke engines are also significantly larger than total road traffic previously estimated in global inventories (Junker and Liousse (2008) with United Nations database for 2002; Bond et al., 2004). The ensuing discussion illustrates the importance of two-stroke engines in the West Africa transport sector and the strong need for inventory updating.