Estimating the susceptibility of gases emission on land cover and air quality from cement plant factory using Gaussian Plumes Model and GIS

Emissions of gases from cement plants are considered to be environmental pollutants. It is significant to detect and monitor such emissions in the appropriate manner. These pollutants represent significant risks and consequences to human health, air quality, the growth of plantations, and the development of urbanised areas. Various studies have been conducted to identify the pollutants that are released into the air by industrial sources and to assess their influence on air quality and land cover. One of the widely adopted method is Gaussian Plume Model for measuring the concentration of pollution from point sources. This model; however, has several significant limitations. Among these is its reliance on a single wind speed value, disregarding the spatial variations across the study area. The second issue is that the model although incorporates the Digital Elevation Model (DEM) to account for ground height variations, it fails to consider the impact of urban structures, forests, and complex terrain. Furthermore, the model assums linear pollutant transimission neglecting the dynamic of wind patterns influenced by various atmospheric conditions. Moreover, the model does not take into account the effects of temperature and humidity on pollutant dispersion. Given these limitations, coupled with the environmental challenges posed by cement factory emissions and inadequate regulatory frameworks in the study area, there is a pressing need to develop a more comprehensive and efficient predictive model. This research first extracted pollution concentration using Gaussian Plume Model implemented in QGIS considering parameters such as stability class based on wind speed values and solar radiation and terrain characteristics. Then, the study estimated pollutant concentration using the same model by incorporating the DEM of the study area and cross section of plume pathways by means of three-dimensional Gaussian Plume. The study further assessed the impact of temperature and specific humidity on pollution dispersion. Finally, the study developed a pollutant prediction model using deep learning methods. The spline interpolation was used to generate wind speed layers for four seasons 2020 from 11 NASA Meteorological stations in the study area. The land cover classes were extracted by applying the maximum likelihood (ML) classification on Landsat images. An Analytic Hierarchy Process (AHP) was used to do a comparison of the primary and secondary directions in terms of their potential exposure to cement plant emissions in the year 2020. The wind speeds across the four seasons were between 3.07 and 4.35 meters per second. Sand, often known as barren ground, is the most frequent type of land, accounting for 75.75 percent of the area that was surveyed, followed by vegetation cover (13.35%) and urban areas (7.97%). The quantity of water is the smallest, making up only 4.67% of the total study area. Both the overall accuracy and the Kappa coefficient of ML were calculated to be 0.9736 and 98.2143%, respectively. According to the pollution risk assesmenr, four levels of severity were assigned to the effects of pollution risk: very high, high, medium, and low. The spring season records the maximum value of very high risk pollution, which ranges from 52.428 to 1264.332 g/m3 , while the winter season of 2020 records the lowest value of very high risk pollution, which ranges from 0 to 0.017 g/m3 . During the summer, 8.573 km2 of urban areas had the highest levels of pollution. During the summer, 5 km2 of plantation land contained the areas with the highest amounts of pollution. There were a total of 60,974 km2 of polluted sand during the summer. During the summer, there were 2.667 km2 of contaminated regions in water bodies. The results of 3D Gaussian Plume model indicated that the simulated pollutant concentration varied significantly by season, with higher concentrations observed in spring and summer, and lower concentrations observed in autumn and winter. The results were achieved via implementing the model to two important places surrounding the study area: Kirkuk City and the Laylan District. Pollutant concentrations in Kirkuk City are higher in autumn and winter than in spring and summer. As one gets farther away from the source, the concentration of pollutants rises progressively until it peaks at 500 meters. Currently, the maximum concentrations of pollutants are found in spring, summer, fall, and winter, with values ranging from 11.69 to 87.76 μg/m3 . With a maximum value of 87.65 μg/m3 at 500 meters from the source, the data for Laylan District showed that the greatest concentration levels were recorded during the winter. The maximum concentration of the pollutants increased with temperature adjustment and decreased with specific humidity correction, according to the results of the Gaussian Plume Model with temperature and specific humidity corrections. The concentrations of the pollutants increased in the fall and winter after the temperature and specific humidity correction parameters were added. They peaked at 168.28 μg/m3 and increased to 1677.84 μg/m3 and 563.027 μg/m3 for the fall season and 108.19 μg/m3 for the winter, which increased to 1418.7 μg/m3 after temperature correction and 611.282 μg/m3 after specific humidity correction. Results also showed that the model is effective in both the Allahabad and Laylan stations, with closer results for the data in the Laylan station, which is based on the Gaussian equation simulated data. Loss functions in Laylan range from 0.0221 for the CaO parameter to 0.0041 for the Fe2O3 and MgO parameters, which corresponds to 0.038 for Xylene at the Allahabad station, and vice versa. The highest loss function value in Laylan is 4.466 for the AQI parameter.

Text FK 2024 60 - Full Text.pdf Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (6MB) Text FK 2024 60.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (7MB)

Actions (login required)

View Item