Document Type : Research Article
Authors
1
Msc. Student in GIS, Department of Civil Engineering, Faculty of Technical and Engineering, Ferdowsi University of Mashhad,Iran
2
Civil Engineering Department, Engineering Faculty, Ferdowsi University of Mashhad (FUM)
Abstract
One of the factors of air pollution is the phenomenon of dust, which has caused a lot of damage to various economic, social and human resources. The phenomenon of dust occurs in parts of the world, including arid and semi-arid regions, which is caused by natural and human factors. This research has identified the origin of dust in Ilam city using spatio-temporal tensor of aerosol optical depth (AOD) with Madis sensor data in the period from March to June 2022. First, dusty days were extracted from meteorological data and the spatio-temporal tensor of aerosol optical depth was produced. The reason for using the tensor was to examine the changes of a large volume of data in a spatial and temporal manner in a study period simultaneously. The results of comparing the relevant tensor with the corresponding meteorological data showed that whenever the aerosol optical depth is higher than 0.5, there is dust in that range. The spatio-temporal dust tensor analysis showed that the amount of dust is directly related to the wind speed and when the wind speed exceeds 15 m/s, dust occurs. Finally, by identifying the spatial changes of AOD, there are four sources of dust (Beld, Mesopotamia, Misan, and Wasit) in the study area, and Mesopotamia was identified as one of the potential dust areas. The analysis of the time pattern of AOD indicates its increasing trend in May. The highest value of AOD with 3.85 in May indicates the amount of dust. By examining the correlation between Ilam dust and the identified centers, the regression model of Ilam city is more related to Wasit region and its correlation coefficient is 82.96%.
Introduction
Dust storms are among the most significant atmospheric hazards in arid and semi-arid regions, often causing substantial environmental and socio-economic damage. This phenomenon is particularly prevalent along the global dust belt, stretching from the western coasts of North Africa to the Middle East. Dust particles suspended in the atmosphere may originate from natural sources such as soil erosion and volcanic activity, as well as anthropogenic pollutants. In western Iran, especially in Ilam Province, dust storms primarily originate from neighboring Iraq. Due to the region’s climatic conditions and lack of adequate infrastructure, Ilam is highly vulnerable to frequent and intense dust storm events. Recent research has demonstrated the growing application of remote sensing data and geospatial information systems (GIS) for monitoring and analyzing dust phenomena across Iran. In this context, several studies have focused on the spatial and temporal analysis of dust events in various parts of the country. The present study specifically investigates the spatiotemporal variability of dust storms in Ilam County. Utilizing MODIS satellite imagery and tensor analysis tools within the Google Earth Engine (GEE) platform, this research aims to assess dust distribution patterns over time and space. The findings help identify primary dust sources and characterize the temporal evolution and spatial extent of dust activity in the region, resulting in the development of detailed spatiotemporal dust distribution maps for Ilam. This section outlines the study area, datasets, proposed methods, and data preprocessing steps undertaken for the analysis. Ilam County is located in the northwest of Ilam Province, Iran. It is bordered by Eyvan County to the north, Sirvan to the east, Mehran to the south, and Iraq to the west. The region features a temperate mountainous climate, with an average annual precipitation of approximately 619 millimeters and absolute temperature extremes ranging from -13°C to 41°C. Geographically, Ilam County holds strategic significance due to its shared border with Iraq, a primary contributor to dust generation in the Middle East. Given the transboundary nature of dust transport in this region, the spatial extent of the study area was expanded to include zones adjacent to the Iraqi border. This allowed for a more comprehensive analysis of the spatiotemporal distribution of dust storms affecting Ilam.
Material and Methods
Satellite Data and Processing Workflow
This study utilizes satellite data acquired from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor, which is mounted on NASA's Terra and Aqua satellites for global atmospheric monitoring. MODIS features 36 spectral bands with varying spatial resolutions. For the retrieval of dust-related parameters, the Deep Blue (DB) algorithm was employed to extract the Aerosol Optical Depth (AOD) index, which quantifies the concentration of airborne particulate matter. The primary dataset used in this research is the MCD19A2 daily AOD product, originally provided at a spatial resolution of 1 kilometer. For analytical purposes, the data were resampled to a 2-kilometer resolution to optimize computational efficiency. Dust storm days were identified and extracted within the Google Earth Engine (GEE) platform, and further processing and visualization were conducted using MATLAB software.
Google Earth Engine
Google Earth Engine (GEE) is a web-based cloud processing platform launched by Google in 2010, providing free and rapid access to remote sensing data. As one of the largest platforms for spatial data analysis, GEE is widely used in climate studies, meteorology, and remote sensing applications. Its use leads to significant time, cost, and human resource savings, as well as enhanced accuracy in studies due to access to up-to-date and comprehensive data repositories.
Tensor Analysis
To simultaneously analyze spatial and temporal phenomena, multidimensional structures are required, which are provided by tensor analysis. In this study, a third-order (three-dimensional) tensor was employed in the MATLAB environment to model and analyze the data based on geographic latitude, longitude, and time. Unlike traditional statistical methods, the use of tensors enables the examination of complex spatiotemporal phenomena with greater precision and differentiation. Dust storms are not only a nuisance but also have profound environmental and human consequences. Identifying the origins of this phenomenon, particularly in the border regions of western Iran, such as Ilam County, is of critical importance. The majority of these dust storms originate from the deserts of Iraq, Saudi Arabia, and Syria, where poor water management practices and hydraulic construction projects have contributed to land degradation, making it more susceptible to dust production. In this study, we used remote sensing technology and the Google Earth Engine platform to investigate daily dust storm variations during the spring of 2022. Unlike previous studies that primarily focused on periodic or seasonal analyses, we employed daily MODIS satellite data in conjunction with tensor analysis to provide a more precise and visual representation of dust storm events. The tensor-based plots for March revealed two significant dust waves on the 4th and 6th of March, originating from the Wasit region of Iraq, which impacted Ilam County and the Maysan region. The intensity of the first wave was such that horizontal visibility in Ilam dropped to just 800 meters. The second wave, although less intense, was still notably impactful. In April, three distinct dust waves affected Ilam, originating from the Balad and Mesopotamian regions. These storms occasionally reduced horizontal visibility to as low as 500 meters. However, the most severe dust events occurred in May, with five consecutive waves originating from the Al Anbar, Wasit, and Maysan regions. On May 23, visibility in Ilam dropped to zero, marking the worst conditions during the study period. June also saw two waves originating from Balad and Maysan, with visibility in Ilam reduced to 2000 meters. The analysis showed that May recorded the highest Aerosol Optical Depth (AOD) values, indicating the greatest dust density.
Validation of Results
To validate our findings, the AOD data were cross-verified with ground station measurements from the source regions and Ilam. Specifically, a strong correlation was found between AOD variations in the Wasit region of Iraq and Ilam County. Regression modeling confirmed this correlation, indicating that whenever dust storms occur in the Wasit region, Ilam County is likely to be affected as well.
Conclusions
This study, utilizing spatiotemporal tensor analysis, successfully identified the origin and daily pattern of dust storms in Ilam County from March to June 2022. The data used in this research were extracted from the MODIS sensor (MCD19A2 product), and dust storm days were selected based on reports from the Meteorological Organization. A key advantage of applying tensor analysis in this study is its ability to simultaneously represent spatial and temporal variations of dust storms in a three-dimensional space, a feature not available in traditional two-dimensional methods, such as satellite imagery. This approach allowed for a precise examination of the dust storm origins on a daily basis, marking an innovative step in atmospheric monitoring. The results of the analysis revealed that the dust storm situation in the study area was critical. Four major regions were identified as the primary sources of dust: Balad, Mesopotamia, Maysan, and Wasit. Among these sources, the Mesopotamian region played a dominant role, particularly in May, with an AOD index of 3.85, indicating the highest frequency and intensity of dust storms. Furthermore, correlation matrix analysis showed the most direct and significant relationship between dust storms in Ilam County and the Wasit region of Iraq. This study demonstrated that tensor analysis is a powerful tool for monitoring and forecasting dust storm behavior, offering valuable support for environmental policymakers and disaster management authorities in making informed, scientific decisions. It is recommended that future studies incorporate additional variables such as temperature, vegetation cover, wind patterns, and climate change to develop more comprehensive models for dust storm monitoring.
Keywords
Main Subjects
)
Send comment about this article