Introduction
Sand and dust storms mainly occur in arid and semi-arid areas of the world. One of the main causes of air pollution in the desert provinces of the country, especially Yazd province, is the occurrence of dust storms containing high concentrations of suspended particles with an aerodynamic diameter of less than 10 μm (PM10) which their severity and frequency have increased over the past and had adverse effects on the health of the people. The meteorological parameters such as surface winds speed, relative humidity, temperature and rainfall are the most important factors affecting the emission rate of these particles. Accordingly, the main objectives of this study are to investigate the trends of PM10 concentration during the six-year statistical period (2012-2017), to identify the most important meteorological parameters affecting the concentration of these particles and to determine its relationship with the occurrence of wind erosion based on Dust Storm Index (DSI).
Materials and Methods
In this study, meteorological data related to the prevailing wind speed, maximum wind speed, relative humidity, average monthly temperature, maximum temperature, minimum temperature and precipitation values of Yazd synoptic station and daily amounts of suspended particulate matter were less than 10 μm obtained from relevant organizations. Then, the average PM10 concentrations were calculated at the monthly, seasonally and annually time scales, and the trend of temporal changes were investigated. In order to determine the kind of relationship between meteorological parameters and PM10 concentration, 10 functions of bivariate regression functions (linear, logarithmic, inverse, power, exponential, quadratic, Cubic, growth curves, compound, logistic and curve S) were used. Using the four criteria of correlation coefficient, F-Statistic, error estimation and root-mean-square error were determined the most accurate function and the most important factor influencing the trend of temporal changes of PM10 concentrations. Finally, the dust storm index was calculated and its correlation with the mean values of PM10 was determined.
Discussion and Results
The highest concentrations of particulate matter were in April (145.57 µg/m3), spring (120.77 µg/m3) and 2015 (129.83 µg/m3). There was no significant linear relationship between the studied meteorological parameters and the concentrations of suspended particles (P value> 0.05), but there were significant nonlinear relationships between the mean monthly temperature and maximum temperature with the concentrations of suspended particles in the study area (P value<0.05). The most accurate nonlinear relationship, correlation coefficient, F statistic, estimation error, and mean square error in establishing regression relations between monthly average temperature and suspended particle concentration were curve S, 0.67, 0.86, 0.1, 0.09 and in the relationship between the maximum temperature and the suspended particle concentration, the curves S, 0.65, 7.3, 0.11and 0.08, respectively. According to the obtained determination coefficient by the bivariate regression analysis between PM10 and DSI, 24% of the annual changes in particle concentration were due to increased activity of sand and dust events over the study period, the most of which was caused by changes of variables in summer season (38%). These results may indicate the intensification of desertification due to the wind erosion phenomenon in summer and especially in recent years. According to the results of this study, the main cause of these changes was the average temperature fluctuations of Yazd city during these years. Given that the most important source of these particles is the Yazd-Ardakan plain which is located in the northwest of the study area, it is expected that the bulk of these particles originated from this critical center.
Conclusions
The highest concentrations of particulate matter were in April (145.57 µg/m3), spring (120.77 µg/m3) and 2015 (129.83 µg/m3). There was no significant linear relationship between the studied meteorological parameters and the concentrations of suspended particles (P value> 0.05), but there were significant nonlinear relationships between the mean monthly temperature and maximum temperature with the concentrations of suspended particles in the study area (P value<0.05). The most accurate nonlinear relationship, correlation coefficient, F statistic, estimation error, and mean square error in establishing regression relations between monthly average temperature and suspended particle concentration were curve S, 0.67, 0.86, 0.1, 0.09 and in the relationship between the maximum temperature and the suspended particle concentration, the curves S, 0.65, 7.3, 0.11and 0.08, respectively. According to the obtained determination coefficient by the bivariate regression analysis between PM10 and DSI, 24% of the annual changes in particle concentration were due to increased activity of sand and dust events over the study period, the most of which was caused by changes of variables in summer season (38%). These results may indicate the intensification of desertification due to the wind erosion phenomenon in summer and especially in recent years. According to the results of this study, the main cause of these changes was the average temperature fluctuations of Yazd city during these years. Given that the most important source of these particles is the Yazd-Ardakan plain which is located in the northwest of the study area, it is expected that the bulk of these particles originated from this critical center.