Loading...

Development and Assessment of a Numerical Model for Investigation of Regional Dust Distribution

Najafpour Mollabashi, Nategheh | 2021

195 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 54557 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Afshin, Hossein; Firoozabadi, Bahar
  7. Abstract:
  8. In recent years, dust storms have been recognized as one of the most crucial problems for people worldwide. Provided that the dust masses transport mechanisms are properly known, the main dust sources can be identified, and therefore, appropriate actions can be taken. Therefore, the main goals of the present study are estimating dust concentration, determination of dust distribution with an emphasis on Iran (especially Tehran), and finding the main dust sources which will be followed in five steps. In the first step, by analyzing the measured data during dust observations in Tehran through 2013-2016, several simple and multivariate nonlinear regression models have been established in order to estimate PM10 dust concentration in terms of horizontal visibility, aerosol optical depth (AOD), and meteorological parameters. While the results of the first step make a quick tool for dust concentration calculation, it also provides a comprehensive vision of dust distribution in different parts of the country. In the second step, in order to better understanding dust distribution, one of the severe dust events in Tehran (February 20-22, 2016) is selected and the flow field parameters are predicted using the BSC-DREAM8b numerical model. Also, in this step, the effects of dust masses on meteorological conditions and atmospheric stability are investigated by physical analysis of the results. The results indicate that on February 19-20, a low-pressure system is formed in the central regions of Iran, which led to atmospheric instability, increased the wind speed, and consequently, increased the planetary boundary layer height to 5 km. As well, this low-pressure system caused the dust mass created in the eastern regions of Syria and central Iraq to enter the central regions of Iran, including Tehran, from the southwestern border cities such as Ahvaz. By scrutinizing the results of radiative forcing, it is found that due to the absorption and scattering of radiation flux by particles, the atmosphere is stabilized, the boundary layer height is declined and the 2m temperature is decreased when the dust concentration is maximized. Afterwards, in the third step, a sensitivity analysis is performed to find the effect of various parameters on the accuracy of numerical modeling. To this end, using an open source numerical model (WRF-Chem), different relations and algorithms governing dust distribution are applied through different schemes embedded in the model and the results are compared. Generally, the results show that using GOCART and AFWA dust emission schemes, MYJ and YSU boundary layer schemes, and Noah land surface model, the numerical results have a good correlation with the observations. In the fourth step, the nested grid method is adopted in order to reduce the computational costs and also increase the possibility of obtaining a more appropriate depiction of topography and dust sources in the desired areas. Comparison of the uniform and nested grid results shows that the predicted dust concentration trend is closer to the observations using nested grid. Considering that it is shown by the sensitivity analysis that the dust emission scheme (determining the dust sources) is the most important scheme in the dust modeling, the fifth step has been organized based on the improvement of this scheme. Accordingly, an attempt is made to increase the accuracy of numerical results, especially in Tehran, by considering experiences gained from the previous steps, using nested grid, and also implementation of a newly developed code in the numerical model in order to calculate the dust flux. It is found that the modified numerical model improves the difference between peak dust concentration calculated by numerical modeling with that obtained by observations, the standard deviation ratio, and NRMSE by 22.2 %, 0.33, and 0.18, respectively. In this step, by focusing on identification of local and external dust sources, it is found that the largest part of particles originates from the external sources located in Iraq and eastern Syria. Also, it can be said that deserts locating in the eastern and southeastern areas around Tehran are the local dust sources for Tehran region. In the studied dust event, 95.1 Tg of dust particles are lifted from all areas, of which about 1.3 Tg are related to Iran. This value is equivalent to 1.4 % of the total lifted particles, while the ratio of the area of Iran to the total area of the computational domain is 3.3 %
  9. Keywords:
  10. Numerical Modeling ; Sensitivity Analysis ; Regression Models ; Nested Lattices ; Regression Models ; Dust Concentration

 Digital Object List

 Bookmark

No TOC