Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

You searched for id:"oai:etd.ohiolink.edu:ucin1490350586489513". One record found.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters

1. Agharkar, Amal. Model Validation and Comparative Performance Evaluation of MOVES/CALINE4 and Generalized Additive Models for Near-Road Black Carbon Prediction.

Degree: MS, Engineering and Applied Science: Environmental Engineering, 2017, University of Cincinnati

Black carbon is formed owing to incomplete combustion of fossil fuels, biofuels, and biomass. Past research has concluded that mitigation of black carbon from on-road diesel vehicles has a potential to slow down global warming and climate change. Although black carbon is a component of a criteria pollutant (PM2.5) under National Ambient Air Quality Standards (NAAQS), mitigation of regional PM2.5 may or may not result in the abatement of black carbon. Hence, previous studies have underscored the necessity to study black carbon as an individual species of PM2.5 speciation. Limited availability of high-resolution black carbon data, traffic data and representative meteorological data for near road sites has resulted in inconclusive attempts to validate existing models to predict black carbon from on-road vehicles. In this study, we have presented a comprehensive and comparative model evaluation and validation of two different modeling philosophies. The first method was a traditional line source Gaussian dispersion model CALINE4, whereas the later philosophy employed a statistical generalized additive model (GAM). The predictions from these two systems were validated using black carbon data, traffic data and meteorological data. Both models were evaluated based on graphical screening techniques and thorough evaluation of descriptive statistics. CALINE4 exhibited R2 value of 0.6928, whereas it was found to be 0.9415 for GAM. Slope of linear regression line was 0.7341 for CALINE4 and 1.094 for GAM. Both models exhibited acceptable performance based on the model performance evaluation guidelines, which made use of fractional bias (FB), normalized mean square error (NMSE), Pearson’s correlation coefficient (R), and Factor-of-Two envelope (Fa2). Model sensitivity analysis indicated the similarities and differences in influential parameters for both the models. Meteorological variables such as wind speed, wind direction, atmospheric stability, and mixing height (5 m – 100 m) were sensitive variables for CALINE4, especially for near parallel winds. GAM exhibited strong sensitivity for meteorological variables such as humidity, temperature, mixing height, wind direction, and pressure. Wind speed was not found to be statistically significant for GAM, which can be associated with a large period of calm winds. Hence, comparative evaluation of GAM and dispersion models is recommended. T-tests were performed on predictions of CALINE4 and GAM and on-site monitored data. Both CALINE4 (p = 0.9482) and GAM (p = 0.9492) showed a good agreement with the measured data. The comprehensive evaluation and validation of the models concluded that CALINE4 and GAM can be used to predict roadside black carbon. Future studies should be undertaken to check applicability of GAM to other site locations using on-road traffic parameters such as vehicle speed, vehicle specific power, and lane numbers to understand pollutant dispersion in a more comprehensive way. Further comparative assessment studies involving validated dispersion models and statistical… Advisors/Committee Members: Lu, Mingming (Committee Chair).

Subjects/Keywords: Environmental Engineering; Black Carbon; Mobile Source; Dispersion; CALINE4; GAM; Air Pollution

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Agharkar, A. (2017). Model Validation and Comparative Performance Evaluation of MOVES/CALINE4 and Generalized Additive Models for Near-Road Black Carbon Prediction. (Masters Thesis). University of Cincinnati. Retrieved from http://rave.ohiolink.edu/etdc/view?acc_num=ucin1490350586489513

Chicago Manual of Style (16th Edition):

Agharkar, Amal. “Model Validation and Comparative Performance Evaluation of MOVES/CALINE4 and Generalized Additive Models for Near-Road Black Carbon Prediction.” 2017. Masters Thesis, University of Cincinnati. Accessed August 23, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1490350586489513.

MLA Handbook (7th Edition):

Agharkar, Amal. “Model Validation and Comparative Performance Evaluation of MOVES/CALINE4 and Generalized Additive Models for Near-Road Black Carbon Prediction.” 2017. Web. 23 Aug 2017.

Vancouver:

Agharkar A. Model Validation and Comparative Performance Evaluation of MOVES/CALINE4 and Generalized Additive Models for Near-Road Black Carbon Prediction. [Internet] [Masters thesis]. University of Cincinnati; 2017. [cited 2017 Aug 23]. Available from: http://rave.ohiolink.edu/etdc/view?acc_num=ucin1490350586489513.

Council of Science Editors:

Agharkar A. Model Validation and Comparative Performance Evaluation of MOVES/CALINE4 and Generalized Additive Models for Near-Road Black Carbon Prediction. [Masters Thesis]. University of Cincinnati; 2017. Available from: http://rave.ohiolink.edu/etdc/view?acc_num=ucin1490350586489513

.