Date of Award

2020-01-01

Degree Name

Doctor of Philosophy

Department

Mathematical Sciences

Advisor(s)

rosa fitzgerald

Abstract

Improvement of the Accuracy and Forecast capability of the Numerical Weather Prediction (NWP) models and Air Quality Models (AQM) are critical issues in today's scientific study of Meteorology and Air Pollution. The models for these predictions are dependent on topography, climate, initial and boundary conditions, domain size, and computational efficiency. Different techniques such as Data Assimilation, Ensemble Methods, Increased Computing Capacity to achieve higher model resolution, and Improved Physics Schemes can be used to address this problem. In this study, the NWP models, the Weather and Research Forecast (WRF), and the HYSPLIT models, were enhanced for the Paso Del Norte (PdN) region by applying these techniques. In addition, the CAMx model was refined and successfully implemented for this region. The PdN region comprises El Paso, TX, Ciudad Juarez, Mexico, and some neighboring cities in New Mexico, an ideal region to perform air quality studies. Several sources of experimental data, such as Radiosonde, Ozonesonde, Satellite-based sounder profile, Continuous Ambient Monitoring Stations (CAMS), and Ceilometer, were used in this study. Selecting the best Physics and Chemistry schemes for the models was also a challenging part of this study. Different data assimilation techniques like Incorporating Satellite observations from METOPS and NOAA-18/19, METAR data, NWS data was another objective of this work. To validate the NWP models' results for the PdN region, they were inter-compared with meteorological satellite data, ground stations, and radiosonde datasets. In this study, the ozone results from CAMx were extensively inter-compared with ozonesonde datasets. Meteorological variables such as temperature, pressure, relative humidity, wind speed, and ozone concentrations were also analyzed at several locations in the PdN region. Additionally, several other studies using statistical analysis and machine learning were performed to predict ground-level pollutant concentration. An in-depth sensitivity analysis of the planetary boundary layer using different meteorological schemes was also conducted. This improved the accuracy of the PBL retrieval for this region. Retrieval methods using observational data (Ceilometers, radiosonde), NWP models (WRF, HYSPLIT), and Satellite data (CALIPSO) were inter-compared for validation and calibration.

Language

en

Provenance

Received from ProQuest

File Size

164 pages

File Format

application/pdf

Rights Holder

Suhail Mahmud

Download

Share

COinS