Date of Award

3-2023

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Engineering Physics

First Advisor

Peter A. Saunders, PhD

Abstract

A crucial component of weather forecasting in numerical weather prediction (NWP) is the analysis of the initial state of the atmosphere. Inaccurate analysis of the environment can lead to amplifying errors in the forecast which can cause devastating effects to the population and its resources. In this study, a weather model simulation was performed over the Pacific Northwest to evaluate wind speed forecast performance by assimilating Global Navigation Satellite System (GNSS) radio occultation (RO) data. Two separate events were examined: 24 - 26 October 2021 during which a strong extratropical cyclone struck the area, and 7 - 9 November 2021 during which high pressure dominated with relatively light winds. The Advanced Research Weather Research and Forecasting (WRF-ARW) model was employed with observation data obtained from National Centers for Environmental Prediction (NCEP) and assimilated using the Gridpoint Statistical Interpolation (GSI) 3D-variational (3DVAR) system. Four separate model runs were conducted for each event, to include a control without assimilation, and three with combinations of conventional, satellite, and RO observation data. Verification was performed using surface observations and radiosondes over the forecast period. It was shown that RO observations improved vertical profile wind forecasts at verifying locations, having the smallest average RMSE for both events. Surface wind forecasts improvements proved inconsistent.

DTIC Accession Number

AFIT-ENP-MS-23-M-083

Download

Included in

Meteorology Commons

Share

COinS