Author ORCID Identifier:
Date of Graduation
7-2026
Document Type
Thesis
Degree Name
Master of Science in Statistics and Analytics (MS)
Degree Level
Graduate
Department
Statistics and Analytics
Advisor/Mentor
Majumder, Reetam
Committee Member
Zhang, Qingyant
Second Committee Member
Robinson, Samantha
Keywords
Calibration; Climate; Deep learning; Machine learning; Statistical; Wind
Abstract
This thesis investigates statistical calibration methods for wind data from Global Climate Models (GCMs), with emphasis on wind speed and wind direction. The analysis uses 30 years of daily observed and GCM wind records from 9 spatial locations in the South East USA. Wind is represented in polar coordinates, but also often modeled in Cartesian coordinates. Four calibration models are compared: a marginal wind speed model, an independent Cartesian model. The calibration models are based on semi-parametric quantile regression, a flexible conditional density estimator fitted using neural networks. The neural network objective functions are optimized using maximum likelihood estimation (MLE) and maxi mum a posteriori estimation (MAP) approaches. Performance is evaluated using Wasserstein distance, 95% quantile error, spatial correlation error, and cross-correlation error. The results show that the marginal model provides the strongest overall calibration for wind speed but does not model wind direction. The independent Cartesian model performs poorly for the distribution of wind direction. Introducing conditional dependence improves directional calibration. MLE generally performs better than MAP for the Cartesian models. These findings support the use of conditional Cartesian modeling when joint calibration of wind speed and direction is required.
Citation
OURU, A. (2026). Statistical Machine Learning For Calibrating Climate Model Wind Fields.. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/6420
