Extension of the Analog Ensemble Technique to the Spatial Domain
Open Access
- Author:
- Clemente-Harding, Laura
- Graduate Program:
- Geography
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 10, 2019
- Committee Members:
- Guido Cervone, Dissertation Advisor/Co-Advisor
Guido Cervone, Committee Chair/Co-Chair
Alexander Klippel, Committee Member
Robert George Crane, Committee Member
George Spencer Young, Outside Member
Sue Ellen Haupt, Outside Member
Cynthia Ann Brewer, Program Head/Chair - Keywords:
- Analog Ensemble
uncertainty quantification
Search Space Extension
Schaake Shuffle
ensemble prediction
spatio-temporal consistency
computationally efficient
spatial optimization
optimal predictor weighting - Abstract:
- The Analog Ensemble (AnEn) technique enables probabilistic predictions using fewer computational resources than traditional ensemble prediction methods. The technique performs well for wind and solar energy prediction, air quality forecasting, 2-m temperature, 10-m wind speed, and select downscaling efforts. The original AnEn-Independent Search (IS) technique presented by Delle Monache et al. (2013) was implemented at a single point in space over a n-point time window and uses a historical repository of corresponding deterministic predictions and observations. This research extends and deepens scientific understanding of the AnEn through the following subject areas: • Optimal predictor weighting of the Analog Ensemble (AnEn) technique for short-term (0-48 hour) probabilistic forecasts. • Artificial expansion of the historical search repository. • Spatial and temporal consistency. Knowledge generated is domain-independent with implications in physical science subject areas, in which 1) single deterministic predictions, past predictions, and their corresponding observations are available; 2) it is necessary to have quantifiable and justifiable measures of uncertainty; and 3) effective management of computational resources is paramount.