Predictability Assessment and Improving Ensemble Forecasts
Abstract
The PI is examining atmospheric predictability with the goal of improving ensemble forecasts at ranges of 12 hours to 10 days. The research is addressing several issues, including: a) Documentation of analysis uncertainty from mesoscale and global analyses. b) Calibration of ensemble forecast system (EFS) output by artificial neural networks. c) Design of optimal EFS's, with an emphasis on precipitation forecasts. d) Design of stochastic physics parameterizations that improve under-dispersion in EFS s.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 2001
- Accession Number
- ADA625671
Entities
People
- Steven L. Mullen
Organizations
- University of Arizona