The Automation of Nowcast Model Assessment Processes
Abstract
Nowcast model assessment involves applying model verification techniques to generate error statistics to improve model performance and the accuracy of forecasts produced by US Army Research Laboratory's nowcast model, Weather Running Estimate-Nowcast (WRE-N). This report documents the design and implementation of the automated process of generating domain-level error statistics that can be used by modelers to improve the accuracy of WRE-N model forecasts. This process allows multiple user configurations, and produces a controlled data structure that could be easily used in data analysis and the evaluation of model improvements.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2016
- Accession Number
- AD1017503
Entities
People
- Jeffrey A. Smith
- John W. Raby
- Leelinda P. Dawson
Organizations
- United States Army Research Laboratory