A194-03
A New Technique for Spatio-Temporal Reconstruction of Analog Ensemble Predictions
Tuesday, 15 December 2020: 08:38
Virtual
Laura Clemente-Harding, Engineer Research and Development Center, Geospatial Research Laboratory, Alexandria, VA, United States, Guido Cervone, Pennsylvania State University Main Campus, Department of Geography and Institute for Computational and Data Sciences, University Park, PA, United States, Martina Calovi, Pennsylvania State University Main Campus, University Park, PA, United States, Weiming Hu, Penn State University, State College, United States, Sue Ellen Haupt, National Center for Atmospheric Research, Boulder, CO, United States and Luca Delle Monache, Center for Western Weather and Water Extremes (CW3E), Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA, United States
Abstract:
The Analog Ensemble (AnEn) technique has been shown to generate well calibrated, bias adjusted ensemble predictions for renewable energy forecasting, air quality forecasting, 2-m temperature, 10-m wind speed, tropical cyclone intensity, and select downscaling tasks. The AnEn technique produces these ensemble predictions using fewer computational resources than traditional ensemble forecast methods. The computational implementation of the AnEn is performed at a single point (or grid) in space over a 3-point time window. This independent grid by grid implementation enables efficient processing, however, it also degrades the spatio-temporal consistency of the forecast field.
The Schaake Shuffle (SS) technique, originally developed by Clark et al. (2004), reconstructs spatial and temporal consistency for temperature and precipitation forecast fields. This research develops a deeper understanding of the continuity in space and time as reconstructed by the SS. Next, this work presents a new alternative to the SS technique. This new technique uses historical information, available through the AnEn results, to inform development of the Copula function used in the spatial and temporal reconstruction process. This new alternative SS technique is applied to extreme heat event forecasting in New York City. Extreme heat predictions over New York City are post-processed using the new alternative SS technique to produce spatially and temporally consistent maps used to analyze hazards related to extreme heat events.
