spatial predication comparison test

Wednesday, April 25, 2018 - 1:00pm - 1:30pm
Eric Gilleland (National Center for Atmospheric Research)
With the advent of high-resolution forecasts came new challenges for analyzing their performance. It is no longer satisfactory to report only the root mean square error or similar grid-point-by-grid-point accumulative summaries. Often such scores give misleading information caused by either double penalties (when a small error in timing or spatial placement leads to a miss and a false alarm accumulation for this same error) or over accumulation of small-scale errors.
Subscribe to RSS - spatial predication comparison test