Metrics for evaluating linear features
Novel metrics designed to evaluate the skill of geophysical models in simulating discontinuities and linear features are introduced and tested using a sea-ice model and remotely sensed observations of leads. The metrics are formulated as frequency-based indices of agreement, thus maintaining the sim...
Published in: | Geophysical Research Letters |
---|---|
Main Authors: | , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
American Geophysical Union
2008
|
Subjects: | |
Online Access: | http://hdl.handle.net/2440/86181 https://doi.org/10.1029/2008GL035086 |
Summary: | Novel metrics designed to evaluate the skill of geophysical models in simulating discontinuities and linear features are introduced and tested using a sea-ice model and remotely sensed observations of leads. The metrics are formulated as frequency-based indices of agreement, thus maintaining the simplicity desired in model development and operational applications, while remedying known shortcomings of common existing skill metrics. User-selectable spatial scales and features of significance allow for the general use of the metrics in variable applications, scales, and observation types. Gad Levy, Max Coon, Giang Nguyen and Deborah Sulsky |
---|