Fig. 5: Relationship between rain-snow distribution overlap density and method accuracy. | Nature Communications

Fig. 5: Relationship between rain-snow distribution overlap density and method accuracy.

From: Machine learning shows a limit to rain-snow partitioning accuracy when using near-surface meteorology

Fig. 5

The average accuracy (a) of all precipitation phase partitioning methods (benchmarks and machine learning methods) and the rain-snow distribution overlap (b) for the synoptic and crowdsourced datasets are both plotted against air temperature. Average phase partitioning method accuracy is plotted against the rain-snow distribution overlap, where the dashed lines are the lines of best fit computed from ordinary least squares regression (c). Source data are provided as a Source Data file.

Back to article page