Skill summary charts are available in the same way as reliability diagrams via the dashboard. In addition to score summaries per grid point, the hindcast summary endpoint provides accuracy metrics per region and custom shapefiles. Using this feature allows the user to quickly see the skill that Salient forecasts provide over either reference models or climatology.
The chart includes a summary of accuracy metrics for a given variable, season, and location over the entire backtesting period in a separate tab. The reference model is based on the lead time.
The skill metric, absolute model error, is a measure of how well forecasts match observed outcomes. Lower skill metric values are better where 0 = totally accurate and 1 = totally inaccurate. However, a higher skill score is better, because it shows the incremental improvement over the reference model.
Figure 1 - Temperature skill metrics for weekly timescale in Jun/Jul/Aug for Salient Blend compared to climatology at 29.3N, -98.4W (San Antonio, TX).
https://api.salientpredictions.com/v2/documentation/api/#/Validation/hindcast_summary