Table S3.

Evaluative metrics of DA risk model performance

Evaluative metricsWAORORORCACACACA
Actual number of DA maxima captured*1088856610
Expected number of DA maxima1919202012131313
Number expected during undersampled years038107352
Adjusted number of expected events§19161210510811
Approximate number of false positives98420421
Percent false positives#47503330040259
DA maxima occurring during a month with elevated risk, i.e., warm, upwelling-favorable conditions635445NaNNaN
DA maxima occurring during the 1 mo following upwelling season121310NaNNaN
Total number of upwelling-associated annual DA maxima756755NaNNaN
Percent of DA maxima during warm upwelling season conditions**3731507010050NaNNaN
Total number of months with a sample2912311459857129129142
Total number possible months300300300300192192192192
  • Rows 1 to 6 report calculations for estimates of false positives for each bin of DA data. Rows 7 to 10 report calculations for the percentage of annual maxima that coincided with upwelling season. Upwelling-related values for south and south central CA were not calculated due to the nearly year-round upwelling season that occurs there.

  • * Number of annual DA maxima values exceeding 20 ppm in bin.

  • Positive risk values for 3 consecutive months at any latitude in bin.

  • Number of expected years that were undersampled, i.e., had less than 6 mo with at least one DA sample in the bin.

  • § Number of expected years minus undersampled years.

  • Actual minus expected minus undersampled years.

  • # Number of false positives divided by adjusted number of expected years × 100.

  • Sum of the number of annual DA maxima that occurred during elevated risk (i.e., upwelling-favorable conditions) and those that occurred 1 mo after upwelling season terminated, i.e., during the annual transition from upwelling to downwelling-favorable conditions.

  • ** Sum of the total number of upwelling associated maxima divided by adjusted number of expected years × 100.