On Wind Chill…

The climatology of wind chill in the PNR is investigated in order to improve the effectiveness of the wind chill warning bulletin.  We compiled event and bulletin statistics for 366 sites utilizing several differing sets of warning criteria, including a new proposal to eliminate all but the minimum equivalent temperature requirement.

Under current criteria, the PNR experiences an average of 736 events per year with a mean duration of 7.4 hours and standard deviation of 2.2 hours.  Data variability is large, with 12% of sites having never logged a single event while numerous Arctic sites see an average of 15 events per year.  Increasing the minimum required wind speed to 20 km/h yielded a 38% reduction in yearly events, while independently increasing the minimum time duration to 6 hours yielded a 46% reduction; the elimination of both requirements entirely dramatically increased the event count by 334%.

Individual site data were then collated into regions and a simple 24-hour forecast and reassessment cycle was used to estimate the production of new bulletins.  Bulletin statistics specific to Alberta nad the southern Prairies showed a similar result.  Under current criteria, 44 original WWCN bulletins must be issued for 187 regions on average each year.  The increase to 20 km/h showed a reduction to 31 bulletins (~30%) and 108 regions (~42%), while the increase to 6 hours gave a reduction to 33 bulletins (~25%) and 130 regions (~30%).  The elimination of wind and duration again showed a large increase to 95 bulletins (+116%) and 487 regions (+160%).  However, objectively selecting the best particular set of criteria still proved difficult; wind speed is the limiting factor, as a minimum of 70% of the variance of any wind chill calculation was found to result form a calculated 2 knot standard deviation in the wind speed.  The wind chill parameter itself can thus only be determined to within 2 equivalent degrees over a 95% confidence interval.

We conclude that while any increase in minimum wind speed and duration produces a reduction in both events and warning bulletins, a more subjective analysis still needs to be investigated to find a better balance between forecaster workload and value to the general public.

A short summary of the work that Jason and I have done so far on wind chill in the PNR region.  To be presented at this year’s change of season workshop.