Since percentage / confidence level based outlooks are the basis of this webpage, I'd like to give some insight into how I come up with them, and what they mean.
-How I make the snowday percentages-While I have no defined mathematical method of coming up with the numbers for the winter threat percentages, slight deviations in the snowday percentages are important. Therefore, contrary to the idea that I pull numbers out of the air, I do use a program that I made to calculate snowday percentages. The program takes into account the average surface air temperature over the last five days, and hourly accumulations of snow and ice between 12am midnight and 2pm. These inputs have to be entered by me, so of course there is also the error involved in forecasting. I usually use fairly close to the output value from the program, but there does need to be some human observation and intuition involved (for example in the period following the Ice storm of 2008).
So the main factors that affect the snowday forecast are:
Timing (of accumulations, precipitation rate, and precipitation type)
Precipitation type
Precipitation rate
Accumulations
How much of the accumulations will stick to the roads
-What the percentages mean-
I base the percentages off the system the NWS uses for precipitation probability in their forecasts:
0-30% = Slight Chance
30-60% = Chance
60-80% = Likely
80-100% = Definite
In situations that the storm occurs overnight, the delay percentages will count for the cumulative chance for a delay or a cancellation. In other words, At least a delay. Sometimes I will have percentages up for a delay, cancellation, and early release, and people will ask why this is. In these cases, the storm is likely to be at least 6 hours in duration, but the timing of the precipitation type changes and rates is uncertain.
|posted by Sam Lillo @ 1/01/2005 10:10:00 AM
