Is Winter Coming Later to Washington State?
The Office of the Washington State Climatologist sometimes receives questions about whether the timing of the seasons has shifted. Ken Kunkel and collaborators have examined how the length of the growing season has changed across the lower 48 states, and others have documented the changes from a regional perspective (e.g., Abatzoglou et al., 2014 and the Climate Impacts Group). Here we will provide a modest contribution to this body of work, focusing on the winter season and including consideration of the variability at individual locations.
As is the typical practice, we key on the date of the first freeze, i.e., a minimum temperature less than or equal to 32°F. We use the daily minimum temperatures from a number of west side and east side stations for two reasons: (1) minimum temperatures, especially in clear and calm conditions, can vary so much from place to place, and (2) dates of weather events represent inherently noisy time series. We also selected stations away from the large urban areas of central Puget Sound and eastern Washington (Seattle and Spokane – you are off the hook this time). Our analysis spans the period from the late 1940s, when many of the present weather stations across WA state were established, through 2017.
Time series of the first freeze dates for the west side stations of Hoquiam, Longview, Olympia, Olga (southern Orcas Island), and Bellingham are shown in Figure 1. Time series for the east side stations of Richland, Walla Walla, Cle Elum, Ritzville, Chelan, and Newport are shown in Figure 2. Not surprisingly, there are systematic differences in timing among the two groups of stations. For example, of the western group, Olympia generally gets an early frost and Olga tends to be late, and of the eastern group, Cle Elum on the east slope of the Cascade Mountains is early and Walla Walla is late. During some years there appears to be a cold snap such that most stations get their first hard frost within a few days of one another, and there are other years with more than a couple of months of separation between the first freezes at different locations.
Linear fits to these noisy time series have been made to quantify overall trends. For the west side stations, the average date of the first freeze at the end of the record is about 7-9 days later than at the beginning of the record, depending on whether one averages individual trends versus averaging the mean date for the 5 stations. The two stations with greater marine influences of Hoquiam and Olga exhibit relatively small and negligible trends, respectively. For the six eastern WA stations, the first freeze is now occurring about 14 days later. It is recognized that the period of 1948 to 1951 was relatively cold, especially on west side of Cascades, and that computations of trends are sensitive to values near the beginning and the end of records. Therefore, we also computed linear trends with a starting year of 1952, which yielded the results of a bit smaller delays on the west side and not much differences in the overall trends for the east side.
The larger shift in the timing of the first freeze for the east side is consistent with the overall changes in mean minimum temperatures over the period of 1948-2017, which have risen about 1.5°F in the Puget Sound climate division for the months of October and November combined, and about 2°F in the Central Basin climate division for the months of September and October. We are unsure how much these changes may have already impacted the landscape. Presumably there are effects ranging from the color of fall foliage to the timing of wine grape harvests, and that they will become increasingly prominent in future decades.
