A Method for Determining Mid-Latitude Seasons

Date : 16 March 1997

As Spring Equinox approaches, you may be wondering : "I know the calendar says it's Spring, but sure feels like Winter to me" (especially in the NE U.S. early this week). Some of the largest snows have occurred during 'Spring'; e.g., 25 inches during 23-24 March at Kansas City, MO - their largest storm total ! Detroit, MI's largest of 24.5 inches occurred during April ! Perhaps the calendar lies...

Most people are familiar with astronomical seasons :

Winter being from Winter Solstice to Spring Equinox, etc. Equinox is a latin word for equal time, and specifically is the intersection of earth's orbital and equatorial planes, thru our sun's center. If we had no atmosphere, our planet's Northern Hemisphere would be coldest near Winter Solstice and warmest near Summer Solstice. Things would be quite simple, and no life would exist. In our complicated atmosphere though, air and water require time to heat, and circulate to distribute heat acquired. Thus, meteorological seasons are often defined as :

Season Months Winter December, January, February Spring March, April, May Summer June, July, August Autumn September, October, November

indicating almost a month lag from seasons according to solar elevation angle (e.g., lowest 3 months from about 5 NOV to 5 FEB). That is okay as an approximation, but the situation is more subtle. Not only does solar elevation angle change more abruptly near equinoxes than solstices, but geometry causes us to experience a more abrupt change from Winter to Summer (or vice-versa) than it or our calendar indicate :

@ Spring Equinox, our sun has made about 2/3 of its change from Winter to Summer situation, though only about 1/2 the days elapse.

Does a better way to determine seasons exist ? Vegetation, harvest times, and other activities (fish spawning, etc.) all indicate changing seasons. Quite often, growing seasons are referred to. On the diagram to the right, the graph indicates average temperatures (from monthly averages) obtained for Glendive, MT from the Lamont-Doherty Earth Observatory climate browser. I defined 4 seasons according to equal temperature changes during each. (Better ways exist to calculate this, but it can suffice.) You may notice that Summer and Winter according to this are about 4 months, Spring and Autumn about 2. Spring and Autumn are transitional seasons between our long, dark winters and long, hot summers (at middle and high latitudes). Is the data for Glendive typical ? To help answer that question I decided to randomly select 16 locations, one for each Region as defined in the Old Farmer's Almanac. I wouldn't believe many of their daily forecasts, but their seasonal ones can often be useful. The regions are very well-chosen, consistent climates occurring in each, if they must be limited to 16. Below is a listing of seasonal graphs according to Region (Region number, location, lat, lon ; elevation) :

 1 Ripogenus Dam, ME  45.88 °N, 69.18 °W ; 294 m
 2 Charlotteburg, NJ  41.03 °N, 74.43 °W ; 232 m
 3 Fredericksburg, VA  38.32 °N, 77.4 °W ; 27 m
 4 Asheville, NC  35.43 °N, 82.55 °W ; 659 m
 5 Saint Leo, FL  28.33 °N, 82.27 °W ; 58 m
 6 Elmira, NY  42.08 °N, 76.78 °W ; 256 m
 7 Hopkinsville, KY  36.83 °N, 87.50 °W ; 180 m
 8 Scottsboro, AL  34.68 °N, 86.05 °W ; 187 m
 9 La Porte, IN  41.60 °N, 87.62 °W ; 247 m
10 Medford, WI  45.13 °N, 90.35 °W ; 448 m
11 El Dorado, KS  37.82 °N, 96.83 °W ; 408 m
12 Amarillo, TX  35.23 °N, 101.7 °W ; 1099 m
13 Lakeview, OR  42.22 °N, 120.37 °W ; 1455 m
14 Needles, CA  34.77 °N, 114.62 °W ; 279 m
15 Portland, OR  45.60 °N, 122.60 °W ; 7 m
16 Chulta Vista, CA  32.62 °N, 117.08 °W ; 17 m

You may notice a very similar trend among them, mainly because they are all for the continental United States (some almost appear identical). Many more of the graphs are in the eastern than western U.S., and as luck determined, a large void exists among these because Region 13's choice was in Oregon, Region 10's in Wisconsin, and Region 14's in California...the beauty & ugliness of randomness ! Thus, you might wish to think of this a geographical distribution according to population rather than area. Maximum daily average temperature at Needles, CA is interesting (almost 35 °C (95 °F) !), one of the hottest places in the U.S. You may notice that seasonal lag is greater for western than eastern locations, influenced by cold ocean currents during Spring. Averaging for the locations determines seasons as :

   Season        Dates        Duration
   Winter   23 NOV - 16 MAR   114 days
   Spring   17 MAR - 26 MAY    71 days
   Summer   27 MAY - 23 SEP   120 days
   Autumn   24 SEP - 22 NOV    60 days

Average temperatures are near 20 °C (68 °F) at many locations during the beginning of Summer, meaning maximum temperatures are in the 70's to low 80's (°F) at some locations, so this method is realistic (as well as logical).

These seem to agree well with experience, though adjustments for other factors (some mentioned above) might be appropriate. E.g., plants still bloom during May and perhaps even June, but many places become like a little jungle during July & August Ice requires time to melt, and lakes remain warm during 'Autumn', etc. Considering all of such things, the following dates may be more realistic :

   Season        Dates        Duration
   Winter   28 NOV - 22 MAR   116 days
   Spring   23 MAR -  1 JUN    70 days
   Summer    2 JUN - 29 SEP   120 days
   Autumn   30 SEP - 27 NOV    59 days

This does not consider precipitation, which can be responsible for types of seasons. Consider if you wish, annual monthly precipitation averages for 3 locations. Data for Saint Louis, MO is typical of much of the continental United States, heaviest precipitation often with abundant showers during late Spring & early Summer. Data for Phoenix, AZ illustrates the monsoons caused by a large-scale heat Low which often develops over our desert southwest states during late Summer & early Autumn. Another type of late-season precipitation maximum is illustrated for data from Wawa, Ontario (at the east side of Lake Superior), from lake-enhanced precipitation. The maximum amounts during August & September are caused by that and mean position of jet streams aloft, which begin southward progression then. The small maximum during November would be greater if liquid snow amounts were greater - lake-effect snow greatly contributing to it.

Such considerations are important regarding world climates, which often are not so easily defined as in the continental U.S., which is the feature for next week.

Text and all graphics named sea-___.gif are copyright of Joseph Bartlo, though may be used with proper crediting.

Home Page