After A Few Cool Summers, A Cruel Summer
Written by: William R. Deedler, Weather Historian NWS White Lake Mi
Sep 10th 2005

While the past couple of summers across Southeast Lower Michigan have been relatively cool, the Summer of 2005 completely reversed that trend and was more of a cruel summer with extensive heat and humidity. Not since the Summer of 2002, has a summer been even close to being as hot as this past one.

The headline on the Summer Outlook issued at summer''s open read? How abouta warmer summer?? Well, the warmer summer came and then some! So much so,that officially the Summer of 2005 will go down in weather history as the warmest summer on record for the Detroit Metropolitan Area with an average temperature of 74.76 (rounded off at 74.8). Straight away, it mustbe mentioned that the urban heat island in Metro Detroit did help influence by modifying mainly the overnight lows and thus, raise the overall average temperature somewhat. The daytime highs are less influenced by the urban heat island as compared to the overnight lows. Therefore, the summer daytime highs is where a closer examination, further down in this report, will tell more of the story.

The Summer of 2005 was the 7th hottest summer on record at Flint with an average temperature of 71.56 (rounded - 71.6). In sharp contrast, it was just last year that the Summer of 2004 took 5th place in Flint''s coldest summers with an average temperature of just 66.7)! Further north, across the Saginaw Valley and Thumb Region, this summer was relatively "cooler" with Saginaw barely placing in the 20th position for hottest summers with an average temperature of 70.95 (rounded - 70.9).

Temp/Pcpn Temp/Pcpn Temp/Pcpn Temp/Pcpn

Detroit 2005 74.1 / 1.95 75.4 / 5.38 74.8 / 1.33 74.8 / 8.66
Norms 69.0 / 3.55 73.5 / 3.16 71.8 / 3.10 71.4 / 9.81
Departures +5.1 /-1.60 +1.9 /+2.22 +3.0 /-1.77 +3.4 /-1.15

Flint 2005 71.4 / 1.97 72.1 / 5.43 71.4 / 0.91 71.6 / 8.31
Norms 66.2 / 3.07 68.4 / 3.17 68.5 / 3.43 68.4 / 9.67
Departures +5.4 /-1.10 +3.6 /+2.26 +2.9 /-2.52 +3.2 /-1.36

Saginaw 2005 71.7 / 3.56 71.2 / 4.03 70.0 / 1.37 70.9 / 8.96
Norms 66.8 / 3.06 71.2 / 2.50 68.7 / 3.38 68.9 / 8.94
Departures +4.9 / +.50 0.0 /+1.53 +1.3 /-2.01 +2.0 / +.02

NWS White Lake 70.2 / 2.67 70.8 / 5.85 70.5 / 1.25 70.5 / 9.77
+ / N

(Departures based on standard 1971-2000 30yr norms)
Top 12 Warmest Summers in Southeast Lower Michigan
                     Detroit(1870)  Flint(1942)    Saginaw(1898)  
Rank Temp Year Temp Year Temp Year
1- 74.76 2005 72.33 1949 73.07 1931
2- 74.50 1995 72.20 1955 73.03 1933
3- 74.43 1955 71.83 1988 72.40 1955
4- 74.20 1988 71.73 1987 72.23 1995
5- 74.00 1933 71.70 1995 72.07 1930
6- 73.80 1949 71.66 2002 72.07 1921
7- 73.67 1921 71.63 2005 72.03 1988
8- 73.57 1991 71.53 1983 71.93 1937
9- 73.57 1952 71.50 1943 71.87 1936
10- 73.50 1959 71.30 1944 71.70 1998
11- 73.50 1931 **70.86 1952 71.53 1934
12- 73.46 2002 70.47 1991 71.53 1932
* * 20- 70.95 2005

**Note: While doing this report, a significant mistake was caught in the 1952 summer records at Flint. The Flint summer average temperature for 1952 should read 70.86 instead of 72.53. This was due to the fact that the August 1952 average temperature should have been 68.7 rather than 73.7. This considerably lower summer average temperature resulted in the Summer of 1952 to fall from first place to 11th.
The following Map (link) shows the Summer of 2005 temperature departure across the Upper Midwest and Great Lakes Region.

There is no doubt about it, on an average statistical level, the Summer of 2005 will go down as the hottest summer on record in Detroit. While the average of the high and low temperatures is used for the official summer average, maybe a more pertinent comparison of the summer heat would be seen if one examines the average summer high and low in the hottest of summers at Detroit.

                              Fig 3
1 2 3 4 5
2005 1995 1955 1988 1933
Ave High Ave High Ave High Ave High Ave High
Jun 84.1 Jun 82.0 Jun 78.5 Jun 83.9 Jun 85.3
Jul 84.9 Jul 84.5 Jul 89.6 Jul 89.3 Jul 85.0
Aug 84.1 Aug 85.3 Aug 85.2 Aug 85.3 Aug 80.4
Sum: 84.4 Sum: 83.9 Sum: 84.4 Sum: 86.2 Sum: 83.6
2005 1995 1955 1988 1933
Ave Low Ave Low Ave Low Ave Low Ave Low
Jun 64.0 Jun 61.1 Jun 58.5 Jun 56.9 Jun 63.9
Jul 65.9 Jul 65.0 Jul 68.5 Jul 64.9 Jul 65.8
Aug 65.6 Aug 69.0 Aug 66.2 Aug 64.8 Aug 63.5
Sum:65.2 Sum: 65.0 Sum: 64.5 Sum 62.2 Sum: 64.4

The above statistics give a better comparison of daytime heat during our hottest of summers in Metropolitan Detroit while addressing the question that may arise about this summer. While the Summer of 2005 was hot, how could it be hotter and rank higher than 1988? That was one miserable summer with all it''s 90''s and 100''s?? As one can see when comparing the two summer''s average highs above, this is a legitimate question. However, while the average high in the Summer of 1988 was a full (and significant) 1.8 degrees above the Summer of 2005, it was the average of the lows in the Summer of ''88 that dropped the overall average significantly (especially in June of ''88).

More parallels can be drawn between the Summer of 2005 and 1955 where, the summer average statistics are nearly identical. Both summers had the exact average high and similar average lows. True, the Summer of ''55 had a bit cooler average low but when consideration is given to the expansive urban heat environment the past 50 years (and is estimated and then deducted out of the low this past summer), the two summer averages may have very well been closer or even identical. While the summers of 2005, 1995 and 1955 were all hot and humid summers, and as mentioned above, the summer of 1988 was one unusually hot but relatively dry (rainfall and humidity - though the rains did start to come the second half). I say unusually hot because of its record amount of 90s and five 100+ temperatures. The Summer of 1936 with its granddaddy of all heat waves, only supersedes the amount of 100+ degree days felt in 1988. Ironically, the Summer of 1936 is not even on the hottest summers list. You see, June of 1936 was unseasonably cool, averaging only 65.9 or 3.1 degrees below normal and thus, pretty much cancelled out the hot July (average of 74.7). (As an interesting side note, there was nearly a ten degree difference in the average temperatures between the two back to back summer months)!


While there was the occasional break in the heat during the summer, the heat was fairly persistent as partially evidenced by the limited range of the three monthly average high temperatures. In all three months, the average high temperature varied less than a degree (84.1-84.9) in Detroit, something not seen in any of the other hot summers.
While the amount of high temperatures reaching 90 or above were above normal atall three sites, they were nowhere near the records. Detroit totaled up 20 days this year, whereas the record in 1988 is almost double at 39! (even as recent as 2002, there were 25 days with 90(+) degrees reported). In Flint, on 13 days the temperature reached 90(+) degrees this summer, also a far cry from their record of 36 back in ?88 (in the Summer of 2002, 24 were recorded). Saginaw totaled up ten days in which the mercury hit 90 orabove and the record for Saginaw is also 39 days back in 1931 (in 1988 there were 37 days and in 2002, 13 days).
Another fact in the persistence heat camp was the amount of days that the temperature rose to 80 or above. There were 74 days (out of the possible 92) this past summer where the mercury rose to 80 degrees or above at Detroit. In the longest climate record base available (Detroit back to 1870) this happened only one other year, 1991 which placed 8th for overall warmest summer. This also helps explain why the summer ranked so high for heat yet there were no record highs nor temperatures even above 95 at Detroit. This, despite the fact, that the other top 4 hot summers listed all contained temperatures in the 100s. In Flint, 68 days the temperature rose to 80 or above during the past summer, just second to 1995 and 1983 when there were 69 days. Interesting, there was only one record high (91/6-9-05) and believe it or not, a record low (44/6-19-05) ten days later at Flint for the summer. In the relatively cooler Saginaw Valley, Saginaw had 61 days of 80(+) this summer, far removed from the heat of 1933 when 77 days were totaled. The above data also gives credence to the fact why only one day this summer met the heat advisory criteria. It was simply a case, that whether it be the temperature, duration of that temperature and/or amount of relative humidity, it did not fulfill the criteria.

Finally, and to add "fuel to the fire" so to speak, not only did persistence play a role in this past summers uncomfortable climate, the Summer of 2005 also contrasted quite effectively to our rough and snowy winter which was followed by one of the coolest springs on record. And, as briefly touched on above, the past few summers across Southeast Lower Michigan as a comparison have been relatively cool, especially last summer where both Flint and Saginaw placed in the top 10 coolest summers, 5th and 9th respectively.


A wet July this past summer was sandwiched in between two dry months, June and August across Southeast Lower Michigan. The wet July was pretty wide spread across the entire area. Note the following rainfalls for July from our regular three cities and co-operative weather observers, who fill the open areas in nicely.

 Location      July rainfall amount       Location     July rainfall amount
Ann Arbor 6.67 Newport 7.27
Cass City 4.58 Oxford 7.06
Chelsea 8.04 Port Huron 5.29
Corunna 6.30 Richmond 7.13
Detroit /DTW/ 5.38 Saginaw 4.03
Durand 6.41 Saline 7.57
Flint /FNT 5.43 Sandusky 5.19
7W Flint 7.25 Tecumseh 5.09
Hudson 7.20 Tipton 6.70
Milan 5.33 White Lake /DTX/ 5.85
Morenci 5.48


Not only did the severe weather season start late and light (only 4 events in May) but picked up with a vengeance in June (the most severe weather events and warning on file for June since 1986 and second most events and warnings for any month. While July was relatively active with severe weather it stalled to virtually nothing in August (and remans that way through mid September). This was the first time since 1986 anyway, that no severe weather events occurred in August.
Previously, the lowest severe weather events for August was just one event, back in 1994. While it is not uncommon for the severe season to "dry up" somewhat by August, to have none is unusual. It''s somewhat ironic, given the fact that it came right after the "stormiest month" (Jul ''05) on record at Detroit, when thunderstorms were observed on 15 days (more feast or famine scenarios). The only other summer month that had no severe weather recorded since 1986 was June of ?88, when it was exceptionally dry and hot. The last severe weather day across Southeast Lower Michigan was July 26th, nearly a month and half ago as of this writing. Note, that while June and July were somewhat deficit in rainfall, the wet July helped bring the overall summer rainfall across Southeast Lower Michigan up closer to normal.
For a regional map of rainfall and departures, click here.


Our preferred summer stats in the Summer Outlook (released early June) stronglysuggested a warm to hot summer with temperatures averaging normal to above for the summer. From the Outlook:
"our analogue summers, recent hemispheric patterns and projections hint that the dominant below normal temperatures of the spring will fade out in June."
Will Summer Sizzle or Fizzle?
In the summers that followed our two closest analogue winters (1926/1952) indicate that the Summer of ''26 started very cool and then warmed closer to normal by mid-late summer, while the Summer of ''52 was a warm to hot summer. This overall warmer pattern displayed in those two summers (over the past few summers) also influenced in the final temperature projection of this outlook."

As can be seen in the original outlook (see Chart 2) when lookingat the 15 summers in the original Outlook research the preferred warm to hot summers were great indicators for this past summer. In our Outlook, the summers listed on the chart that were the best indicators include 1995, 1987, 1952, 1932 and 1919. In addition, our recent Summer of 2002 in many ways mirrored this past summer. Here, a very weak El Nino (or negligible pattern) was in effect too and thus, it too had little or no affect on our summer. Finally, the summer''s of 1995 (former hottest summer in Detroit), 1987, 1952, 1932 and 1919 (along with 2002) all appear in the top 20 hottest summers on record in Southeast Lower Michigan (and this is going as far back as 1870). In the end, both projections from 1952 and 1995 worked out the best for summer above normal temperature projections (also were the first and eighth hottest summers on record in Detroit)


From the Outlook:
"While our analogue summer years Chart 2 indicate a variety of summer conditions with variable precipitation amounts, one trend that is evident is that about half of the summers tended to start out dry but became wetter (or normal to above normal rainfall) as they evolved. Note, here we are interested more of the overall rainfall trend of the summer, and not necessarily each specific month's amount. This idea of starting the summer on a dry note certainly reflects and would continue the trend of the last three months with Spring of 2005 being one of the driest springs on record! This trend will have to be watched because of its persistence. As of this time, however, look for rainfall to total around normal by summer's end."
The expected trend of dry to wet for the summer happened (and strongly) this past summer but failed to remain there. We went from a dry June to a wet July then back again to a dry August. The larger pattern, however, has really been one of dryness with the dry period that commenced in early spring pretty well dominating all months except July. Remember too, it was in July that the whole area received beneficial rains from the remnants of Hurricane Dennis which added nicely to the total rainfall.
This concern for longer periods of wet and dry regimes was commented on Outlook and appears to be happening (wet late fall into winter - dry spring into much of the summer) But it is still too early to commit to the trend. Note the following taken from the Outlook:
"It is felt that the pattern of relatively dry and tranquil weather followed by notably wet and stormy periods of the past several seasons continues to persist (and each, possibly now for longer durations). While this still confirms the "feast or famine" regime we've become more accustom to the past several years, we will have to keep watch for the possibility of the evolution of longer durations of each (wet and dry)."
Scanning the summer rainfalls and patterns suggest that the Summer of ’52 again had a handle on the rainfall amounts, actual pattern of dry, wet, dry and even location with the wetter areas in the north sections of our CWA (Saginaw Valley and Thumb Region). It is interesting to note our specific trend rainfall guidance was not that far off when it estimated 9.34â€Â or below the normal of 9.81â€Â for Detroit (actual was below at 8.66â€Â). Flint with a specific estimate of 8.31â€Â also below normal, received a close 8.53â€Â (below normal) while Saginaw, with a forecasted estimate of 9.67â€Â (the highest of the three and actually above normal did indeed total just a bit above normal at 8.96). Therefore, our projected precipitation trend and location for the summer, worked out fine. Since projected rainfall amounts ranged from just above normal to just below, the near normal was used for rainfall over the entire Southeast Michigan region rather than splitting it up (which is not advisable) but this was somewhat too generous in some areas.


As with many of our nicer early falls, this fall indicates temperatures averaging above normal early fall, then normal or below in mid to late fall. Frost and Freezes should arrive near their normal dates but timing was extremely variable due to the quick resurgence of the impressive roller-coaster temperature pattern seen during the cooler months of the past several years under the weak El Nino to Neutral regime. The present Neutral Phase inthe Pacific is slated to hold on at least into early winter. This pattern could help make for some interesting fluctuations in temperatures again but remember, that also tends to give us a better chance for Indian Summer weather - as it did last autumn.

Sep - early Oct Temperatures averaged above normal in 11 of the 15 Septembers (or 73%) and some by several degrees. This warm trend with above seasonal levels seems realistic given the recent trends, dry soil and similar past early falls. October''s temperature data gives less of a dominant signal with 7 above normal Octobers, 6 below and 2 normal (or equalMid Oct - Nov chances). This also seems to reflect a change in the prevelent above normal trend beforehand to a more normal to below temperature trend mid - late autumn. Interestingly, November''s data does seem to back that up with a the majority (8) below normal, 3 normal and just 4 above.

Precipitation: While rainfall amounts and frequency increase during the fall, the overall dry pattern of late may be hard to budge. It is interesting to note that September of ''52 and ''95 (our good analogues for the summer) both show below normal rain, especially 1995 where only 0.62" fell in September. With that in mind, look for normal to below normal rain this fall. Again, some early snows showed up mid - late autumn (not unlike last autumn) but this season just 6 out of 15 reflect above normal amounts in November.
We shall see.

Look for the Winter 2005-06 Outlook late in October or early November.

Dates of Interests for the Autumn (all data based on Detroit data base)
Autumn arrives: September 22nd at 623PM EDT
Average date of the first killing frost/freeze (not light frost): October 21st
Earliest killing frost/freeze in Detroit: September 22nd, 1974
Average first date of snow (Trace): Approximately Halloween (or last week of October)
Earliest appearance of snow recorded in Detroit (Trace): October 1st, 1974 is the U.S. government's official web portal to all federal, state and local government web resources and services.