Challenging the bleak climate forecast for Georgian Bay


By E. Craig Jowett

“State of the Bay” is an initiative of the Georgian Bay Biosphere Reserve (GBBR) which has forecast that water temperatures in Georgian Bay, Ontario, “are warming” and could rise by “10°C this century within coastal ecosystems”, and warns of “catastrophic events”. Upon learning of this claim, I undertook to review actual and historical government weather station records, which are available for anyone to examine.

These records include air temperatures from stations all around Georgian Bay, plus surface water temperatures from moored buoys in northern and southern Georgian Bay.

Temperature and CO2

In regard to the role of “current rates of carbon emissions” or CO2 in air temperature, climate history reveals no telling correlation between world CO2 and air temperatures in SW Georgian Bay, Parry Sound, or Toronto, Ontario. All parameters in Figure 1a are normalized to their respective values in 1980, similar to a percentage change for better visualization. As CO2 climbs predictably and regularly, air temperatures vary unpredictably and do not correlate with CO2.

Subscribe to our Newsletter!

The latest environmental engineering news direct to your inbox. You can unsubscribe at any time.

In the mid-1950s to mid-1970s, Parry Sound was relatively cooler (10% -15% warmer than it was in 1980) compared to SW Georgian Bay (15% – 20% warmer than in 1980); otherwise, they follow similar variations. Downtown Toronto, at 2 – 3°C warmer than Georgian Bay, follows similar relative trends overall, but rises relatively less from its value in 1980 in both directions, and also has no connection to CO2. (Removing standard elevation and latitude effects, Toronto would be 1 – 2°C warmer than expected.)

As described in my article, A mosaic perspective of climate due to natural and societal influence”, which appeared in the August 2019 issue of ES&E Magazine (Jowett 2019), a very close correlation between world population growth and CO2 concentration, plus a time order and a mechanism, infer a causal effect between the two variables. However, with no correlation between Ontario temperatures and CO2 concentration to begin with, there is no method to infer a causal effect at any time over this 80-year period.

The 1 – 2°C temperature anomaly between Georgian Bay and Toronto since WWI supports the notion that urban areas are permanent anthropogenic “domes” of heat, CO2 and water, which are products of combustion. Any collateral warming from a regional CO2 greenhouse effect is not evident.

Figure 1a. Annual temperatures of combined Owen Sound + Wiarton + Chatsworth stations (blue dots), in comparison with downtown Toronto (red dots) and Parry Sound temperatures (violet dots), and with NOAA World CO2 annual estimates (black circles). Parry Sound differs from SW Georgian Bay in the mid-50s to mid-70s but is otherwise similar.

Historical Parry Sound air temperature records are available on the Environment and Natural Resources Canada website for the periods 1875-1976, 1979-1991, and 2002-2019 (the last being at the Canadian Coast Guard (CCG)). Similar to Owen Sound and Toronto, Parry Sound records global and regional events such as the Krakatau volcano eruption, the 1916-1917 drop of 2.1°C, 1921 high, etc. (Jowett 2019).

To compare directly with Wiarton, Ontario, a starting date of 1948 and hinge point of 1980-1981 are used. Average temperatures for full years only are plotted in Figure 2a with a polynomial curve fit. (A month of data in the CCG record is missing in 2005 so this year was excluded.) A linear trend shows a -0.33°C/decade cooling from 1947 to 1980 and a +0.16°C/decade warming from 1981 to 2006 (with a 13-year gap in the record). CCG 12-month rolling averages show cooling of -0.12°C/decade for 2008-2019 and -0.31°C/decade for 2010-2019.

Figure 2a. Annual air temperatures of Parry Sound and CCG (after 2002) stations since 1948, with 3-order polynomial trend line, show general temperature cooling to 1980 and warming to present day, but with gaps in the data.

The cooling-warming-cooling cyclicity in Figure 2a over the past 70 years is very noticeable. A straight-line trend through this data would provide little insight into actual data trends, and would provide incorrect extrapolation into the future. As an example, a person in 1970 Parry Sound might reasonably assume that a linear forecast to 1990 could average a very cool 3°C. But since temperatures didn’t follow a linear trend, it was a warm 6°C instead. Similarly, forecasting linearly through the 1970-2005 warming segment would also likely be wrong, easily by 1.5°C or more.

The RCS weather station on the west side of Georgian Bay in Tobermory provides information from 2007 to present day, with gaps in the earlier records. Cooling of -0.81°C/decade for 2010-2018 is calculated using annual averages (not shown), and cooling of -0.29°C/decade is calculated from 12-month rolling averages through the July 2010 to June 2019 time period (Figure 2b).

Figure 2b. Tobermory temperatures in 12-month rolling average format from July 2010 to June 2019 indicating a -0.29°C/decade cooling trend.

Wiarton A station located in the southwest part of Georgian Bay has operated continuously since 1947 and shows a general cooling from 1948 to 1980, with a general “noisier” warming period since 1981 (Figure 2c). The major drop at 1991-1992 coincides with the Pinatubo eruption, and the major jump at 1997-1998 coincides with an El Nino event. The polynomial curve fit mimics Parry Sound and Owen Sound to a good degree.

A linear trend through the same data shows a -0.44°C/decade cooling from 1948 to 1980, with +0.24°C/decade warming to 2018. However, with cyclical rhythms, linear trends should only be used to interpolate general trends, but not to extrapolate or forecast, and the actual data show cooling at -0.26°C/decade since 2005.

Figure 2c. Annual air temperatures of Wiarton A station since 1948, with 3-order polynomial trend line, showing general cooling to 1980, then warming, followed by levelling off to cooling after 2000-2005.

When monthly average temperatures are averaged on a rolling basis over a 12-month period, the trends are quite striking (Figure 2d). The large fluctuations of 2012-2016 coincide with major cooling La Nina and warming El Nino events, and the 1997-1998 El Nino warming event is very evident in Wiarton and Owen Sound records (Figures 2c, 2h).

Figure 2d. Twelve-month rolling averages of monthly air temperatures for Wiarton A (red circles) and Parry Sound CCG (blue circles), with July 2005 to June 2019 start and end points. Internal linear trends show distinct cooling at rates of -0.49°C/decade for Wiarton and -0.44°C/decade for Parry Sound.

Rolling average trends of air temperatures since 2005 show cooling for Wiarton and Parry Sound at -0.49°C/decade and -0.44°C/decade respectively. Two months in the 2005 Parry Sound record were infilled using polynomial curve fitting.

Representing the southeast part of Georgian Bay, the Town of Midland’s records start at 1974 and end in 2017 with a large gap in the 1980-1990s. The polynomial curve fit in Figure 2e is similar to Figures 2a, 2c for Parry Sound and Wiarton, but no data is available after early 2017. A linear trend for September 2004 to August 2016 shows cooling of -0.08°C/decade for Midland.

A statement in the GBBR website of Air temperature is up an average of 1.9°C since 1974 (Midland station) is misleading as Midland had cooled first to 1981 before warming, and now is cooling again (Figure 2e), as is Georgian Bay in general. A very different picture is revealed when the first and last 24 months (June to May) are used as start-end points. Between 1974-1976 and 2014-2016, temperatures cooled from 7.36°C to 7.21°C, a rate of -0.04°C/decade. When using the first and last three years, temperatures warmed only +0.08°C in that 40-year period, or +0.02°C/decade, and the last years of operation saw cooling from 2004 to 2016.

Figure 2e. Twelve-month rolling averages of monthly air temperatures for Midland for 1974 to 2016 with 3-order polynomial line.

The weather station at Honey Harbour – Beausoleil, which is 15 km north of Midland, ceased operation in 2007. However, annual averages show warming at +0.67°C/decade from 1975 to 1990, and cooling at -0.47°C/decade from 1998 to 2006, similar to Midland. Several missing months of data were infilled using a 6-order polynomial curve fit of the data before and after so that entire years could be used.

Figure 2f. Annual averages of monthly air temperatures for Honey Harbour – Beausoleil for 1975 to 2006 with 3-order polynomial line.

In the southeast sector of Georgian Bay, 35 km SW of Parry Sound and 45 km NW of Honey Harbour, Western Islands station experienced rapid warming of +5.9°C/decade in 1995-2002, and after a long gap in the record, very rapid cooling of -6°C to -21°C/decade in 2015-2019, depending on the time variables used. A linear trend through the 12-month rolling average temperatures in the 2015-2019 era shows a -9.1°C/decade cooling (Figure 2g). These large warming-cooling rates are anomalous and likely not sustainable, but do corroborate the present cooling in southeastern Georgian Bay.

The distinct contrast of temperatures with CO2 values measured by Environment and Climate Change Canada (ECCC) at Egbert Ontario just 110 km away show the complete lack of correlation between the two variables (Figures 1a and 2g). Actual CO2 values in Figure 2g are overall yearly averages of non-filtered hourly data and range from 408 ppm in 2015 to 415 ppm in 2019. They were adapted to fit the temperature scale with the relative trends preserved. CO2 is going up and temperature is going down with no apparent connection between them.

Figure 2g. Twelve-month rolling averages of monthly air temperatures for Western Islands, Georgian Bay (blue circles). Linear trend shows a rapid cooling rate of -8.9°C/decade (July 2015 to June 2019), with -6.7°C/decade using annual averages for 2016-2019 (not shown). This cooling trend is contrasted with the upward trend of ECCC CO2 values from nearby Egbert Ontario (black circles).

Owen Sound in the southwest part the Bay has operated intermittently since the 1800s with location changes in 1965 and 2007. The 12-month rolling averages for each month are shown in Figure 2h, with a cooling-warming-cooling cyclicity similar to Parry Sound and Wiarton. A linear trend from April 2005 to March 2019 (not shown) yields a cooling trend of -0.32°C/decade.

Figure 2h. Owen Sound temperatures in 12-month rolling average format with 3-order polynomial fit outlining the cooling-warming-cooling trend similar to Wiarton and Parry Sound.

At the south end of the Bay, Collingwood has a large data gap from mid-1970s to mid-1990s, but has been cooling since 1997 (Figure 2i). Annual averages were calculated treating random missing data as blanks (blue circles) and also with adjacent averages inserted into missing data dates (red dots). The latter likely provides a better representation, but both show distinct cooling over the 20-year period at -0.24°C/decade and -0.28°C/decade.

Figure 2i. Collingwood annual temperatures show cooling since 1997. Blue circles treat missing data as blanks and cool at -0.24°C/decade. Red dots have adjacent averages inserted into missing dates and cool at -0.28°C/decade.

Historical air temperatures all around Georgian Bay (Wiarton, Owen Sound, Collingwood, Midland, Honey Harbour – Beausoleil, Western Islands, Parry Sound, Killarney, Tobermory and Kincardine nearby on Lake Huron) follow similar stable or cooling since 2000-2010, with Sandfield on Manitoulin Island missing 2018-2019 and being inconclusive. “Surprisingly rapid rates of warming” leading to catastrophic events are not in the picture, not now.

Water temperatures in Georgian Bay

Data records for Georgian Bay water temperatures started in 1989 and 1997 at moored Buoys C45137 and C45143 during the general 1980-2005 warming period. Hourly temperatures of the surface water (SSTP) are recorded throughout the year, but most winter months have weeks or months of data missing. November and May have reasonable data records in many years, but are erratic enough that real and aberrant temperatures could not be differentiated. The 5-month “year” of June 1 to October 31 is used. Source data sets are from

North Georgian Bay: Buoy C45137 (Latitude=45.54°N; Longitude=81.02°W) is moored 85 km WNW of Parry Sound and 55 km NE of Tobermory, and has been operated by ECCC since October 1989. Hourly water temperatures at C45137 were averaged for each 5-month year and graphed.

Some years had insufficient data and were not used. 2005 had 10 days of early June data estimated by interpolation and was used. The polynomial trend line in Figure 3a indicates water warming for 1991-2004 followed by a levelling off and cooling at -0.02°C/decade for 2004-2019 (Figure 3b) and at -0.37°C/decade since 2010 (not shown).

Figure 3a. Northern Georgian Bay Buoy C45137 surface water temperatures averaged over June 1 to October 31 for each year. The third-order polynomial curve indicates temperatures levelling off after 2004-2005.
Figure 3b. Linear trend line shows northern Georgian Bay water is cooling slightly at -0.02°C/decade, or virtually no change in water temperature after 2004.

South Georgian Bay: Buoy C45143 is moored 65 km southwest of Parry Sound and 45 km northeast of Wiarton (Latitude=44.94°N; Longitude=80.63°W), and has been operated by ECCC since August 1997. Hourly temperatures from June 1 to October 31 of each year were graphed with start-end points on the same day in August to minimize bias in a linear trend line. The years 1999 and 2002 had too much data missing to be used. The Bay between Wiarton and Parry Sound has been cooling at -0.12°C/decade from 1997 to 2005 and at -0.10°C/decade from 2005 to 2019 (Figure 4a).

Figure 4a. Hourly Georgian Bay Buoy C45143 surface water temperatures of June 1 to October 31 since August 2005. The trend line indicates a cooling of -0.10°C/decade.

The hourly water temperatures depicted in Figure 4a were averaged for each year as an alternative trend calculation (Figure 4b). Most years had 3000-3600 data points of the potential 3672 values. Those with missing data distributed evenly over the five months were considered representative. Years 1997, 1999 and 2002 had too few values and were excluded.

Although 2005 had two weeks data missing in early June, a sensitivity test showed that the true temperature was likely only 0.1°C lower than the measured, and so 2005 was included. The 2014-2016 warming jump as seen in Figures 3b and 4b coincides with an El Nino event similar to the 1997-1998 jump.

The linear trend shows substantial cooling at -0.28°C/decade for 2004 to 2019 (Figure 4b), and of even greater cooling at -0.75°C/decade for 2010-2019 (not shown and with only N=10 values). These results are similar to the trends for northern Georgian Bay (Figures 3a,b) but are cooling at a greater rate.

Figure 4b. Five-month “annual” temperatures averaged for June 1 to October 31 at Buoy C45143 for surface water. The linear trend line indicates cooling at -0.28°C/decade for 2004-2019.

Average water temperatures for the individual months used were calculated for Buoy C45143. From 2005 to 2019, the months of August and September warmed at +0.03°C and +0.56°C/decade, respectively, October stayed flat at +0.003°C/decade, July cooled at -0.68°C/decade, and June (using average of 2004 and 2006 as start point) cooled substantially at about -1.5°C/decade. Over the last 15 years, Georgian Bay water has been cooling in the late spring and early summer, especially in June, and warming in the early autumn, staying constant in October, with an overall cooling trend.

Extreme air temperature trends

Extreme air temperatures at Wiarton A for the same time period as nearby Buoy C45143 were analysed for potential insight. The stations normally record the single hottest and the single coldest temperatures for each month as the “extreme high” and “extreme low” for the month, and these 12 values were averaged for each year (Figures 5a, 5b).

From 1997 to 2019, overall cooling trends are -0.42°C/decade for extreme highs (Figure 5a) and -0.50°C/decade for extreme lows. The 1997 start date coincides with the start up date of southern Georgian Bay Buoy C45143, and being before the 1998 El Nino jump, it is not an artificial starting point. For 2000-2019 (Figure 5b), cooling is similar with -0.42°C/decade for extreme highs and of -0.48°C/decade for extreme lows.

Figure 5a. “Extreme high” air temperatures for Wiarton A from 2000 to 2019 showing a cooling trend of -0.42°C/decade.
Figure 5b. “Extreme low” air temperatures for Wiarton A from 2000 to 2019 showing an overall cooling trend of -0.48°C/decade.

An analysis of the available data shows that from the late 1940s, Parry Sound, Wiarton and Owen Sound all follow a cycle of cooling to about 1980, warming to mid-2000s, and then cooling over the past 15 years. Collingwood has cooled at a rate of -0.28°C/decade since 1997. Tobermory has been cooling since 2010. Since 2005, cooling trends have been -0.51°C/decade for Wiarton, -0.44°C/decade for Parry Sound, -0.39°C/decade for Owen Sound, -0.43°C/decade for Kincardine, and -0.31°C/decade for Killarney. Sandfield shows warming or cooling depending on the method used. Western Islands outside Parry Sound have cooled rapidly over the last 4-5 years.

In the northern sector, Buoy C45137 surface water warmed from 1991 to 2003, then levelled off and cooled at -0.37°C/decade for 2010-2019. In the southern sector, Buoy C45143 surface water has been cooling down at -0.28°C/decade for 2004-2019. Since 2005, Buoy C45143 water has been cooling in late spring and early summer, especially in June, warming in early autumn, and staying constant in October. “Extreme high” and “extreme low” air temperatures from nearby Wiarton show steady cooling of -0.42°C/decade and -0.50°C/decade from 1997 to 2019.

Air and water temperatures since 2005 have been steady or cooling; there is no sign of warming over the last 15 years. Forecasts are especially difficult for natural rhythms, and critically flawed when linear trends are extrapolated through non-linear cyclical data.

The stark juxtaposition of rising CO2 and dropping temperatures all around Georgian Bay causes concern as to the validity of GBBR’s dire forecast of Georgian Bay waters rising 10°C in the next century. The Canadian government goes to great lengths to supply accurate temperature data from around the country, for public safety in sea and air transport as well as climate forecasting and investigation. If we accept that our public service scientists are basically honest, the data they provide can be accepted as factual.

Craig Jowett Ph.D., P.Eng. did his Masters and Ph.D. at Toronto between working stints as a field geologist from coast to coast in Canada and the Arctic Islands, Eastern & Western Europe, and SW USA, followed by NATO Science Fellowships at Michigan and Cornell. After a time with University of Waterloo as a research professor, he founded his own Ontario-based wastewater manufacturing and research company. His research contributions can be found at ResearchGate.


Please enter your comment!
Please enter your name here