WunderBlog Archive » Weather Extremes

Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.

World’s Hottest Nights/Highest Minimum Temperatures Yet Measured

By: Christopher C. Burt, 9:31 PM GMT on November 21, 2016

World’s Hottest Nights/Highest Minimum Temperatures Yet Measured

One of the interesting facets of global warming has been how, over the past several decades, nighttime minimum temperatures have become warmer relative to daytime maximum temperatures. There are several scientific explanations for this. A recent study (published in the International Journal of Climatology in February 2016 by Richard Davy et al at the Bjerknes Centre for Climate Research posits that the primary reason is that “night-time temperatures are inherently more sensitive to climate forcing”. See for more about this hypothesis and this for the full article. The focus of this blog, however, is simply an attempt to correlate a list of the extremes for record maximum-minimum temperatures from available data.

The blog consists of the figures of record daily maximum-minimum temperatures observed in the world and particularly the U.S. I should note, of course, that there is a difference between the ‘hottest night’ and ’the ‘warmest daily minimum’ temperatures since the temperature may, during the evening fall lower than the previous morning’s low temperature. It is difficult to determine just what the ‘hottest night’ on record may be for any given location. Temperature data is formatted and published to measure ‘calendar days’ rather than ‘24-hour periods’ since Excel files can’t handle more than one figure per column (at least that is the reason I’m given by those I’ve asked about this). That story is for a future blog. The statistics provided in this blog are for ‘daily record maximum-minimum temperatures’.

World Records for Highest Daily Minimum Temperature

Global average temperature trends over the past 50 years as researched by Richard Davy et al in the recent study (published in the ‘International Journal of Climatology’ in February 2016. For the northern hemisphere the most pronounced effect was during the winter months: +0.18°K whereas the summer months saw about a 0.11°K rise. Graph from ‘International Journal of Climatology’ February 2016.


It would appear that two sites in the world more or less tie for the warmest daily minimum temperature yet measured: Death Valley in the U.S. with 107°F (41.7°C) on July 12, 2012 and 41.7°C (107.1°F) at Khassab Airport, Oman on June 27, 2011. Since the conversions between F° and C° can make a difference of 0.1°F it is a statistical tie between the two sites. It should also be noted that since the NWS in the U.S. rounds down minimum temperatures and up maximum temperatures to the nearest 1°F, the actual temperature measured in Death Valley could have been anywhere from 106.6°F to 107.5°F (41.4°-41.9°C).

Below is table of the highest daily minimum temperatures measured by continent:



Record maximum-minimum temperatures so far observed by continent. This information is courtesy of Maximilliano Herrera.


Aside from the U.S. and Australia, it is difficult to pin down max-min temperature records for other individual nations or cities. Most national weather service web sites do not provide this information. From gleaning information I’ve found on the internet (words of caution there!) here are a few contenders.

United Kingdom: 23.9°C (75.0°F) at Brighton on August 4, 1990.
Netherlands: 24.1°C (75.4°F) at Limburg on July 2, 2015.
Germany: 26.7°C (80.1°F) at Weinbeit bei Neustadt on Aug. 7, 2003
Greece: 35.8°C (96.4°F) at Paleochora on June 27, 2007

It is frustrating that I am unable to find what Canada’s record for this might be. Toronto’s record is 26.6°C (79.9°F) on July 21, 2011, but it is likely other locations (probably many) have seen warmer lows elsewhere in the country, especially in southern Saskatchewan and Manitoba during the intense summer heat waves of the 1930s (Bismarck, North Dakota had a minimum of 83°F/28.3°C in July 1936). Hopefully, one of the readers of this blog knows the answer.

The national weather service for France, Meteo France, maintains an excellent web site full of weather-record information but not so far as record high minimums. These most likely would have occurred during the deadly heat wave of August 2003 (see national record for Germany above). Apparently, Paris endured its hottest night ever on August 12, 2003 when the temperature failed to fall below 25.5°C (77.9°F), but just what the national record for France might be escapes me. Perhaps Jerome or Max can answer this.

U.S.A Records for Highest Daily Minimum Temperature

The trend in the U.S. also reflects (to a smaller degree) the global increase in daily minimum temperatures versus daily maximum temperatures over time, as reported in Davy’s article.






The top graph(s) shows how on an annual basis (in this case the 12-month period from November-October, 2016) the maximum temperatures have been rising by an average of +0.13°F per decade whereas the minimum temperatures by +0.15°F per decade. In the bottom graph(s) we see the temperature changes, since 1901, broken down into winter and summer seasons. In these cases there is a pronounced bias toward warmer summer temperature minimums relative to warmer summer temperature maximums: +.14°F per decade for minimums versus +.07°F for maximums, in other words the summer nights are warming twice as fast as the summer days in the U.S. The winter months show very little difference in this regard. Note that this does not correlate with the Davy report. However, the Davy report was for hemispheric observations and with a 50-year POR baseline. The NOAA report has a 100-year POR baseline. So the two graph’s are not comparable all things being equal. Graphs from NOAA: Climate at a Glance.


Researching this subject for individual sites in the U.S. is relatively easy using the NCDC’s ‘NOW Data’ available for thousands of weather stations (including COOPs) across the country. The biggest problem is that there are some quality control problems with this database; some obvious.

Nowhere in the country (and perhaps the world) has nights as hot as Death Valley where, on a few days some summers, the temperature fails to fall below 100°F. The only other 100°F+ minimum site I can find is Bullhead City, Arizona which claims a low of 102F° on July 26, 2006 (this according to NOW data, WU has the 102° figure occurring on July 22nd). Death Valley’s world-tying low of 107°F on July 12, 2012 does not include a pair of very dubious reports from Greenland Ranch in August 1924 of lows of 110°F (43.3°C) on Aug. 25 and 26 and again on July 5, 1918 (equally dubious). Looking at the COOP form for August 1924 one can see a number of irregularities: extremely small diurnal temperature spreads of only 6° on Aug. 19th, a twelve-day continuous stretch of minimums of 100° (and 110°). In the modern history (since 1934) of Death Valley’s weather record the longest stretch of consecutive days above 100°F has only been three.



The COOP form for Greenland Ranch in Death Valley for August 1924. Two nights reportedly had minimum temperatures of 110°F (43.3°C). However, a close examination of this form shows a number of irregularities. The two 110°-figures cannot be considered genuine. Similar scrutiny of the COOP form for July 1918, when another minimum of 110° was reported, is also suspicious.


Outside of the desert Southwest the highest daily minimum temperatures I’ve found so far were 92°F (33.3°C) at Leavenworth, Kansas on Aug. 10, 1934 and 91°F (32.8°C) at Lincoln, Nebraska on July 25, 1936.



One of the hottest nights ever endured in the U.S. outside of the desert Southwest occurred on the night of July 25, 1936 at the unlikely location of Lincoln, Nebraska, when the temperature fell to only 91°F (32.8°C). The citizens of the city spent the night outdoors trying to sleep on the lawn of the state capitol. Photo from the archives of the Nebraska State Historical Society.


On the other side of the coin, the U.S. site (including Alaska) with the record coolest highest minimum summer temperature on record is 51°F (10.6°C) at Leadville, Colorado on July 20, 2010. Leadville is located at an elevation of 10,152’ (3094 meters) and is the highest incorporated city in the entire country. It should be noted, however, that the POR (period of record) for Leadville only goes back to 1976. Eureka, California has the coolest record daily max-min for any first-order site with a long POR: a 63°F reading last set on January 18, 1981! There were two other occasions with a daily record max-min of 63°: Feb. 26, 1980 and Aug. 27, 1894. In Alaska, Barrow holds this record (for a first order site) with a 53°F record (see table below)

Below is a table of the highest minimum temperatures on record for most major cities in the U.S., arranged by both temperature and date of occurrence. Most of these sites have POR’s dating back to the 19th century. Although one can see a plethora of post-2000 records. However, many of these were tied with earlier dates and, in addition, this is too small a sample of sites to draw any conclusions from, although one could argue there is a trend here. The summer of 2016 was a great example of how record high minima made all the difference so far as the summer’s ranking as the fifth hottest on record (it was the warmest summer on record so far as minimum temperatures were concerned. See this blog posted by Bob Henson on WU in September for more details.




NOTES ABOUT THE TABLE ABOVE: NOW data for Anchorage, Alaska only goes back to 1953, thus I did not include this significant site on the table. Since 1953 the max/min temperature record for Anchorage is 63° on July 10, 2005.

A minimum of 76° on June 26, 1915 reported in NOW data for Fairbanks, Alaska is suspicious.

Denver, Colorado NOW data has two highly suspicious minimums of 77° and 76° both set in the summer of 1881. No other year since then (or before) has seen a daily minimum warmer than the 73° reading in 2003.

Indianapolis has two suspicious minimums of 85° and 84° both reported during the summer of 1874.

Hopefully, some readers of this blog can suggest new data (especially for non-U.S. sites) to build on this initial effort.

KUDOS: Thanks to Mark Stroud of Moon Street Cartography for the table production above, and Maximilliano Herrera for the continental highest minimum temperatures.

Christopher C. Burt

Weather Historian

Heat

The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.

Reader Comments

Canada highest min. temp. is 32.2C (90F). in July 1937, same episode of the 45.0C (113F) max.
Although the July 1936 heat wave was much longer in Central/Eastern Canada, in July 1937 a 32C isotherm at 850hpa briefly entered Alberta and Saskatchewan with insane temperatures and Mojave-like extremes of 32.2-45.0 . Note that Fahreneit were used at that time in the Maple Republic.
The downtown Charleston SC station has reported minima as high as 86°F. KDMH in Baltimore Inner Harbor reported a low of 88°F in July 2011 but it is unofficial.
Glad you pointed this out Brian. The Charleston weather record is actually one of the oldest in existence in the U.S. yet NOW data only goes back to 1938 for the city. A perfect example of the problem with the NOW data set. For one thing, it does not include the downtown Charleston records, ONLY Charleston Airport (which had 81 on Aug. 1, 1999 when downtown measured the 86 you refer to). In yet, the NOW data set is supposed to be threaded for ALL official sites in a city when they say, for instance "Charleston Area". Not in this case, obviously.

Why does NOW data not include any weather records for the city prior to 1938? Smithsonian records for Charleston begin 1738..

Quoting 2. BaltimoreBrian:

The downtown Charleston SC station has reported minima as high as 86%uFFFDF. KDMH in Baltimore Inner Harbor reported a low of 88%uFFFDF in July 2011 but it is unofficial.
Christopher, with just a cursory review, I did find that the Needles, CA airport recorded a Minimum Temperature of 100° on July 22, 2006 (source: NCEI Climatological Data publication 7/2006). If I find any more, I'll repost.
You are correct about Needles Belmont Jeff, It is in the NOW data set as well. I thought I checked all the potential sites in the NOW data set for 100F min-max's (including Needles) but I must have 'glazed' over that one. That kind of happens when you spend weeks looking at thousands of figures of data. I'll bet I missed some other contenders as well!...and, if you find others (with long PORs), please let us know.

Thanks!

Quoting 4. BelmontJeff:

Christopher, with just a cursory review, I did find that the Needles, CA airport recorded a Minimum Temperature of 100%uFFFD on July 22, 2006 (source: NCEI Climatological Data publication 7/2006). If I find any more, I'll repost.


Here is a list of UK data, as supplied by the "Meteorological Office" of Great Britain.

Highest daily minimum temperature records

Country Temperature ( °C) Date Location
England 23.9 3 August 1990 Brighton (East Sussex)
Wales 22.2 29 July 1948 Swansea, Victoria Park (West Glamorgan)
Northern Ireland 20.6 31 July 1868 Armagh (County Armagh)
Scotland 20.5 2 August 1995 Creebridge (Wigtownshire)

According to ogimet.com on June 27, 2011 the daily minimum for Khasab Airport was 41.9 C. Is it a wrong data?
Quoting 7. Vesca:

According to ogimet.com on June 27, 2011 the daily minimum for Khasab Airport was 41.9 C. Is it a wrong data?


It depends on the standard used. We use statistics for 00:00 to 23:59 daily minima. Using different spans of time like 7am to 7am or 9am to 9am or other standards, the statistics are different and higher minima exist but they are spread between 2 days. on 26th Khassab dropped to 41.1C during the early night hours, while on 27th it dropped to 41.7C during the following evening at midnight local time,00:00, the first observation on the 28th (21H UTC time in summer is midnight in Oman). So it's true this synop belongs to the 28th and 41.9C should be considered the minimum temperature of the 27th, in absence of other synops with partial minimum temperatures recorded between the tri-hourly synops. 41.7C might have been the minimum temperature between the local 9pm synop and the midnight synop (very likely it was 41.7C 1 minute before midnight too) but at Khassab Airport there is only one transmission of minimum temperature. Unfortunately many countries don't use the 24 hours standard of minimum temperatures and the transmission is given is between the early hours of each day and it's something complicated to translate them into that format with only 1 daily tranmission of the minimum temperature.
Quoting 8. maxcrc:



It depends on the standard used. We use statistics for 00:00 to 23:59 daily minima. Using different spans of time like 7am to 7am or 9am to 9am or other standards, the statistics are different and higher minima exist but they are spread between 2 days. on 26th Khassab dropped to 41.1C during the early night hours, while on 27th it dropped to 41.7C during the following evening at midnight local time,00:00, the first observation on the 28th (21H UTC time in summer is midnight in Oman). So it's true this synop belongs to the 28th and 41.9C should be considered the minimum temperature of the 27th, in absence of other synops with partial minimum temperatures recorded between the tri-hourly synops. 41.7C might have been the minimum temperature between the local 9pm synop and the midnight synop (very likely it was 41.7C 1 minute before midnight too) but at Khassab Airport there is only one transmission of minimum temperature. Unfortunately many countries don't use the 24 hours standard of minimum temperatures and the transmission is given is between the early hours of each day and it's something complicated to translate them into that format with only 1 daily tranmission of the minimum temperature.


I see, thank you for your answer.
In France, the highest minimum temperature ever recorded was 30.3 °C in Menton, which lies between Nice and the italian border. That was during the extreme heat wave on August the 6th in 2003, during which the record for Paris was recorded.

https://donneespubliques.meteofrance.fr/donnees_l ibres/bulletins/BCM/200308.pdf

The maximum temperature for Menton for this day stood "only" at 35 °C.
http://www.meteopassion.com/almanach-meteo-du-6-a out.php
Ponca City, OK also had a minimum of 91F on July 26, 2011 -

Link

"More compelling were the daily lows, with 6,171 records set for the highest minimum temperatures for July. In Ponca City, Oklahoma, the overnight low on July 26 was 91 degrees Fahrenheit (33C). The pattern of warmer nights fits with a longer trend noted in a recent NOAA report."
Did anyone predict any of this? I doubt it. Real science is predictive.
Quoting 12. comitatus:

Did anyone predict any of this? I doubt it. Real science is predictive.


Predict what?
The highest daily minimum are mostly higher in large cities. These cities are large heat sinks with concrete and mortar. That's why results are more significant for the minimum, also at night wind speed usually drops. Would be interesting to see charts with max population of cities at 8,000. Just saying for Minnesota I see Duluth and Minneapolis not StJames or Chisolm.
Basrah international airport Iraq, on July 28,2016 recorded a high of 129 and recorded an over night low of 101.
The highest minimum temperature ever recorded in Slovakia was 27,3 C in Piestany on 29.7.2013.
http://nun.sk/
- in Hungary 27,9 C in Pecs on 20.07.2007
http://owww.met.hu/eghajlat/Magyarorszag/rekordok /
- in Czech Republic 26,9 C in Bystrice pod Hostynem on 29.8.1992
http://www.in-pocasi.cz/archiv/

In Slovakia and Czech Republic its actually daily minimum between 21-21 UTC.
Are there any data from the Danakil or Qattara depressions in Africa? I understand that political factors may be a problem in maintaining instruments at such sites. Not so much for the Dead Sea area--what sort of numbers come out of there? One also wonders about the Dasht-e-Lut region of Iran, due to the high humidities sometimes reported there, and likewise about inland locations of the Namib Desert, where humidity can also be high.

I grew up in St. Louis and know that morning lows in the 80-85 range are not unknown during major heat waves, although I cannot remember any such readings exceeding 85. On rare occasions the dew point has exceeded 80 degrees. It would not surprise me if similar numbers did in fact come from Indianapolis, which is only a bit farther north and east.
The highest minima for St. Louis are 86, on July 24, 1901 and July 25, 2012. The earliest in the year min of 80 or more was 80 on June 15, 2016 and the latest was 81 on September 15, 1931.

For Indianapolis IN the highest min was 85 on June 28, 1874. No other mins of 82 or more. Many high mins set in 1872, 1873 and 1874 which makes me think there could have been a measurement problem.

Source: http://threadex.rcc-acis.org/
19. japr
Hi can you tell me  ;about the whether in my  city dhamar yemen
Those people sleeping on the lawn of the state Capitol in Nebraska are not an unusual scenario for those times. My mom grew up in Chicago in the 1930's and it was very common for whole families to go down to the park and sleep on the grass at night in summer. In my mom's case, the park they went to was 3/4 of a mile from home. Amazingly, nobody get mugged or raped.

But why are the people close enough to see in the photo sleeping with their shoes on?
Quoting 17. Snowfire:

Are there any data from the Danakil or Qattara depressions in Africa? I understand that political factors may be a problem in maintaining instruments at such sites. Not so much for the Dead Sea area--what sort of numbers come out of there? One also wonders about the Dasht-e-Lut region of Iran, due to the high humidities sometimes reported there, and likewise about inland locations of the Namib Desert, where humidity can also be high.

I grew up in St. Louis and know that morning lows in the 80-85 range are not unknown during major heat waves, although I cannot remember any such readings exceeding 85. On rare occasions the dew point has exceeded 80 degrees. It would not surprise me if similar numbers did in fact come from Indianapolis, which is only a bit farther north and east.


Yes, there are several regional stations in the Dash e-Lut desert, its highest temperature so far is 52.5C in 1998. Keep in mind Dasht e-Lut minimum elevation is few hundreds feet above sea level.
Its lowest point might get as hot as 53 but not more than 54C.
Qattara depression is hardly capable to reach 50C. There are stations near the depression about at sea level and none of them of ever recorded more than 49C. It's all plain , not far from the Mediterranean and not near mountains, so it's weak point is not being favored by the topography.
I estimate the Highest temperature ever recorded in the Qattara depression is around 50.5C.

We had a station for a decade in the hottest point of the Danakil depression: Dallol.
Its temperatures are very stable in the summer months,indeed its yearly pattern, due to its topography and latitude, is very regular, there are no "heat waves" or "cold waves". Every year is almost a photocopy of the previous year.
Its highest temperature ever recorded in 48.9C (120F). Ethiopian Meteorological Service estimates highest possible temperature in the bottom of the Danakil and Lac Assal depressions (Djibouti) doesn't exceed 51C with today's climate.
Highest minima can be over 35C in July and August but I don't think they can exceed 40C.
Yearly average temperature is the highest on Earth, but I estimate a reliable yearly average to be around 32C without the effect of the hundreds of geysers which are present in the area of Dallol and in my opinion, has given a "aggregated value" of an average of 2C yearly.
To be sure we would need to install a station in Hamed Ela, the nearest village , which is now well connected with a paved road to Addis Abeba and reachable in a matter of hours and not days like few years ago. (thanks to the Chinese govt.).
I think a weather station in Hamed Ela would be extremely interesting and it would require only a very small investment by the WMO or if they have the resource, the Ethiopian Meteorological Institute.
And it continues

SYDNEY sweltered through its hottest December night on record as the mercury refused to budge to acceptable levels. The Bureau of Meteorology confirmed the overnight temperature dipped to 27.1C.
It was the highest minimum on record. The previous record was set on Christmas Day 1858, when it dropped to 26.3C.

Link
Something ain't right. There's too much heat in the night.

Oh NO3Z we're all gonna die in a burning earth fire!
I am currently residing in the Ft Lauderdale, Fl area for the winter and I keep a daily watch on the weather on the WU website. I have noticed that the morning lows have been running 5 to 10 to 20 degrees above average since November and the current 10 day forecast continues this trend. Also the coastal water temp here is 82 and directly west across the state in Naples it is 72. Can you give an explanation of why the morning lows have been so high recently? Thanks, JR
I'm just one of those guys who doesn't understand climate from weather. It's -15 down here in Southern Colorado. Please send us some global warming.

Yeah, don't waste your breath. ----- This country bumkin just doesn't understand.
Well, your data may be wrong. THe records this site keeps and reports and not the same as the actual temps measured in my area. I think you guys are frauds myself. It was 16 degrees at my house and my brothers house this morning. The official temp read 15.5 degrees on the web page this morning you guys used an official temp of 21. The old record was 19. Frauds.
It has been found that 90% of heat increases are from heat islands. Maybe you should take another look at where they are placed. I could change the ones in a 50 mile radius of me and lower the temps by 4-8 degrees depending on about 20 variables. I have a weather station and I watch it regularly. Th planet always warms between ice ages, this is why we are not in an ice age now, lol
"It has been found that 90% of heat increases are from heat islands". Oh brother. Almost certainly the most ignorant comment of the week. Heat islands do not account for warming oceans. I suggest reading articles in Climate News and Oceanography News from Science Daily. Avail yourself of them.
GIGO.. Garbage in.. Garbage out.. if your hypothesis is not supported by the data you change the hypothesis, not the data..

Investigation of methods for hydroclimatic data homogenization

We investigate the methods used for the adjustment of in homogeneities of temperature time series covering the last 100 years.

Homogenization practices used until today are mainly statistical, not well justified by experiments and are rarely supported by metadata. It can be argued that they often lead to false results: natural features of hydroclimatic time series are regarded errors and are adjusted.

While homogenization is expected to increase or decrease the existing multiyear trends in equal proportions, the fact is that in 2/3 of the cases the trends increased after homogenization.

The above results cast some doubts in the use of homogenization procedures and tend to indicate that the global temperature increase during the last century is between 0.4°C and 0.7°C,

E. Steirou, and D. Koutsoyiannis, Investigation of methods for hydroclimatic data homogenization, European Geosciences Union General Assembly 2012, Geophysical Research Abstracts, Vol. 14, Vienna, 956-1, doi:10.13140/RG.2.2.23854.31046, European Geosciences Union, 2012.

https://www.researchgate.net/publication/25861467 6_Investigation_of_methods_for_hydroclimatic_data_ homogenization
Quoting 28. Proveit:

It has been found that 90% of heat increases are from heat islands. Maybe you should take another look at where they are placed. I could change the ones in a 50 mile radius of me and lower the temps by 4-8 degrees depending on about 20 variables. I have a weather station and I watch it regularly. Th planet always warms between ice ages, this is why we are not in an ice age now, lol


The Arctic region is the fastest warming region on the planet. Just how much urban sprawl do you thing is going on up there?
Quoting 28. Proveit:

It has been found that 90% of heat increases are from heat islands. Maybe you should take another look at where they are placed. I could change the ones in a 50 mile radius of me and lower the temps by 4-8 degrees depending on about 20 variables. I have a weather station and I watch it regularly. Th planet always warms between ice ages, this is why we are not in an ice age now, lol


"I have a weather station and I watch it regularly". Okay...good for you. Do you have it properly sited, accounting for exposure to winds, buildings, direct sun? Do you want to go to Oak Ridge National Laboratory and advise the scientists there that do urban heat island research that their methods are wrong? You have nothing to back up your argument. You have a weather station, big woop. Local variability and siting differences can change the temperature. Heat islands are accounted for in climate understanding.

Quoting 26. Skeptic27729:

I'm just one of those guys who doesn't understand climate from weather. It's -15 down here in Southern Colorado. Please send us some global warming.

Yeah, don't waste your breath. ----- This country bumkin just doesn't understand.


Far easier to understand than taxes
weatherhistorian has created a new entry.
weatherhistorian has created a new entry.
"It is frustrating that I am unable to find what Canada’s record for this might be. Toronto’s record is 26.6°C (79.9°F) on July 21, 2011, but it is likely other locations (probably many) have seen warmer lows elsewhere in the country, especially in southern Saskatchewan and Manitoba during the intense summer heat waves of the 1930s (Bismarck, North Dakota had a minimum of 83°F/28.3°C in July 1936). Hopefully, one of the readers of this blog knows the answer."

Chris..

You can use try using the High Plains Regional Climate Center climate program (CLIMOD) http://climod.unl.edu/ to analyze climate data for Canadian sites. It can do a lot of climate data analysis including record high minimum temperatures for any climate site. Winnipeg Manitoba for example has climate data back to 1872 (using data from an older city station, not the airport) Like Bismarck, its record highest minimum temperature was 28.3C (83F) recorded during the great 1936 heat wave on July 12 1936. I suspect this extraordinary reading is the highest minimum temperature of any major site in Canada.