“As cold as ever I knew it.” Joseph Colen, February 22nd, 1802. The Changing Climate of Manitoba
by Dr. Tim Ball
MHS Transactions, Series 3, Number 33, 1976-77 season
An entry in the daily journal of the Hudson Bay Company at York Factory for the 20th of January 1787 reads, "For days past we have had the severest weather ever remembered by the oldest Native on the Plantation." The journal was kept by Joseph Colen the Chief Factor who was required to provide a daily synopsis of the weather, as well as the activities of the men, the success of the hunt and other information.
It is interesting to speculate on the value of these daily weather reports to the Governors of the Company. It does not seem possible that they could have been of anything more than passing interest. True, some of the Governors were also members of the Royal Society and therefore had more than a passing interest in any natural phenomena, however, these interests were usually pursued more actively. For example, Samuel Weggs is responsible for the maintenance by employees of several meteorological journals which were subsequently submitted to the Royal Society.  Other journals were apparently kept for personal interest. Peter Fidler's records were maintained fairly consistently over a period of thirty years and seem to be a result of his lifelong interest in the world around him. 
Regardless of the reason for these observations, journals, and accounts of the weather, they provide an excellent opportunity for reconstructing weather patterns of northern North America over a long period of time. In fact they are almost unique in the world with regard to the length of the observations at one location and also in the extent and detail of phenomena recorded. There are, of course, many difficulties to be overcome before the rigors of the scientific method can be satisfied, nonetheless it provides an excellent base for continued understanding of climatic change.
Before looking at the actual record it is essential to understand how and why the information is so important to an understanding of the historic climate. Three elements must be understood before the relevance of the data can be seen. These elements are: the significance of the area in which these observations were taken; the status of research in other areas of the world; and finally the application of the results to a better understanding of climatic fluctuations.
One of the ironies in Manitoba history is that the earliest European settlements occurred along the shores of Hudson Bay, an area that is struggling for its very existence today. As with any settlement this early pattern is a result of the mode of transportation and the natural highways that result. The practice of latitude sailing resulted in sailors from England proceeding to a point north and west of Scotland. From here they merely followed a specific latitude to the West until landfall was made in North America. There were two advantages to this procedure. Firstly, it provided for ease of navigation and lack of need to determine longitude. Secondly, because the latitude was well to the north of the Atlantic it meant that the full force of the North Atlantic Drift and Gulf Stream was avoided. The resulting landfall brought the vessels into Davis Straits and from there it was a natural passage into Hudson Straits and then to Hudson Bay.
One of the earliest forts was established by the Hudson's Bay Company at York Factory. It appears on Thornton's map of 1685 as Hayes Fort, however, the official name was York Fort. Following a brief occupation of the area by the French under the hands of Jeremie the fort was returned to the Hudson's Bay Company by the Treaty of Utrecht in 1713.  Governor Knight and his party arrived to commence trading activities in September of 1714. At this point the incredible sequence of daily journal entries begins and continues almost uninterrupted through the next 150 years.
The daily journal was a mandatory bookkeeping function of the factor or his deputy. Normally a rough copy was maintained which was subsequently transcribed into a final copy that was sent with the yearly packets to the Governors in London. Each day's entry was supposed to include various pieces of information, the most notable being, daily hunting successes, provisions provided by the Indians or Eskimos in trade, the deployment of the men, and the daily weather. Sometimes the journal was written by the Factor but more often it was transcribed by a writer.
The type of entry varies considerably, as would be expected, however there are surprising consistencies throughout. Variations occur in wording and emphasis to reflect the individual style and perspective of the Factor. They also occur to reflect the changing terminologies of the scientific world. Despite these variations the majority of the Factors were trained as navigators or surveyors and therefore were very conscious of the weather and viewed it with the expertise of these trainings. Even those factors who had not been formally trained as navigators had a very perceptive eye for the weather, much as a farmer whose livelihood and very existence make him a surprisingly reliable observer.
Most of the variability in the record can be overcome by statistical methods or at least the degree of variability can be determined. Some of the comments are of a qualitative nature and some are quantitative. These are interspersed with complete years where actual measurements are recorded in Meteorological journals. Instrumental records of this type create other problems that will be discussed later.
In order for data of a mixed nature to be used in a continuous and meaningful manner it has to be adapted or converted to a form that will create a homogeneous record. Only then can an attempt be made to study the significance of variations in the record as a reflection of actual climatic change. The method used in this case was to establish a numerical code for the various weather elements discussed in the journal that could then be punched onto computer cards for ease of storage and analysis. It is not the purpose of this paper to discuss the details of the methodology but rather to show how significant the Hudson Bay Company Journals are to an understanding of the climate coincident with Manitoba's European history. Also it is intended to add significantly to the world data base of weather fluctuations over the recent or historic past.
What is the significance of this information in building a world picture of the weather during the historic period? Systematic daily meteorological observations using instruments began in England shortly after the formation of the Royal Society in 1660. As a result Dr. Gordon Manley has been able to develop a sequence of monthly mean temperatures for central England commencing in 1659.  Subsequent work has been continued by Dr. H. H. Lamb who was responsible for the formation of the Climatic Research Unit at the University of Anglia. Lamb has reconstructed general circulation patterns and synoptic charts for various time periods back to 990 A.D.  This work is now being continued and intensified by J. A. Kingston also at East Anglia. 
Similar records of temperature variability have been reconstructed for Japan using religious records that denoted tree blossoms and lake freezing dates with regularity and accuracy.  These records are extremely valuable but rather isolated in relation to most work covering this period of time. At present there is no other record to provide even fragmentary correlation for Asia, Africa or South America. Some North American research is being carried out at the Climatic Research Unit under the auspices of Reid Bryson in Wisconsin. However, most of the historical studies, that is those studies that specifically use historical sources, have been carried out by David Ludlum. Much of his work can be read in the journal Weatherwise, the house organ of the American Meteorological Society. His most significant study to date is gathered in a book entitled Early American Winters; 1604-1820.  This is composed of verbal comments on the weather conditions in major regions of the United States gathered from a variety of sources including personal diaries, voyages of discovery (for example the Lewis and Clark expedition) as well as military records from various campaigns, most notably the Civil War.
Two major problems exist with this type of historical information. First, it is extremely varied and subjective. Second, it is very scarce both temporally and spatially. The Hudson Bay Company Journals offer an opportunity for continuous long term records that will fill many gaps. Beyond this they have an enormous benefit because they provide coverage of an area, northern Canada, that even today has very few weather stations. In fact it might be one of the ultimate ironies of the Canadian North that a more complete network of weather observations points existed through the eighteenth and nineteenth centuries than exist for the twentieth century.
The North American continent is extremely important in its influence on the climate of the northern hemisphere. It provides a land mass running from the Tropic of Cancer to a line several degrees of latitude north of the Arctic Circle. Within this zone can be found three of the five major zonal belts of world climate. These are the subtropical Highs that account for the hot deserts of the world; the zone of the Westerlies with its sequence of cyclones and anti-cyclones bringing weather so familiar to Western Europeans; and finally the Polar Easterlies of extreme northern latitudes. The zone of the Westerlies marks the area of major confrontation between the cold Arctic air and the warm tropical area along a line that is referred to as the Polar Front. In the upper atmosphere the Westerlies form waves that vary in frequency and amplitude and undulate along the polar front from west to east. They are recognized as a major component in the transfer of surplus heat energy from tropical latitudes to polar regions. Lamb has shown that the Rocky Mountains acts as a barrier to this Westerly flow of air creating a disturbance that results in almost permanent troughs being established on the downstream side of the mountain barrier.  The cyclones and anti-cyclones, that bring an increased variability of weather, are steered around these large semi-permanent troughs.
The purpose of the preceding paragraph is to develop the picture of the basic weather pattern and circulation that exists in North America. If any change in the frequency of cyclones and anti-cyclones occurs it is most likely to be a reflection of a change in the formation of the trough and the movement of the upper air. The connection between variability of weather at the Hudson Bay Company posts and the variability in the general circulation pattern can provide a further piece in the puzzle of climatic variability. In any scientific research data must be gathered to prove or disprove a theory. Up to now the amount and accuracy of data collected is very sparse.
The weather records of the Hudson Bay Company Journals take two forms. From a scientific point of view the most useful form seems to be those records that are based upon instrumental readings. Certainly if any comparison to present day climate is to be achieved actual measurements would appear to offer the best opportunity. Unfortunately things are complicated by the problem of variance in the quality of the instruments and the knowledge of the recorder. W. E. Middleton illustrates this problem.
The problem is still evidenced in 1771 in the opening remarks to Thomas Hutchins Meteorological Remarks recorded at York Factory.
Hutchin's was aware of the importance of the temperature readings being taken from a thermometer that was not exposed to the sun, and that the height above sea level of a barometer is extremely significant. Despite this awareness his scientific endeavours were severely limited by such distinctive local requirements exemplified by the following quotation:
How the instrument was standardized after the quicksilver was replaced is not recorded, thus throwing subsequent readings into question.
The daily journals show a greater variety of entries although a majority are as follows:
It is important to note that in most of these entries there are some weather elements that appear almost without fail. These are wind direction, wind strength, cloud cover, and the occurrence of precipitation. As a result it is possible to build a fairly accurate picture of the daily pattern of weather conditions. For most years the data is complete enough that interpolation of the few days on which data is missing is relatively easy. On a very broad scale it is a simple matter to determine which years had excessive snow or rainfalls, droughts, or were extremely cold, and as a result coarse patterns of weather can be determined.
The journal keepers assist greatly in this task by making note of unusual seasons. For example:
The last entry relies of course upon the accuracy of the writer's memory. Modern studies have shown that peoples' memories of weather events are surprisingly accurate.  The abundance of comments reflecting on previous years throughout the journals suggests that the Hudson Bay Company men were as capable. This accuracy is further proved by checking back through the journals in order to compare the accuracy of the statement with the event.
At the present time work is progressing on using all of this historical information to measure the variations in Manitoba's weather for the last 250 years. Preliminary information suggests that the periods from 1730 to 1740 and 1790 to 1820 were extremely bad climatic periods. The latter is extremely important for historians as it covers the period of arrival for the first settlers. It would appear that they could not have chosen a more inappropriate time as far as weather was concerned. Of course it could be argued that their time of arrival was not of their choosing but more dictated by land policies in their former country.
Several patterns are emerging. The first and most immediately apparent is that the weather sequences change from relatively little variation in temperature and precipitation extremes to periods of extreme fluctuation in the range of weather experienced. A graph of the first occurrence of snow in the fall suggests that from 1714 to approximately 1738 the date was quite variable, while from 1740 to 1800 the date was fairly constant. This pattern appears to correlate with European research suggesting that 1740 marks the end of the Little Ice Age and the onset of a period of amelioration, at least in the Northern Hemisphere.
In summary it can be stated, with little fear of contradition, that the Hudson Bay Company Journals provide one of the most complete and extensive sources of climatic data presently known. Hopefully it will serve to increase and extend our knowledge of climatic variability. Perhaps it will provide the key to understanding these variations and thus lead to the ultimate goal of science, namely prediction.
1. Ruggles, Richard. 1976 Governor Samuel Wegg: The 'Winds of Change!' Beaver Autumn 307:2 pp. 10-21.
7. (a) Arakawa, H. 1954. Five centuries of freezing dates of Lake Suwa in Central Japan, Arch. Met., Wien, B., Vol. 6, pp. 152-166. (b) Arakawa, H. 1956. Climatic change as revealed by the Blooming dates of the cherry blossoms at Kyoto, J. Met., Vol. 13, pp. 599-600.
Page revised: 22 May 2010