The extent of summer sea ice in the Canadian Arctic during the 19th century was remarkably similar to present ice conditions
The widely perceived failure of 19th-century expeditions to find and transit the Northwest Passage in the Canadian Arctic is often attributed to extraordinary cold climatic conditions associated with the “Little Ice Age”evident in proxy records. However, examination of 44 explorers’ logs for the western Arctic from 1818 to 1910 reveals that climate indicators such as navigability, the distribution and thickness of annual sea ice, monthly surface air temperature, and the onset of melt and freeze were within the present range of variability.
The quest for the Northwest Passage through the Canadian archipelago during the 19th century is frequently seen as a vain and tragic failure. Polar exploration during the Victorian era seems to us today to have been a costly exercise in heroic futility, which in many respects it was.This perspective has been reinforced since the 1970s, when paleoclimate reconstructions based on Arctic ice core stratigraphy appeared to confirm the existence of exceptionally cold conditions consistent with the period glaciologists had termed the “Little Ice Age”(Figure 1a), with temperatures more than one standard deviation colder relative to an early 20th-century mean [Koerner, 1977; Koerner and Fisher, 1990; Overpeck et al., 1998]. In recent years, the view of the Little Ice Age as a synchronous worldwide and prolonged cold epoch that ended with modern warming has been questioned [Bradley and Jones, 1993;Jones and Briffa, 2001; Ogilvie, 2001].
This article demonstrates the use of historical instrument and descriptive records to assess the hypothesis that environmental conditions observed by 19th-century explorers in the Canadian archipelago were consistent with a Little Ice Age as evident in proxy records. We find little evidence for extreme cold conditions.
Comparison of Four Climate Indicators
While there is an extensive literature on the history of 19th-century Arctic exploration, surprisingly little use has been made of the detailed scientific and meteorological observations compiled during many expeditions. There were more than seventy expeditions or scientific enterprises of various types dispatched to the Canadian Arctic in the period between 1818 and 1910.From this number,we analyzed 44 original scientific reports and related narratives; many from expeditions spanning several years.The majority of the data come from large naval expeditions that wintered over in the Arctic and had the capacity to support an intensive scientific effort.A table listing the expeditions and data types is located at www.pmel.noaa.gov/arctic/history.The data cover about one-third of the possible number of years depending on data type, and every decade is represented.
Our analysis focuses on four indicators of climatic change:summer sea ice extent, annual sea ice thickness, monthly mean temperature, and the onset of melt and freeze as estimated from daily mean temperature. Historical observations in these four categories were compared with modern reference data; the reference period varied, depending on data availability. Both sea ice extent and the onset of melt and freeze were compared to the 30- year reference period 1971–2000; monthly means are compared to the 50-year period 1951–2000. Modern sea ice thickness records are less continuous, and some terminate in the 1980s; the reference period is therefore based on 19 to 26 years of homogeneous record.
In common with most forms of visual and instrument data, questions regarding the accuracy and reliability of historical observations arise. For example, the measurement of extremely cold winter temperatures was problematic throughout the 19th century.Thermometers were particularly liable to inaccuracy below the freezing point of mercury (-38.8°C). Air temperature measurements in summer are also more likely to reflect cool values, as temperature measurements were generally suspended as soon as ice conditions were favorable enough to permit ships to sail.While a particular historical data series is subject to unknown error, the overall validity of observations is supported by concurrent, independent lines of evidence, which include measurements recorded simultaneously at nearby locations, descriptions of related physical processes,and occasionally,drawings or photos.
The extent of summer sea ice during the 19th century, insofar as it is shown in patterns of navigability inferred from ship tracks, the direct observations of explorers,and a number of native accounts, is remarkably similar to present ice climatology.A chart of northern Canada (Figure 2) shows the routes followed by discovery expeditions and their wintering locations between 1818 and 1859, and also displays the frequency that sea ice has occurred during the recent 30-year reference period 1971–2000. It is perhaps surprising that most of the Northwest Passage was navigated during the 19th century,with expedition ships coming within 150 km of completing the passage on a number of occasions. Most significant is that even in years that were recognized as unfavorable at the time, ships were still able to reach locations that would be consistent with the worst ice conditions that have occurred during the modern reference period. Of 33 expedition or supply ships bound for the western part of Lancaster Sound between 1819 and 1859, only two failed due to unfavorable ice conditions.
Eighteen measurements of maximum annual sea ice thickness were recorded between 1819 and 1876; four of these are incomplete, as they were interrupted before the end of the ice season.The remaining 14 observations had a mean of 215 cm and a maximum of 250 cm. Ice conditions for Resolute Bay, Sachs Harbor, and Mould Bay over 19, 23, and 26 years in the modern era give a mean maximum annual thickness of 199 cm and a mean of the absolute maximum at the three stations of 248 cm. Thus,the historical observations are consistent with 20th-century values.