Details of Data

The search for the ice data presented here cannot pretend to be exhaustive; there are far too many potential sources, such as ship's logs, sailor's diaries, books, newspapers, etc., that it is doubtful if that boast will ever be made, but a sufficient amount of material was found to permit a reasonable assumption of what one might expect ice conditions to be at a given time. All records are referenced and the sources can be found in the references. While the IIP data has been plotted, the records themselves are not reproduced here due to the abundance of material, so that the reader is referred to the appropriate IIP Bulletins. As explained, after about 1923 there was little need to search further documents, hence there is little relevant text to go with the maps shown here since it is all contained in the bulletins, the only additional text included here being what was found during incidental reading.

The study was primarily concerned with the areal extent of the pack ice for the winter months. However, the search through the various sources being fairly laborious, all mention of ice in the North Atlantic, including icebergs, was noted. For the earlier years, when any mention of ice was quite scarce, ice reports from outside the months January to April were also noted. It is also worth mentioning that while the month of March was the focus of the study, since that is usually the month in which the maximum ice extent occurs, it was also important to enlarge the window of observation to get as accurate a feel as possible for that year. There is also the apparent conundrum that occasionally in a month of unusual ice extent and severity there are very few ice reports. This is because in these instances it is well known that severe ice conditions exist and ships have completely avoided the area. These cases are often indicated by numerous ice sightings in January and February showing larger than usual ice extents, followed by unusually few reports in March. Further reports from April then help in the characterisation. Since ice reports from January were not common being early in the year, any January information is included with the February data and one map has been drawn for the combined months.

A large number of typographical errors were noted in the published records in which geographical co-ordinates were obviously incorrect. These have been denoted by [!] and where records have been illegible, as was found occasionally in microfilms, by [?]. In many instances educated guesses could be made, otherwise these entries were ignored. An individual sighting of sea or pack ice is plotted as a red cross. Where sightings of pack ice included more than one position, such as following the edge of the pack or traversing through the pack from one position to another, a red line has been drawn between the recorded positions. Individual sightings of icebergs have been plotted as green triangles and similarly if a number of bergs were sighted along a route, the traverse has been indicated with a green line. Modern terminology is fairly consistent but a century and a half ago there was not the same consistency and sometimes it is not exactly clear what kind of ice was meant. Icebergs then were known as ice islands and when pieces of ice were mentioned they could be either icebergs or pieces of pack ice. When in doubt, intuitive guesses were made and it is hoped that when plotted any anomalous sightings will be apparent. The red boundary line shows the expected maximum average ice extent for that month based on the information available for that year and how the ice normally shapes itself in years of similar extents.