Computer analysis of an inconsistent data source: the case of probate inventories

Computers are becoming increasingly popular for analysing data in historical geography and are being applied to an extending range and variety of historical sources. The methods described in most computer texts for geographers and historians, using the langueage allied to punched cards, are only suitable for numeric data which are in a standardized and consistent form. Rather than attempting to apply these methods to sources not organized in this way, it is more sensible to tailor a technique to a particular class of data.This paper considers some advantages of computer analysis, using English probate inventories as an example. Limitations of conventional methods are discussed and a technique is described for dealing with sources whose information is inconsistent between records. The method relates the logical structure of inventories to the storage and use of data by the computer, and considers the forms data can take for input to the machine. An example from the analysis of inventory data prepared by this method is briefly presented, describing two aspects of agricultural change in East Anglia between 1580 and 1740.     

Extreme weather events in northeastern New Brunswick (Canada) for the period 1950–2012: Comparison of newspaper archive and weather station data

In the context of global warming, changes in extreme weather and climate events are expected, particularly those associated with changes in temperature and precipitation regimes and those that will affect coastal areas. The main objectives of this study were to establish the number of extreme events that have occurred in northeastern New Brunswick, Canada in recent history, and to determine whether their occurrence has increased. By using archived regional newspapers and data from three meteorological stations in a national network, the frequency of extreme events in the study area was established for the time period 1950–2012. Of the 282 extreme weather events recorded in the newspaper archives, 70% were also identified in the meteorological time series analysis. The discrepancy might be explained by the synergistic effect of co‐occurring non‐extreme events, and increased vulnerability over time, resulting from more people and infrastructure being located in coastal hazard zones. The Mann Kendall and Pettitt statistical tests were used to identify trends and the presence of break points in the weather data time series. Results indicate a statistically significant increase in average temperatures and in the number of extreme events, such as extreme hot days, as well as an increase in total annual and extreme precipitation. A significant decrease in the number of frost‐free days and extreme cold days was also found, in addition to a decline in the number of dry days.