In this study a comprehensive quality assurance (QA) system for both station and sea level pressure data is designed and applied to hourly pressure records (for 1953-2002) from 761 Canadian stations, to produce a high quality database of station and sea level pressures for various climate studies. The main principles of the QA system are described in detail, followed by a brief emphasis on the decision making and error correction algorithms. The general performance of the QA system and the main problems in the Canadian historical pressure database are discussed and illustrated through various examples.

The results show that there exist serious systematic errors (i.e., sudden changes in the mean, or mean-shifts) in the Canadian pressure database, which are caused either by the use of wrong station elevation values in the reduction of barometer readings to station or sea level pressure values (e.g., the “50 feet rule” or station relocation without updating the station elevation), or by transposition/swapping of station and sea level pressure values, or by mistakes made in the archive data ingestion or data recording/digitization processes (e.g., use of a wrong base number). Random errors also exist and are mainly due to transposition of two digits or miscoding of one or two digits. These errors must be corrected before the data are used in various climate studies, especially climate change related studies (e.g., assessment of climate trends and variability, weather regimes).