The development and updating of the Stockholm Historical Weather Observations data set is a collaboration between the Swedish Meteorological and Hydrological Institute and Stockholm University.
Homogenization of the Stockholm temperature series, i.e. removal of artificial changes of observed temperatures that are not due to changes in climate and weather, is described in Moberg et al. (2002). These corrections account for several changes in observation hours, for the urban warming trend after 1870, for missing or poor early observations in 1756 and 1763 and for an incorrectly calibrated thermometer used 1819-1825. The adjustment for the urban warming trend is applied such that the most recent homogenized temperatures are made somewhat colder than the observed temperatures, in order to make the entire series being respresentative of the rural conditions that prevailed before the mid-19th century.
An additional adjustment has been made to temperatures from May to August during 1756-1858, because it has been concluded (Moberg et al. 2003) that the thermometer in this period was not sufficiently well protected against direct sunlight, reflection of sunlight and radiation from the building and surrroundings. The effect of this likely caused the observed temperatures to be too high. The correction applied is -0.3°C in May and August and -0.7°C in June and July. The size of this adjustment is an ad hoc expert judgment, rather than a result from strict statistical analysis, but is in agreement with findings by Moberg et al. (2003) and was used for the first time in Moberg et al. (2005). Apart from this correction, the monthly temperatures are the same as in the homogenized data described above. The data that include the adjustment 1756-1858 do probably more truly reflect the real temperature variations in the hypothetical situation of Stockholm being still under the same rural conditions as before the mid-19th century.
Non-homogenized monthly temperatures that only account for changes in observation hours, for the missing or poor early observations in 1756 and 1763 and for the incorrectly calibrated thermometer used 1819-1825 are also avaialble. These data may be of interest in particular for studies where no adjustment for the urban warming trend is wanted.