The weather diary of Georg Christoph Eimmart for Nuremberg, 1695–1704

Abstract. Weather diaries have long been used to reconstruct past climate. However, they could also be used to reconstruct past weather. Weather reconstructions could help to better understand the mechanisms behind, and impacts of, climatic changes. However, reconstructing the day-to-day weather requires many diaries from different regions covering the same period, ideally combined with instrumental measurements. In this paper, I describe the weather diary of Georg Christoph Eimmart from Nuremberg, covering the period 1695 to 1704, which was particularly cold in Europe. The diary was imaged from the Russian National Library in St. Petersburg and then digitized. It contains twice daily weather conditions in symbolic form, wind direction, and information on precipitation and temperature in text form. Symbols changed during the first two years, after which a much reduced (and stable) set of symbols was used. Re-coding all days according to the later set of symbols, I find no signs of inconsistency over time in symbols, wind direction, and precipitation information extracted from the text. Comparisons with other sources confirm the day-to-day weather information in the diary. For instance, the wind direction in Nuremberg agrees with the daily pressure gradient between Jena and Paris. Three case studies further confirm the meteorological correctness of the information. This is shown on behalf of an eight-day sequence of stormy weather in 1702, a study of the severe winter of 1697/8, and of the summer of 1695, which was cold and wet, possibly related to tropical volcanic eruptions. The examples underline the consistency of the weather diary with other information and suggest that weather reconstructions as far back as the late 17^th century might become possible. However, the spatial information is limited, and any approach arguably needs to make good use of the temporal sequence of information.