After an overview of the problems concerning the early rain-gauges and their thresholds, a study is made to investigate the impact that the instrumental threshold of a rain-gauge has on the ...distribution of precipitation frequency and amount. Tests are performed using two historical datasets, i.e., the observations by Giovanni Poleni in Padua from 1725 to 1760, and Jacopo Bartolomeo Beccari in Bologna from 1723 to 1765, and two modern rain-gauge records, i.e., taken at the Botanical Garden, Padua, and at the Hydrographic, Bologna, from 1990 to 2019. The tests involve applying a filter to the datasets to simulate the action of an instrumental threshold. The result is that the threshold has an enormous impact on the frequency and a smaller one on the total amount. The study includes how the threshold affects the percentile distribution of precipitation amounts. The results provide indications to correct and interpret early records and to test their quality. Moreover, they are useful in the analysis of long time series composed by datasets derived from different instruments for climate studies.
The article offers an overview of the time frames used in instrumental series and how to transform them into modern units. In the early instrumental period, time was measured with sundials or ...mechanical clocks regulated every day with the culmination at noon, making reference to the apparent solar time (AST) and the local meridian. Every day had a slightly different duration, start, and end for the apparent changes of speed of the Sun. When canonical hours were used, hours were computed starting from twilight. In the late eighteenth century, the start was established at midnight. In the mid-nineteenth century, when precise clocks were available, it was possible to adopt an average time, with all days having the same duration, but related to the local meridian. A further step was to unify the time of all cities of a country adopting the time of the capital. Finally, the interregional rails, the telegraph, the telephone, and the international contacts required to unify the different time frames. This lead to the creation of the Coordinated Universal Time (UTC) and the time zones (TZ). The change from AST to UTC introduced two important time differences: one related to the variability of the apparent solar motion and one related to the longitude of the site. Instrumental records, especially the longest ones, are affected by changes in time frames that may cause bias. In this paper, the time changes of 92 selected European cities when they passed from AST to Western European Time, Central Europe Time, or Eastern Europe Time, are considered, and the departures during the calendar year are calculated. Moreover, the bias in temperature derived from the time frame change has also been evaluated in ten case studies over Europe. This paper will assist historians and climatologists to recognize and correct the time departures that affect meteorological series concerning temperature, barometric pressure, and other variables with daily cycles. This correction is crucial to assess climate changes. Specific aims are as follows: to make a friendly explanation of the methodology to pass from old time frames to UTC; to provide transformation equations to remove the bias for the time difference; to pass from time difference to temperature bias to homogenize early records with modern ones.
The earliest temperature series in Paris, from May 1658 to September 1660, taken by Boulliau is made available at daily resolution after a careful work of homogenization, correction and calculation ...of the average values. New results are achieved concerning the Little Florentine Thermometer, building location, thermometer exposure, observation methodologies followed by Boulliau and weather in the mid-seventeenth century. Two methods are used and compared to calculate the daily average from readings taken at random sampling times. The first one is based on the reading needing the smallest correction to be transformed into a daily average; the second considers all the readings of the day and makes a bulk average of the individual results. The series is compared with the temperature record by the Grand Duke Ferdinand II in Florence, which was the primary station of the Medici Network (1654–70). In addition, the comparison of the earliest temperature series in Paris and Florence with their respective 1961–1990 reference period gives a clear image of the change of weather conditions in Europe in the middle of the seventeenth century. Key features were a strong variability for warm-air and cold-air outbreaks, severe winters and cold summers in Paris, not affecting Florence.
This paper deals with the earliest meteorological observations in Florence after the Medici Network (1654–1670), i.e., from mid-seventeenth to mid-eighteenth century, and puts them in the context of ...the history of meteorology in Florence. After the gap caused by the Inquisition, observations started again in the eighteenth century, made by Cipriano Antonino Targioni (1728–1748), Giovanni Targioni-Tozzetti (1737–1740), Pietro Gaetano Grifoni (1751–1766), Leonardo Ximenes (1752), and Luca Martini (1756–1772). The first two records were lost, and this paper considers those by Grifoni and Ximenes. The latter is affected by severe bias; the former is of good quality and has been recovered and analyzed. Both the observers made only one reading a day and the metadata are scarce. The paper discusses several issues: the conversion from the apparent solar time to the Central Europe Time; the transformation from a single reading to a daily average; the identification of the thermometric liquid and the scale; the test made with the snow benchmark; the comparison with the contemporary series in Bologna. The comparison of the reconstructed series with other periods, i.e., 1654–1670, 1881–1910, 1961–1990, and 1991–2017, reveals that in mid-eighteenth century the temperature reached the lowest levels, especially in summer, and showed a sudden warming in the most recent decades.
In the framework of the meteorological observations in Florence after the Medici Network (1654–1670), the earliest surviving record by Pietro Gaetano Grifoni (1751–1766) was followed and partially ...overlapped by the temperature series by Luca Martini (1756–1775), which has been recovered from different sources and analyzed in this paper. Martini series should be divided in two periods, the turning year being 1765, when he changed house, still in Florence. A critical data analysis and the comparison with the Grifoni contemporary series in Florence clarify the thermometric liquid and scale used in the different periods. The reconstruction of the 1756–1775 daily average temperatures is carried out here following a methodology that includes the conversion from the apparent solar time to the Central Europe Time; the transformation from single/double readings to a daily average; the analysis of the hourly temperature variation during the calendar year in the 1961–1990 reference period in Florence; the test made with the snow benchmark; and the correction of the bias due to the local microclimate in the first period. The final series (1751–1775) has been composed combining Grifoni and Martini observations, covers one quarter of the eighteenth century, and constitutes the only surviving outdoor temperature record in Florence. The comparison with other periods from the mid-seventeenth to the early-twenty-first century confirms that in the mid-eighteenth century, the temperature reached the lowest levels and that a marked warming has characterized the most recent decades. Information on the pre-industrial climatology in Florence can be useful for climatic change study.
Long instrumental records are fundamental to study climate changes, but their reliability and quality have to be checked before any statistical research. Moreover, when metadata are used to solve ...some problems, data interpretation may change and require further work to refine the series. A thorough revision of the three-century precipitation series in Padua (1713–2018) shows that the results of previous analyses were affected by serious biases. This paper clarifies key features concerning early instruments, exposure, relocation and observational protocols. The daily analysis pointed out a number of problems, bias, irregular sampling and underestimations that have passed unobserved so far. A comparison with the parallel series of Bologna and Venice made it possible to distinguish bias from the climate signal or to reconstruct missing data. The instrumental threshold was recognized to be fundamental to determine the frequency of precipitation, but less important with respect to its amount. This paper provides a methodological example to test the goodness of long instrumental series, in particular to identify problems related to metadata and observations, which is the preliminary step to perform a sound correction and obtain a reliable series. It also includes the set of original raw data, transformed into modern units.
This paper analyses the response of alcohol thermometers in relation to the departure from linearity and the choice of the calibration points. The result is that alcohol thermometers are affected by ...large departures that reach a maximum (i.e. −6 °C) at 50 °C ambient temperature. This may have caused a severe bias in early records, when alcohol thermometers were popular, especially during the Little Ice Age. Choosing a lower temperature for the upper point, calibration may substantially reduce this bias. Examples are given with thermometers in use in the 17th and 18th centuries. A careful correction of long series is necessary to avoid misleading climate interpretations.
The precipitation record taken by J.B. Beccari in Bologna from 1723 to 1765, with three observations a day, has been recovered from the original Logs and analysed. Although metadata are scarce or ...even missing, a thorough investigation of the contemporary sources, the interactions between Beccari, his colleagues in Padua, and the Royal Society London, as well as the record analysis, have allowed reconstruction of most of the information concerning the instrument, its location and exposure, the measuring protocol and units. Daily, monthly and yearly amounts and frequencies, as well as extreme events, have been analysed. The first decade (i.e. 1723–1733) was drier than the 1961–1990 reference period and the subsequent decades wetter. During the calendar year, summer was dryer and October was characterized by a stronger activity of Atlantic perturbations.
The Padua temperature series is one of the longest in the world, as daily observations started in 1725 and have continued almost unbroken to the present. Previous works recovered readings from the ...original logs, and digitalized and corrected observations from errors due to instruments, calibrations, sampling times and exposure. However, the series underwent some changes (location, elevation, observing protocols, and different averaging methods) that affected the homogeneity between sub-series. The aim of this work is to produce a homogenized temperature series for Padua, starting from the results of previous works, and connecting all the periods available. The homogenization of the observations has been carried out with respect to the modern era. A newly released paleo-reanalysis dataset, ModE-RA, is exploited to connect the most ancient data to the recent ones. In particular, the following has been carried out: the 1774–2023 daily mean temperature has been homogenized to the modern data; for the first time, the daily values of 1765–1773 have been merged and homogenized; and the daily observations of the 1725–1764 period have been connected and homogenized to the rest of the series. Snowfall observations, extracted from the same logs from which the temperatures were retrieved, help to verify the robustness of the homogenization procedure by looking at the temperature frequency distribution on snowy days, before and after the correction. The possibility of adding new measurements with no need to apply transformations or homogenization procedures makes it very easy to update the time series and make it immediately available for climate change analysis.
This paper is focused on the closed-tube Stancari air thermometer that was developed at the beginning of the eighteenth century as an improvement of the Amontons thermometer, and used to record the ...temperature in Bologna, Italy, from 1715 to 1737. The problems met with this instrument, its calibration and the building technology in the eighteenth century are discussed in order to correct the record. The used methodological approach constitutes a useful example for other early series. The analysis of this record shows that the temperature in Bologna was not different from the 1961–1990 reference period. This result is in line with the contemporary record taken in Padua, Italy, confirming that this period of the Little Ice Age was not cold in the Mediterranean area.
PDF
