17th Conference on Applied Climatology

4.2

The temperature series of Verona (1741-2006)

Michele Andrighetti, Atmospheric Physics Group, University of Trento, Trento, Italy, Trento, Italy; and D. Zardi and M. De Franceschi

Introduction: Preliminary results from the reconstruction and climatological analysis of the long series of air temperature and precipitation measurements in Verona are presented. After careful search of both published and unpublished data from observations in various archives, libraries and academies, data covering the period 1741-2006 have been collected. A preliminary assessment about quality and reliability has been performed on the basis of metadata and historical information and complete series of monthly mean have been obtained. Various gaps in the temperature series have been filled by means of correlation coefficients with respect to a reference series, evaluated on the basis of a weighted mean of correlation coefficients with the homogeneous series of Milano, Padova and Mantova. Outliers, detected by means of suitable test, have been removed and the Standard Normal Homogeneity Test allowed to detect discontinuities, which were corrected by application of suitable shifts. After application of a low-pass filter, suitable trends were evaluated.

1. VERONA AND ITS CLIMATE

Verona is nowadays a middle sized city (about 260'000 inhabitants in 2001) at the foothills of the Alps, close to the inlet of the Adige Valley, where the Adige River flows out into the Po Plain, bending thereafter into various meanders, on a couple of which Verona developed. The city is since its origins at the junction of two major transport axes – one between central Europe and the Mediterranean basin, the other between east and west in northern Italy.

These particular conditions favoured the development of various activities in the fields of science and technology, especially in connection with economic activities (e.g. agriculture), among which the observations and analysis of weather and climate phenomena. As a consequence, starting in early ‘700, various scholars started, either independently or associated into academies or cultural circles, systematic observations of meteorological variables.

2. COLLECTED DATA

Collected data cover almost completely the period 1741-2006. In the first period (1741-1750) observations were made by Jean-François Séguier, followed (only for the year 1765) by Anton Maria Lorgna and (for 1768-1774) by Giuseppe Maggi in the castle of Castelvecchio. In 1788 the astronomer Antonio Cagnoli started collecting meteorological data for the Accademia di Agricoltura Scienze e Lettere di Verona, a task which was later undertaken, on behalf of the Academy, by many observers without discontinuities until the end of XIX Century. In 1900 Giovanni Fracastoro started again collecting meteorological data on request of the Ufficio Centrale di Meteorologia e Geodinamica. After him two other observers, Arturo Cassandrini and Pio Franchini, continued collecting data till 1971. In 1948 Emilio Bellavite started careful observations, which he still carries out.

3. DATA ANALYSIS

A preliminary assessment about quality and reliability has been performed on the basis of metadata and historical information. Temperature data prior to 1788 need particular care, due to gaps and possible ambiguities in the interpretation of the procedures adopted by the observers, which at that time were not yet fully standardized. Each daily observation consists in an air temperature value, the time at which the observer took the measure and a note about the sky cover, following Jurin's instructions (Jurin, 1723). Mean monthly values are obtained from an homogeneous series starting from the single daily value and passing through the temperature extremes Tmax and Tmin which have been statistically evaluated (see abstract ID: 141050 in this Conference) on the basis of the mean seasonal cycle. This cycle has been analytically calculated (see abstract ID: 141041 in this Conference) from recent data of an urban meteorological station with high acquisition frequency.

However complete series have been obtained so far for the period 1768-2006: gaps have been filled by means of correlation coefficients evaluated on the basis of a weighted mean of correlation coefficients of the homogeneous series of Milano (1763-1998) (Maugeri et al, 2002), Padova (1725-1997) (Camuffo, 2002), and Mantova (1828-1997) (Maugeri et. al, 2002).

Then the series obtained has been normalized to the reference series in order to identify and remove outliers. Values have been discarded when they were outside of an interval 3σ wide (where σ is the standard deviation) centered on the mean value. About the 90% of outliers were located before 1900, as a consequence of the better quality of recent instruments against older ones.

Finally the Standard Normal Homogeneity Test (Moberg and Alexandersson, 1997; Alexandersson and Moberg, 1997) has been applied to the series, detecting various discontinuities, which were corrected by application of suitable shifts.

After application of a suitable digital recursive low-pass filter (de Franceschi and Zardi, 2003) to the homogeneous series of annual and seasonal mean temperatures, linear trends have been evaluated along with their uncertainties.

wrf recording  Recorded presentation

Session 4, Temperature
Monday, 11 August 2008, 3:30 PM-5:00 PM, Harmony AB

Previous paper  Next paper

Browse or search entire meeting

AMS Home Page