Replacing mechanical thermometers with electronic thermometers in a Mediterranean test station and its impact on climatology

In the last two decades of the 20thcentury, most meteorological networks replaced traditional mechanical thermometers (usually placed inside a Stevenson screen) with electronic ones (usually in the open air, protected by a plastic shelter). Impacts of such instrumental change on the climatology of daily minimum and daily maximum temperatures are assessed here, by analysing eleven years of data from a couple of stations operating in parallel in the same test site in a location with a Mediterranean climate. Seasonal effects (i.e. monthly biases ranging from -0.8!C to +1°C) were detected on maximum temperatures, but they compensate each other in the yearly average. In minimum temperatures, electronic thermometers introduced a -0.75°C bias, regardless of the season. Specific biases due to horizontal winds and to vertical heat fluxes were also detected and studied. Two possible causes of the climatological bias were investigated: local wind conditions and different responses to turbulent heat fluctuations by electronic and mechanical thermometers. Other possible causes exist, but they could not be investigated due to a lack of data. The present study shows that, in the test site, replacing traditional thermometers with modern ones in climatological networks introduced an underestimate in minimum temperatures. This suggests that in locations with similar microclimates the real atmospheric warming could be even greater than that observed to date.