Guest post by Paolo Mezzasalma – meteorologist
Anthony Watts’ surface station survey project inspired my attempt for a similar task devoted to the Italian network. Weather observations in Italy are carried out by essentially three institutions, each with different purposes.
A national synoptic network was established in its current fashion after the second world war by the Italian Air Force (Aeronautica Militare – thereafter AM), devoted to aeronautic assistance and routine meteorological measurements. Many of these stations are located in airports but some are scattered through the Italian territory, even on remote mount tops. At the end of the Eighties the national agency for flight assistance (Ente Nazionale Assistenza al Volo – ENAV) took the duty for weather observations at the main civil airports. Currently 81 stations are managed by AM and 24 by ENAV.
Another widely scattered network of stations was established in the first decades of the 20th century by Ufficio Idrografico, whose main goal was the precipitation measurement and the river monitoring. These stations (a few thousands) were sited mainly in cities and villages and not in unpopulated areas; some of them were also provided with temperature sensors. This network is currently in charge of the 20 regional administrations in which the Italian territory is divided. Stations were upgraded and integrated with the local networks.
A third and historical network for weather readings was developed in the 18th and 19th century by many astronomical observatories in the main cities (about three dozen).
For this first report, the web provided me most of the close satellite imagery imagery (Google Earth, Pagine Gialle), 3D vision pictures (Microsoft Live Maps) and photographs (Google Earth, several sites) of almost all the stations of the synoptic network, without a real need to go and take personally a photo. Furthermore, AM has been reporting very informative description of a few of its observatories at its own Internet site
The use of all these pieces of information collected from the web allowed me to do a satisfactory job for 88 out of the 105 stations of the synoptic network, whose readings are promptly available in the Global Telecommunication System, see the map in figure 1 for their location.
This report in particular deals with the 6 stations that are found in the GHCN and GISS data base with values updated in 2008 (1 Messina; 2 Trapani/Birgi; 3 Marina di Ginosa; 4 Termoli; 5 Pisa/S.Giusto; 6 Trieste). Following reports will eventually deal with stations of the GSN project or which are included in the ECA&D data set.
Starting from Sicily, Messina (station #1) is the city on the shore of the sea strait that separates the island from the Italian mainland. The local observatory is in a now fully developed urban area, as you can see in figure 2; the Stevenson Screen is over a small tower, also visible in the photograph in figure 3. Figure 4 shows the observatory before the catastrophic earthquake of the year 1908.
The area has surely seen a lot of urbanization in the 60s through the 80s while some pinus halepensis trees have been growing immediately to the north of the station, so not to cast any shadow on the shelter but acting as a barrier to the free flow of air and IR scattering.
Figure 5 plots the yearly temperature data and the remarkable adjustment applied by GISS.
At the western tip of Sicily it is located the airport of Trapani/Birgi (#2), a military facility hosting also some civil air traffic. Figure 6 (top) shows that the area around the airport is quite rural, while Giss consider it as being urban; the civil air station is in the building in the south-eastern part. The Stevenson Screen is in the middle of the air field, to the north of the runway (figure 6, bottom).
Someone could contest the attribution of that white spot as to be a Stevenson shelter, but I’m quite confident I have acquired a certain ability to detect the correct sites; my ability was also tested in a few cases. Giss adjustment to raw data is shown in figure 7.
Marina di Ginosa (#3) is marked as being rural and no correction is apparent in GISS plots. Figure 8 (top) shows, anyway, that the area suffered from an extensive urbanization as many coastal areas in Italy did. Zooming in at the area in which I was told the sensors are located doesn’t allow, in my opinion, a very clear detection of the Stevenson Screen: a closer inspection is required. The plot of GISS temperature data is in figure 9.
The meteorological station in Termoli (#4) is located, since the year 1946, on the top of an ancient 50 metre high tower in the old part of the city, see figure 10. The old city is in front of the Adriatic Sea and is surrounded by the sea and the harbour for three quarters of its sides, as figure 11 shows. I think that the urban texture immediately around the tower has not changed a lot in the last decades. Data in GISS dataset have large lacunae (figure 12).
Of course there is a weather station close to the Leaning Tower of Pisa (#5), sited at the local airport which shares military and extensive civil traffic. This is the first station I found in Google maps after I spotted it during a takeoff from Pisa runway, see figure 13. The Stevenson Screen is not very clear in the bird eye vision from live Microsoft (figure 14), probably because of a deep cloud cover over it. As you can see in figure 13, even if the airport is very close to the city, the sensors are between the runway and some cultivated fields. Giss adjustment (figure 15) makes Pisa much colder in the past (a negative urban influence).
Trieste (#6) is in the northern shores of the Adriatic Sea, at the border with Slovenia. The meteorological observatory is at its present location (figure 16) since the middle of 2001 . It was previously located closer to the city centre, as figure 17 shows. Giss adjustment (figure 18) is positive but does not change data of the very first years. As a result, the temperature increase is more monotonic.