'Hydrometrics’ is the measurement of water levels and flow -…
The Flood of Florence
The 4th November is the 52nd anniversary of the flooding of Florence by the Arno River. Flooding was then not limited to Florence, for example Venice saw the highest high-tide on records (1.92m). Florence was probably the worst hit and so that year event is remembered as The Flood of Florence.
During the month of October 1966 heavy and prolonged rainfall was widespread over Italy, reaching a total monthly average of 214mm, equivalent to 188% of normal rainfall, with peaks in Tuscany of 200-300% of the monthly average.
On 3rd November 1966, hence without any significant time to allow for the discharge of the previous rainfall, an exceptionally heavy rain fell on most of the the Arno River basin, reaching a daily average of 200mm, with a peak of 437.3 mm.
In the early hours of the 4th November the Arno River (along with other, but not all, tributaries throughout the city) burst its banks in Florence, flooded the whole city centre, killing 17 people and damaging or destroying millions of artworks, some of which are still being subject to restoration works.
The Arno River Basin
The hydrographic basin of the Arno River is the fifth largest in Italy, covering a total of 8,228 km2, the majority of it at elevations below 300m. The River source is on the Falterona Mount, while the mouth is in the Tyrrenian Sea near Pisa, with a course of 241 km.
The vast majority of the river basin is characterised by the presence of low-permeability sandstone and shale, with an effective permeability of <10%, reflected also in the relatively low flow of springs, generally below 50 litres/s.
More permeable ground is present in the lower part of the basin, where alluvial sediments of sand and gravel are located (although usually covered by lacustrine and alluvial clay). Here groundwater is usually between 2m and 8m from surface. Naturally, the shallow aquifer would feed the River. However, the reverse is often the case, as groundwater is subject to intense abstraction both for industrial activities (the area between Florence and Pisa is famous for its tanneries and shoe factories) as well as for human and agricultural consumption.
Despite the size of its basin, the Arno River has a torrent-like behaviour; while the yearly mean flow in Florence is ca. 60m3/s, the flow can go as low as 2.2m3/s during prolonged summer droughts. On the other hand, the Arno River is well known for its devastating floods, usually in the Autumn. Historical records account for 171 floods in the period 1177 to 1941, with the most devastating event being the 1966 flood, when a flow of ca. 4,100-4,400 m3/s was estimated in Florence (when the maximum amount which can flow through the river embankments is calculated to 2,500 m3/s).
Up to the mid 18th century flooding took place on average every 10 years (there are 55 events recorded between 1261 and 1761, with at least 30 described as heavy and 5 exceptional). It should be noticed that in the last 200 years there have been only 3 floods, two of which described as exceptional. This could be due to the effectiveness of the hydraulic works carried completed in the 18th-19th centuries (see below), which may have managed to reduce flooding risk in Florence and adjacent areas, although not improving but rather worsening the floods in Pisa. Another explanation could be linked to a change in climate at the end of the Little Ice Age, with significant reduction of the frequency and severity of events affecting the whole Arno River basin.
In any case, it is clear that the flooding of 4th November 1966 was caused by an exceptionally severe and prolonged rainfall, due to a peculiar although not unusual interaction between a strong anticyclone based over the Balkans and a deep low pressure located over Italy. The anticyclone acted effectively as a block to the eastward migration of the weather front, causing widespread rainfall over Italy. It is interesting to notice that the above weather situation is always present when exceptional floods occur in Italy. It is also important to notice that rainfall affected most of the Arno River basin, as heavy rainfall only on a section of the river basin would not generate a high flow capable of causing exceptional floods.
In the event of the 1966 flood, a role was also played by the two hydro-electric dams located upstream of Florence (diga di Levane and diga di La Penna). The lakes associated to these dams were full to capacity in those days and water was released in the morning of 4th November. This input proved vital for flooding of the area upstream of Florence, but did not have much of a negative impact on the city. It could be argued that by facilitating flooding upstream of Florence, it somehow alleviated the situation in the city, reducing the volume of water in the Arno River and hence the extent and severity of the flood. Finally, another contributing factor was the thawing of early snow on the mountains in the upper part of the basin due to warm southern winds accompanying the weather front.
Historically, flooding by the Arno River has not been restricted to Florence but affected the whole length of the river. This is an area which for a most of human history has hosted numerous settlements and consequently through the centuries numerous attempts have been made to alleviate the intensity of the floods through extensive hydraulic works.
The first of such works were carried out in the area of Pisa during the 14th century, simply to re-direct flood water away from the city. Extensive works were carried out during the rule of the Medici-Lorena (18-19th centuries) to accommodate the growing demand for residential and agricultural land; swamps were reclaimed, river beds straightened and narrowed, floodplains occupied by ever-sprawling urban settlements, hills and mountains affected by extensive de-forestation. To an extent this extensive intervention managed to create a sort of equilibrium through the creation of a network of canals which contained heavy rainfall and reduced the velocity of surface water run-off. At the same time, the reclamation of Val di Chiana swamps diverting water to the Arno River, away from the natural recipient of the Tiber River basin, added 350-650 m3/s during significant floods, worsening the situation in the Arno River basin.
This equilibrium was dramatically altered during the economic boom following the end of World War II. Economic and social improvement in Italy lead to a significant increase in population, migration from rural areas to urban centers, transfer of workforce from agriculture to industry and services. The network of canals was damaged or destroyed to allow for construction of residential and industrial areas, and the amount of land covered by concrete increased significantly. Moreover, the competent authorities often failed to co-ordinate in planning and intervention and to adequately maintain watercourses, sometimes due to lack of funding.
There is no doubt that these works brought significant economic improvement, alleviated problems during droughts, allowed for the use of the watercourses for transport of goods, and were instrumental in eradicating malaria. However, from the point of view of management of the Arno River basin, they had a dramatic negative impact, increasing volume and velocity of surface water run-off, favouring landslides and soil erosion, aggravating the severity of floods.
According to the 1999 Piano di Bacino – Rischio Idraulico (Arno River Basin Management Plan – Hydraulic Risk Plan published in 1999) storage for ca. 400 million m3, half of which upstream of Florence, is required to avoid flooding by the Arno River not just in Florence, but in the whole river basin. The Comitato Tecnico Scientifico Internazionale stated that existing hydraulic works are inadequate to protect Florence from an event of the same scale as the 1966 flood, with exposure to unacceptable risk of loss of human life and artistic treasures. It is considered that a repetition of a similar event would result in a worse economic outcome.
Given the reality of the land use in the basin, the only realistic solution would be to find storage space for flooding waters upstream of Florence (and it is interesting, and sad at the same time, to notice that this conclusion was already contained in the final report published in 1970 by a committee set up in the wake of the 1966 flood). Unfortunately not much improvement has taken place in the area since the 1966 flood. Works have been carried out in Florence between Ponte Vecchio e Ponte a Santa Trinita, allowing for an increase of flow to 3,400 m3/s from the 2,500 m3/s of 1966. This is surely a welcome improvement, but still not sufficient when compared with the 4,100-4,400 m3/s of 1966 flood. There is a general consensus that not much else can be done within Florence.
Upstream of Florence, hydraulic works to create space for flooding waters have only started in 2014 and are due to be completed in 2022. In the meantime, further settlements curtailed the river floodplains, accompanied by construction of river embankments to protect these built-up areas. A hydro-electric scheme involving an artificial dam (diga del Bilancino) has been completed on one of the Arno River’s main tributaries upstream of Florence. However, this scheme is not considered to be significant in mitigating the flood risk. The national weather forecasting system has greatly improved, which should eliminate the surprise factor which played an important role in 1966 (e.g. these days an early alarm could be raised allowing for some mitigating actions, like emptying the artificial lakes and moving art works to safer locations).
Back to the Future?
Models designed for the prediction and management of floods in the basin indicate that Florence city centre would not be at risk in a case of a flow up to ca. 3,400 m3/s, when the Arno River would break its banks and start flooding the lower part of the city centre (area Santa Croce). However, other parts of the town (Parco delle Cascine, a historical park extending for ca. 160 hectares downstream of the city centre) would be flooded with a volume of 2,800 m3/s. It is clear that with a river flow comparable to the event of 1966, the centre of Florence could escape flooding only if the river heavily flooded the areas upstream of the city. In any case, while the city may escape, there would be extensive and severe floods affecting the area downstream of the city almost to the outskirts of Pisa.