The changes in the CI value underline how events more intense take during the years an important role in determining the total precipitation. Fig. 12 shows the NSI obtained for the simulated hyetographs for the years 1954, 1981 and 2006, and considering different return periods. The NSI index gives an idea of how critical the area under analysis: if the rainfall persists, the faster the network gets saturated, the faster response of the area to the input rainfall. In an area where the drainage is entirely mechanical, this information can be critical, giving an idea of the timing for the ignition of the pumping stations. selleckchem The decrease in storage

capacity from 1954 to 1981 and then 2006 results in a worsening of the situations in all the cases considered. Fig. 13 depicts the average NSI for all the considered hyetographs (a), and the differences in NSI considering: (1) the average performance, (2) the scenario with the highest NSI, therefore the case where the area in 1954 was expected to have the most delayed response to the storm (Sym18); and (3) the worst case scenario (Sym03) where the area in 1954 was expected to have the fastest response to the storm (∼lowest NSI). On average, for the year 1954 the NSI is about 1 h and 15 min for the most frequent events (return period of 3 year), and it decreases to about 40 min

for the most extreme GDC 973 events (return period of 200 year). When considering the conformation of the network

in 2006, the NSI is about 40 min for the most frequent events, and decreases to 15 min for the most extreme ones (Fig. 13a). The highest changes in the NSI index derive from the changes in storage capacity registered from 1954 to 1981, while from 1981 to 2006 the NSI changes slightly. Our empirical data, with a use of a simple index, highlight issues already underlined by other researchers. Graf (1977) showed how the changes in drainage networks due to urbanization can result a reduced lag time. A reduction in the time to peak flow in relation to installation of field drains cAMP was also reported by Robinson et al. (1985) and Robinson (1990). Among others, Backer et al. (2004) and McMahon et al. (2003) drew attention to the increased flashiness of stormflows in urbanized basins. Similar conclusions have been found by Smith et al. (2013) that underlined how the timing of the hydrological response is strictly linked to the management of the artificial drainage network and the storage volumes. Wright et al. (2012), comparing basins with different land use and urbanization degree in Atlanta, found that flood response is strictly influenced, among other factors, by the drainage network structure and the available storage volumes.