Abstract

The discrete three-parameter Lerch distribution is used to analyse the frequency distribution of inter-arrival times derived from 26 daily precipitation time-series, collected by stations located throughout a 28,000 km2 area in NW Italy (altitudes ranging from 113 m to 2,170 m a.s.l.). The precipitation regime of these Alpine regions is very different (latitude 44.5 to 46.5 N) from the typical Mediterranean precipitation regime of the island of Sicily (latitude 37 to 38 N), where the Lerch distribution has already been tested and whose results are compared. In order to verify the homogeneity of the precipitation time series, the Pettitt test was preliminarily performed. In this work, a good fitting of the Lerch distribution to NW Italy is shown, thus evidencing the wide applicability of this kind of distribution, also allowing to jointly model dry spells and wet spells. The three parameters of the Lerch distribution showed rather different values than the Sicily ones, likely due to the very different precipitation regimes. Finally, a relevant spatial variability of inter-arrival times in the study area was revealed from the regional scale application of the probability distribution here described. The outcomes of this study could be of interest in different hydrologic applications.

INTRODUCTION

A robust statistical inference of the alternation of wet spells and dry spells (DS) is still an open issue, despite its fundamental role in several common hydrological applications, including irrigation management (Rallo et al. 2014; Canone et al. 2015), the analysis of precipitation temporal trends (Agnese et al. 2012) and the eco-hydrological modelling of land–atmosphere feedbacks (Agnese et al. 2016).

In most studies, the statistical analysis of the sequences of rainy days, previously denoted wet spells (WS), and that of no-rainy days, DS, was carried out separately. In particular, the observed WS frequencies were commonly modelled by the geometric distribution, consequent to the statement that a rainy day is a ‘success’ outcome of classical Bernoulli trials, with constant rain probability.

As showed by Chatfield (1966), this assumption of constant rain probability is commonly unrealistic for modelling DS frequency distribution; in fact, as stated by the author, ‘the probability that a dry day will be followed by a dry day does increase with the previous number of consecutive dry days’. By using daily rainfall data of London from 1958 to 1965, Chatfield successfully fitted the dry spell frequencies through the logarithmic-series distribution; this distribution allows accounting for the observed persistence of dry states, because of its peculiar pattern having a slowly decaying tail (Kottegoda & Rosso 1997). In a previous study, Agnese et al. (2014) demonstrated that the two-parameter polylogarithmic distribution significantly outperforms the logarithmic distribution, in order to statistically infer the observed DS frequencies.

As mentioned above, here the joint modelling of both WS and DS is performed, instead of treating them separately. In fact, a single occurrence process regulates the alternating sequences of rainy and no-rainy days. A similar approach was carried out by Strupczewski et al. (2013), who derived a two-component model of flood dynamics that combines the duration of high waters and the exceeding of a discharge threshold, thus distinguishing two possible states, namely, flood and no-flood.

A joint modelling of both WS and DS can be obtained by considering the observed frequencies of the so-called ‘inter-arrival times (IT)’ of precipitation events, ITs, which represent the series of times elapsed between two subsequent rainy days, instead of the commonly used number of dry days. Indeed, the probabilities of rainy day sequences (i.e., wet spells, WS) and of dry sequences (i.e., dry spells, DS), can be easily derived from the IT distribution. IT series can be viewed as a realization of a renewal process (Feller 1968; Cox 1970; Buishand 1977), by assuming that ITs are independent and identically distributed (i.i.d.) random variables. Accordingly, the WS geometric distribution directly emerges from the renewal property, whereas the DS probabilities can be easily calculated by rearranging the IT probabilities.

Another important aspect of this work is to use a discrete probability distribution. In the literature, both discrete and continuous probability distributions were used to model the frequency distribution of inter-arrival time-series (Foufoula-Georgiou & Lettenmaier 1986; Rodriguez-Iturbe et al. 1987; de Groen & Savenije 2006; Manfreda & Rodrıguez-Iturbe 2006), although the inter-arrival time is properly a discrete random variable. In fact, Foufoula-Georgiou & Lettenmaier (1986) demonstrated the statistical advantages of treating the rainfall occurrence as a discrete process, rather than a continuous one. In order to describe the rainfall duration, Manfreda & Rodrıguez-Iturbe (2006) considered the exponential distribution, which is the extension to the continuum of the discrete geometric distribution. Eichner et al. (2011) discussed the influence of the persistence of the daily rainfall process on that of the IT by using long-term correlated data sets.

In a previous work, the ITs have already been modelled by Agnese et al. (2014) using the discrete three-parameter Lerch distribution. This distribution is able to account for some peculiar features usually observed in IT samples, which other distributions are not capable of reproducing as well. These features are the high standard deviation and the high positive skewness, the monotonically decreasing pattern of frequencies with a slowly decaying tail and a distinctive ‘drop’ marking the passage from the wet status (IT= 1) to the dry status (IT= 2). The ability of the Lerch distribution in reproducing all of these features allows the use of this distribution for a variety of hydrological applications. In a previous work regarding the period 1975–2005, Agnese et al. (2014) successfully applied the Lerch distribution to the IT frequencies derived from 55 daily rainfall series recorded in Sicily.

However, rainfall regime in Sicily is characterized by a strong seasonality, splitting of the year into two parts, a wet cold season (from October to March) and a dry warm one (from April to September). Consequently, data have been preliminarily partitioned into two groups falling in those two different periods of the year, and then they were fitted separately via the Lerch distribution. For data belonging to the ‘wet’ semester, all records satisfied the chi-square test of the goodness of fit; whereas for data of the ‘dry’ semester, 11 series did not satisfy the test, likely because of the small available sample size.

The primary objective of this work is to test the Lerch distribution in an area ranging from the highest peaks of the Alps to the Tyrrhenian Sea. In order to do that, the IT series were derived from long daily precipitation series (70/90 years duration), collected in 26 stations located throughout the Piedmont and Aosta Valley, in typical Alpine and sub-Alpine areas of NW Italy. The climate of the studied area is strongly spatially variable (Figure 1) and was classified as sub-litoranean, sub-alpine, pre-alpine and sub-continental, and differs completely from that of Sicily which is typical Mediterranean. The different parameter values derived by the statistical inference were compared to those obtained in the study about the Sicily region and results were discussed. Therefore, the application of the three-parameter Lerch distribution also to the area of NW Italy represents an important test to establish the wide applicability of the proposed approach. An attempt to explore the validity of the hypothesis (usually accepted) that ITs are i.i.d.s was carried out.

Figure 1

Physical map of Piedmont and Aosta Valley with the station locations (dots) and the elevation legenda of the altitudes (in metres) at the right top of the figure.

Figure 1

Physical map of Piedmont and Aosta Valley with the station locations (dots) and the elevation legenda of the altitudes (in metres) at the right top of the figure.

Moreover, an objective of this paper is also to use the Lerch distribution in order to describe the sequences of rainy days, interrupted by 1-day DS, which could have a certain interest from a hydrological point of view, representing in a broad sense, a unique perturbed period, not significantly altering the antecedent soil moisture status. The equivalence on using a single IT three-parameter Lerch distribution, in place of using two distributions, namely DS two-parameter polylogarithmic distribution and WS one-parameter geometric distribution, is also argued.

Finally, the expected return periods of dry and wet spells at the regional scale are determined, in order to quantify the spatial variability of the precipitation regimes in the investigated area, and to provide suitable results for the sake of agricultural water management.

STUDY AREA AND DATA

In this work, the daily precipitation data long series (70/90 years) from 26 stations have been used. They belong to the available databases of the National and Regional Observatories (ARPA Piemonte and Centro Funzionale Valle d'Aosta), and of the Italian Meteorological Society (Mercalli & Cat Berro 2008). These stations are located throughout the Piedmont and Aosta Valley area (Italy) and cover an area of 28,000 km2, characterized by different climatic regimes, ranging from sub-continental to sub-litoranean ones, in between 113 and 2,170 m altitude. It is an area characterized by very high mountains like Mont Blanc, Matterhorn and Gran Paradiso, quite close to the plains and to the sea. In addition to the mentioned orographic issue, the irrigated agricultural Po Plain plays a major role in the air humidity pattern because of the widespread surface irrigation, which saturates wide areas. This produces a notable variation in the average statistics of precipitation occurrence through the year and through the study area.

The stations were selected in order to have highly reliable and long records, with few gaps in the data series. Figure 1 shows the location of the stations (dots) on a physical map of the area. Table 1 reports the available years of observations, the elevation, the latitude and the longitude (UTM 32 T, m). For the purpose of this study, a common starting year (the year 1919, if available) was considered for all 26 data series. The IT series were obtained from the recorded precipitation observations using the precipitation threshold h* = 1 mm (Figure 2), which is usually adopted in Italy by regional observatories in order to discriminate between rainy and no-rainy days.

Table 1

Years of observations, elevation and coordinates of the selected 26 stations

ID Station From To Available records (years) Elevation (m a.s.l.) North latitude   
UTM 32 T (m) East longitude 
Alpe Cavalli 1929 2010 82 1,510 431,301 5,104,900 
Aosta 1891 2010 120 583 369,928 5,066,980 
Bra 1862 2010 149 290 408,452 4,949,620 
Casale Monferrato 1913 2011 99 113 455,028 4,999,120 
Ceresole Reale 1926 2011 86 1,579 361,956 5,032,690 
Cuneo 1901 2010 110 536 382,020 4,917,460 
Domodossola 1871 2013 143 302 445,620 5,106,821 
Formazza 1913 2008 96 1,270 456,009 5,136,150 
Gressoney-D'Ejola 1927 2011 85 1,850 408,907 5,071,847 
10 Lago Cignana 1927 2011 85 2,170 390,934 5,081,400 
11 Lago Malciaussia 1936 2011 76 1,810 354,297 5,008,050 
12 Lago Moncenisio 1931 2010 80 2,000 337,919 5,011,030 
13 Lago Piastra 1916 2010 95 956 371,812 4,900,240 
14 Lago Saretto 1913 2010 98 1,633 335,203 4,927,000 
15 Lago Toggia 1932 2010 79 2,160 456,050 5,141,710 
16 Lombriasco 1928 2012 85 241 392,405 4,966,572 
17 Moncalieri 1866 2010 145 225 397,101 4,983,130 
18 Oropa 1961 2012 94 1,180 420,972 5,053,905 
19 Ovada 1913 2010 98 187 472,477 4,942,550 
20 Perreres 1928 2011 84 1,750 392,292 5,085,080 
21 Rosone 1938 2011 74 714 375,024 5,034,280 
22 Torino 1802 2010 209 238 395,177 4,991,310 
23 Ussin 1929 2007 79 1,322 392,196 5,079,520 
24 Valgrisenche 1913 2007 95 1,780 348,869 5,054,490 
25 Varallo 1913 2011 99 453 441,947 5,074,045 
26 Vercelli 1927 2010 84 135 450,518 5,020,260 
ID Station From To Available records (years) Elevation (m a.s.l.) North latitude   
UTM 32 T (m) East longitude 
Alpe Cavalli 1929 2010 82 1,510 431,301 5,104,900 
Aosta 1891 2010 120 583 369,928 5,066,980 
Bra 1862 2010 149 290 408,452 4,949,620 
Casale Monferrato 1913 2011 99 113 455,028 4,999,120 
Ceresole Reale 1926 2011 86 1,579 361,956 5,032,690 
Cuneo 1901 2010 110 536 382,020 4,917,460 
Domodossola 1871 2013 143 302 445,620 5,106,821 
Formazza 1913 2008 96 1,270 456,009 5,136,150 
Gressoney-D'Ejola 1927 2011 85 1,850 408,907 5,071,847 
10 Lago Cignana 1927 2011 85 2,170 390,934 5,081,400 
11 Lago Malciaussia 1936 2011 76 1,810 354,297 5,008,050 
12 Lago Moncenisio 1931 2010 80 2,000 337,919 5,011,030 
13 Lago Piastra 1916 2010 95 956 371,812 4,900,240 
14 Lago Saretto 1913 2010 98 1,633 335,203 4,927,000 
15 Lago Toggia 1932 2010 79 2,160 456,050 5,141,710 
16 Lombriasco 1928 2012 85 241 392,405 4,966,572 
17 Moncalieri 1866 2010 145 225 397,101 4,983,130 
18 Oropa 1961 2012 94 1,180 420,972 5,053,905 
19 Ovada 1913 2010 98 187 472,477 4,942,550 
20 Perreres 1928 2011 84 1,750 392,292 5,085,080 
21 Rosone 1938 2011 74 714 375,024 5,034,280 
22 Torino 1802 2010 209 238 395,177 4,991,310 
23 Ussin 1929 2007 79 1,322 392,196 5,079,520 
24 Valgrisenche 1913 2007 95 1,780 348,869 5,054,490 
25 Varallo 1913 2011 99 453 441,947 5,074,045 
26 Vercelli 1927 2010 84 135 450,518 5,020,260 
Figure 2

Sketch of inter-arrival times (IT), dry spells (DS) and wet spells (WS). A day is a rainy day if the daily precipitation depth h is greater than a threshold h*.

Figure 2

Sketch of inter-arrival times (IT), dry spells (DS) and wet spells (WS). A day is a rainy day if the daily precipitation depth h is greater than a threshold h*.

A simple analysis of the monthly-averaged frequencies of wet days through the year (Table 2), revealed the presence of a ‘wet’ period in spring (peak in May) and a ‘dry’ period in winter (minimum in January), for all the stations. Furthermore, another ‘wet’ period in autumn is observed in the area located at low–medium altitudes, whereas a secondary peak in August is present at the high altitudes. These facts complicate the splitting of the year into two homogeneous periods valid for the whole area, as for the Sicily case.

Table 2

Average number of events per month, <# events/month>, of the 26 selected stations

  Station Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec 
Alpe Cavalli 6.25 6.84 7.67 10.32 12.83 11.22 9.06 10.15 9.06 8.87 8.50 6.72 
Aosta 5.08 5.01 5.31 5.80 6.63 6.77 5.59 5.96 5.71 6.26 6.45 5.76 
Bra 3.90 4.49 5.59 7.39 7.99 6.08 3.77 4.73 5.25 6.28 5.88 5.00 
Casale Monferrato 5.03 4.97 5.51 6.72 7.22 5.89 4.08 4.57 4.81 6.26 6.52 5.34 
Ceresole Reale 5.74 6.22 6.81 9.26 11.11 9.83 7.64 8.74 7.98 7.84 7.51 6.35 
Cuneo 4.46 5.00 7.00 8.67 9.14 7.63 5.02 5.28 6.29 7.35 6.47 5.37 
Domodossola 5.07 5.23 6.61 9.05 11.12 9.26 7.84 8.47 7.58 7.44 7.79 5.46 
Formazza 7.48 7.59 7.84 9.34 11.13 10.36 9.27 10.07 8.74 8.85 8.83 7.73 
Gressoney-D'Ejola 6.74 7.21 7.62 9.84 12.28 11.62 9.32 10.22 7.98 7.87 7.99 7.42 
10 Lago Cignana 7.82 8.01 8.15 9.19 11.48 10.76 8.50 9.78 8.11 7.93 8.22 7.78 
11 Lago Malciaussia 6.53 6.69 7.47 9.62 11.69 10.53 6.90 8.37 7.70 8.19 7.43 6.83 
12 Lago Moncenisio 6.31 6.18 6.43 7.49 9.96 9.06 6.60 7.17 7.34 7.19 6.91 6.48 
13 Lago Piastra 5.10 5.16 6.63 9.04 9.50 7.71 5.35 5.95 6.46 7.42 6.62 5.55 
14 Lago Saretto 5.48 5.96 6.87 8.74 9.61 9.47 6.35 6.88 7.11 7.84 7.02 6.52 
15 Lago Toggia 10.15 10.12 9.59 10.68 12.26 12.23 10.48 11.94 10.09 9.84 11.22 10.24 
16 Lombriasco 4.24 4.36 5.08 7.05 8.22 6.51 4.05 5.31 5.36 5.68 5.82 4.49 
17 Moncalieri 3.91 4.58 5.41 7.61 9.08 7.28 5.18 5.45 5.59 6.26 5.63 4.80 
18 Oropa 5.29 6.39 7.98 11.67 13.53 11.48 8.41 9.03 8.44 8.08 7.56 5.45 
19 Ovada 5.17 5.29 5.60 6.41 5.97 4.16 2.80 3.35 4.66 6.50 7.66 5.83 
20 Perreres 7.19 7.31 7.29 8.73 11.11 10.67 8.92 9.62 8.28 7.87 8.07 7.29 
21 Rosone 5.12 5.72 6.36 9.58 11.60 10.15 7.28 8.54 7.26 7.60 6.55 5.53 
22 Torino 4.10 4.73 5.73 8.26 9.85 8.29 5.54 5.88 6.09 6.57 5.95 4.97 
23 Ussin 6.13 6.10 6.50 7.79 10.34 9.74 7.77 8.41 7.29 7.84 7.29 6.26 
24 Valgrisenche 7.61 7.66 8.02 8.57 8.79 9.59 7.62 8.62 7.69 7.88 8.41 7.84 
25 Varallo 3.88 4.39 6.11 9.26 11.04 9.87 7.62 7.64 7.15 7.39 6.96 4.42 
26 Vercelli 4.69 4.61 5.51 7.29 8.12 6.73 4.54 5.48 5.43 6.62 6.98 5.22 
 <# events/month> 5.71 5.99 6.72 8.59 10.06 8.96 6.75 7.52 7.06 7.45 7.32 6.18 
  Station Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec 
Alpe Cavalli 6.25 6.84 7.67 10.32 12.83 11.22 9.06 10.15 9.06 8.87 8.50 6.72 
Aosta 5.08 5.01 5.31 5.80 6.63 6.77 5.59 5.96 5.71 6.26 6.45 5.76 
Bra 3.90 4.49 5.59 7.39 7.99 6.08 3.77 4.73 5.25 6.28 5.88 5.00 
Casale Monferrato 5.03 4.97 5.51 6.72 7.22 5.89 4.08 4.57 4.81 6.26 6.52 5.34 
Ceresole Reale 5.74 6.22 6.81 9.26 11.11 9.83 7.64 8.74 7.98 7.84 7.51 6.35 
Cuneo 4.46 5.00 7.00 8.67 9.14 7.63 5.02 5.28 6.29 7.35 6.47 5.37 
Domodossola 5.07 5.23 6.61 9.05 11.12 9.26 7.84 8.47 7.58 7.44 7.79 5.46 
Formazza 7.48 7.59 7.84 9.34 11.13 10.36 9.27 10.07 8.74 8.85 8.83 7.73 
Gressoney-D'Ejola 6.74 7.21 7.62 9.84 12.28 11.62 9.32 10.22 7.98 7.87 7.99 7.42 
10 Lago Cignana 7.82 8.01 8.15 9.19 11.48 10.76 8.50 9.78 8.11 7.93 8.22 7.78 
11 Lago Malciaussia 6.53 6.69 7.47 9.62 11.69 10.53 6.90 8.37 7.70 8.19 7.43 6.83 
12 Lago Moncenisio 6.31 6.18 6.43 7.49 9.96 9.06 6.60 7.17 7.34 7.19 6.91 6.48 
13 Lago Piastra 5.10 5.16 6.63 9.04 9.50 7.71 5.35 5.95 6.46 7.42 6.62 5.55 
14 Lago Saretto 5.48 5.96 6.87 8.74 9.61 9.47 6.35 6.88 7.11 7.84 7.02 6.52 
15 Lago Toggia 10.15 10.12 9.59 10.68 12.26 12.23 10.48 11.94 10.09 9.84 11.22 10.24 
16 Lombriasco 4.24 4.36 5.08 7.05 8.22 6.51 4.05 5.31 5.36 5.68 5.82 4.49 
17 Moncalieri 3.91 4.58 5.41 7.61 9.08 7.28 5.18 5.45 5.59 6.26 5.63 4.80 
18 Oropa 5.29 6.39 7.98 11.67 13.53 11.48 8.41 9.03 8.44 8.08 7.56 5.45 
19 Ovada 5.17 5.29 5.60 6.41 5.97 4.16 2.80 3.35 4.66 6.50 7.66 5.83 
20 Perreres 7.19 7.31 7.29 8.73 11.11 10.67 8.92 9.62 8.28 7.87 8.07 7.29 
21 Rosone 5.12 5.72 6.36 9.58 11.60 10.15 7.28 8.54 7.26 7.60 6.55 5.53 
22 Torino 4.10 4.73 5.73 8.26 9.85 8.29 5.54 5.88 6.09 6.57 5.95 4.97 
23 Ussin 6.13 6.10 6.50 7.79 10.34 9.74 7.77 8.41 7.29 7.84 7.29 6.26 
24 Valgrisenche 7.61 7.66 8.02 8.57 8.79 9.59 7.62 8.62 7.69 7.88 8.41 7.84 
25 Varallo 3.88 4.39 6.11 9.26 11.04 9.87 7.62 7.64 7.15 7.39 6.96 4.42 
26 Vercelli 4.69 4.61 5.51 7.29 8.12 6.73 4.54 5.48 5.43 6.62 6.98 5.22 
 <# events/month> 5.71 5.99 6.72 8.59 10.06 8.96 6.75 7.52 7.06 7.45 7.32 6.18 

In spite of these problems, with almost similar characters for all the sites, a simple partition of the whole year into two contiguous periods was chosen. Accordingly, a first period, denoted as the ‘cold season’, including the four months (from December to March) in which lowest precipitation day frequencies are observed ‘on average’ and a second period, namely, the ‘warm season’, covering the remaining months (from April to November), were selected. This choice is supported by Figure 3, which reports, for all the considered stations, the average number of events per month, <# events/month>, during the ‘cold season’ and the ‘warm season’. As can be observed, all the months of the ‘cold season’ show lower average number of events per month than the months of the ‘warm season’.

Figure 3

Average number of events per month, <# events/month>, for the ‘cold season’ and for the ‘warm season’, of the selected 26 stations.

Figure 3

Average number of events per month, <# events/month>, for the ‘cold season’ and for the ‘warm season’, of the selected 26 stations.

Then, for each season, a statistical inference was performed by fitting the IT data frequencies through the three-parameter Lerch (Lch) distribution. For the sake of comparison, the statistical inference was also applied to the whole year data set (Y).

Before performing the statistical inference, the occurrence of a change-point in some average annual IT data series, <IT>, was detected. For this purpose, a number of different tests can be applied (Baiamonte & D'Asaro 2016; Pohlert 2018), such as either the Buishand test or the standard normal homogeneity test. However, for its simplicity and for its wide application in hydrology (Zhang et al. 2009), the Pettitt test (Pettitt 1979) was considered in this study. The Pettitt test localizes the year when most likely a breakpoint (discontinuity or inhomogeneity) occurs. Detecting the change-point allows dividing the entire data series into two intervals, before and after the change-point, forming two homogeneous groups.

The null hypothesis of the test is that the two sequences of random variables, (x1,x2, … xt) and (xt+1, xt+2, …, xT), have a common distribution function, against the alternative hypothesis that a change-point exists. The non-parametric statistic Ut,T is given by:  
formula
(1a)
where  
formula
(1b)
For p ≤ 0.05, the significance probability of KT is approximated by:  
formula
(1c)
If ρ is smaller than the specific significance level, p, the null hypothesis is rejected, then a change point exists, and the time series is divided into two parts at the location of the change-point.

For changing in one direction, the statistics KT+ = max {Ut,T} and KT = −min {Ut,T} were calculated, with KT= max {KT+, KT}, indicating that large KT+ corresponds to a shift down in level, whereas large KT corresponds to a shift up in level from the beginning of the series, respectively.

Table 3 reports the statistics KT+, KT, KT, the p-value significance level, the year of changing and the shift upward or downward for the considered 26 stations.

Table 3

P value, statistics, KT+, KT, KT, and year of changing, of the selected 26 stations

Station p value KT+ KT KT Year of changing Shift 
Alpe Cavalli 1.000 165 198 198 1941 – 
Aosta 0.519 414 129 414 1925 – 
Bra 0.065 200 659 659 1980 – 
Casale Monferrato 0.108 609 508 609 1950 – 
Ceresole Reale 0.278 444 275 444 1949 – 
Cuneo 0.000* 50 1,321 1,321 1964 Upward 
Domodossola 0.582 94 396 396 1937 – 
Formazza 0.068 645 274 645 1945 – 
Gressoney-D'Ejola 0.020* 73 654 654 1958 Upward 
10 Lago Cignana 0.200 82 471 471 1940 – 
11 Lago Malciaussia 0.164 29 414 414 1996 – 
12 Lago Moncenisio 0.000* 1,094 41 1,094 1949 Downward 
13 Lago Piastra 0.000* 1,404 120 1,404 1970 Downward 
14 Lago Saretto 0.005* 829 829 1967 Downward 
15 Lago Toggia 0.480 266 338 338 2002 – 
16 Lombriasco 0.031* 611 171 611 1952 Downward 
17 Moncalieri 0.090 106 627 627 1978 – 
18 Oropa 0.071 190 609 609 1972 – 
19 Ovada 0.155 78 524 524 1973 – 
20 Perreres 0.133 502 248 502 1976 – 
21 Rosone 1.000 171 183 183 1944 – 
22 Torino 0.143 165 579 579 1979 – 
23 Ussin 0.010* 24 666 666 1947 Upward 
24 Valgrisenche 0.001* 114 986 986 1940 Upward 
25 Varallo 0.217 344 500 500 1957 – 
26 Vercelli 0.091 70 546 546 1937 – 
Station p value KT+ KT KT Year of changing Shift 
Alpe Cavalli 1.000 165 198 198 1941 – 
Aosta 0.519 414 129 414 1925 – 
Bra 0.065 200 659 659 1980 – 
Casale Monferrato 0.108 609 508 609 1950 – 
Ceresole Reale 0.278 444 275 444 1949 – 
Cuneo 0.000* 50 1,321 1,321 1964 Upward 
Domodossola 0.582 94 396 396 1937 – 
Formazza 0.068 645 274 645 1945 – 
Gressoney-D'Ejola 0.020* 73 654 654 1958 Upward 
10 Lago Cignana 0.200 82 471 471 1940 – 
11 Lago Malciaussia 0.164 29 414 414 1996 – 
12 Lago Moncenisio 0.000* 1,094 41 1,094 1949 Downward 
13 Lago Piastra 0.000* 1,404 120 1,404 1970 Downward 
14 Lago Saretto 0.005* 829 829 1967 Downward 
15 Lago Toggia 0.480 266 338 338 2002 – 
16 Lombriasco 0.031* 611 171 611 1952 Downward 
17 Moncalieri 0.090 106 627 627 1978 – 
18 Oropa 0.071 190 609 609 1972 – 
19 Ovada 0.155 78 524 524 1973 – 
20 Perreres 0.133 502 248 502 1976 – 
21 Rosone 1.000 171 183 183 1944 – 
22 Torino 0.143 165 579 579 1979 – 
23 Ussin 0.010* 24 666 666 1947 Upward 
24 Valgrisenche 0.001* 114 986 986 1940 Upward 
25 Varallo 0.217 344 500 500 1957 – 
26 Vercelli 0.091 70 546 546 1937 – 

The shift for the stations for which the null hypothesis is rejected is also reported. The symbol *indicates the p value is less than 0.05.

The results showed that the null hypothesis was rejected (i.e., no change-point occurs) only for eight stations out of the selected 26 ones. Thus, 18 data-series were considered as a whole, whereas in the remaining eight series, the period before the changing year (detected by the Pettitt test) was eliminated. In those cases, only the more recent homogeneous period was considered for the application of the Lerch distribution.

The cases where the null hypothesis was rejected are depicted in Figure 4, showing the <IT> time-series, the location of the change-point, and the disregarded period highlighted by a shaded area. As can be observed, the upward or downward shift agrees with the results of the Pettitt test reported in Table 3. However, it should be emphasized that, as recently observed by Serinaldi et al. (2018), such change-points should be accepted as physically meaningful if they can be related with a predictable process based on theoretical models and/or well identified physical dynamics justifying causality. In our case, neither of the occurrences was possible to find.

Figure 4

Average annual IT data, <IT>, location of the change-point (vertical straight line), and the disregarded period highlighted by a shaded area.

Figure 4

Average annual IT data, <IT>, location of the change-point (vertical straight line), and the disregarded period highlighted by a shaded area.

METHODS

Let be a time-series of precipitation data spaced at a uniform time-scale, τ, e.g., day; a day k is a ‘rainy’ day if , otherwise k is a ‘no-rainy’ (or dry) day; is any precipitation threshold.

From the H series (of size N), the inter-arrival time series, , representing the succession of the times elapsed between two subsequent rainy days, could be easily derived; IT size, M, equals the number of rainy days. Accordingly, an isolated rainy day is identified by the conditions ITj > 1 and ITj+1 > 1, whereas a wet spell, WS, of duration n+ 1 is identified by any sequence, in IT, of n consecutive 1 values. Moreover, by neglecting any ITk = 1, the DS sequence, can be easily obtained, by DSj = ITj 1; then, if r indicates the frequency of ITs equal to 1, the size L of DS (or, equivalently, of WS) results as: . Figure 2 reports an example of the terms and sequences defined above.

Sample means of IT, DS and WS (in τ units) can be easily expressed by the following relationships:  
formula
(2)
With the hypothesis that ITs are independent random variables with identical distribution (i.i.d.), each sample series can be interpreted as a realization of a renewal process, the main feature of which is that it probabilistically re-starts at each arrival epoch, independently of the previous history (the so-called ‘renewal property’, Cox 1970); consequently, WS must be geometrically distributed:  
formula
(3)
with parameter (1 − p1), where p1 equals the probability that IT= 1.
Furthermore, the occurrence probability of DS of size k, could be easily derived from IT distribution as follows:  
formula
(4)
where denotes the probability that DS equals k.
Conversely, if WS is geometrically distributed with parameter (1 p), the probability distribution of IT can be easily deduced as follows:  
formula
(5a)
and  
formula
(5b)
The empirical frequency distributions of IT (or, equivalently of DS) at the daily scale, usually exhibited high standard deviation and skewness, and a monotonic decreasing pattern of frequencies with a slowly decaying tail. Moreover, IT distributions showed a peculiar ‘frequency drop’ from the modal value at IT= 1 to the value at IT= 2; this drop is generally absent in the DS distributions, since it properly accounts for the natural tendency of precipitation to produce clusters.

In order to reproduce the main features of the above-mentioned empirical IT distributions, Agnese et al. (2014) suggested applying the three-parameter Lerch distribution.

The three-parameter Lerch probability distribution represents the most general form of a family of discrete probability distributions extensively used in the scientific literature (Aksenov & Savageau 2005; Kemp 2006; Gupta et al. 2008). The probability mass function (pmf) of the ‘family’ of Lerch distributions is:  
formula
(6)
where θ, s and a are the distribution parameters, with the constraints indicated in Table 4, and Φ(., ., .) is the Lerch transcendent function (Erdélyi et al. 1981).
Table 4

Lerch family of probability distributions with the corresponding parameters domains, and the theoretical means (arithmetic, AM, geometric, GM, and/or harmonic, HM) that match (x) or not (–) with the empirical ones, according to the MLE method

ID Probability distribution Code θ s a AM GM HM 
3-par Lerch Lch 0 < θ < 1 >0* > −1 
2-par polylogarithmic Poly 0 < θ < 1 >0* – 
1-par logarithmic Log 0 < θ < 1 – – 
1-par geometric Geo 0 < θ < 1 – – 
2-par extended log Ext Log 0 < θ < 1 > −1 – 
2-par Hurwitz Hur >1 > −1 – 
1-par zeta (Zift) Zeta >1 – – 
ID Probability distribution Code θ s a AM GM HM 
3-par Lerch Lch 0 < θ < 1 >0* > −1 
2-par polylogarithmic Poly 0 < θ < 1 >0* – 
1-par logarithmic Log 0 < θ < 1 – – 
1-par geometric Geo 0 < θ < 1 – – 
2-par extended log Ext Log 0 < θ < 1 > −1 – 
2-par Hurwitz Hur >1 > −1 – 
1-par zeta (Zift) Zeta >1 – – 

*The condition s > 0 allows obtaining monotonically decreasing pmf, with mode equal to 1.

In the most general form, the distribution reported in Equation (5), hereafter denoted as Lch, depends on the three parameters θ, s and a (Table 4). Positive s values ensure that the probability mass function (pmf) is a monotonically decreasing function with mode X= 1.

A special case of the Lch distribution is the polylogarithmic (hereafter referred to as Poly) distribution (Kemp 2006), which can be obtained by setting a = 0 into Equation (5):  
formula
(7)
where is the polylogarithm function (Jonquière 1889). Good (1953) and Kemp (2006) suggested using the Poly distribution in order to fit the observed population frequencies of species per genus. For s= 1, Equation (6) reduces to the logarithm-series distribution (Log), widely used in the past for describing DS frequencies (Chatfield 1966; Buishand 1977, 1978; Kottegoda & Rosso 1997):  
formula
(8)
where
For s = 0, Equation (6) reduces to the well-known geometric series (hereafter referred to as Geo):  
formula
(9)

Therefore, the distributions expressed by Equations (8) and (9) are different forms (particular cases) of Lch distribution.

In Table 4, the Lerch family of the probability distributions and the corresponding parameters domain are resumed (ID = 1–7) (Zörnig & Altmann 1995).

In order to perform the statistical inference of the previously introduced pmfs, the well-known maximum likelihood estimation (MLE) method was used. The analytical solution of the MLE method for the general three-parameter Lch implies the equality of the arithmetic, geometric and harmonic expectations to the corresponding sample means (Gupta et al. 2008; Agnese et al. 2014). Obviously, two-parameter and one-parameter distributions of the Lerch family do not fulfil all equality conditions; in Table 4, conditions that are satisfied or not satisfied by each distribution are also indicated.

The classical χ2 procedure is not appropriate to test the goodness-of-fit of long-tailed distributions with several small classes (less than five elements) and strong asymmetry, as in our case, as pointed out by Martínez-Rodríguez et al. (2011). Therefore, the evaluation of the goodness-of-fit was performed by operating with artificial IT synthetic samples generated via Lerch distribution and numerically reconstructing the null hypothesis of χ2 via Monte Carlo simulation. A set of 2,000 replicates of the IT sample was generated by randomly sampling from the theoretical distribution obtained by inference, and the associated p-value was defined as the fraction of the 2,000 replicates for which reconstructed χj2 values (with j = 1, 2, …, 2,000) are greater than χref2, obtained by the classical procedure (Hope 1968). Accordingly, the null hypothesis was accepted if the p-value was greater than 0.05, the value that is usually assumed as the significance level.

Moreover, in order to establish if the adoption of the general three-parameter Lch is justified with respect to the other distributions with one or two parameters (Table 4), the likelihood-ratio (LLR) test was used, which is suggested when using distributions that are particular cases (logarithmic distribution, polylogarithmic distribution, etc.) of the general one (Lerch distribution) (Wilks 1938). In fact, it is obvious that the more general model will always fit at least as well as the special cases; the LLR test allows detecting if the improvement in terms of log-likelihood endorses the introduction of the extra parameters. Each of the two competing models, the null model (ID = 2–7) and the alternative model (Lch, ID = 1), is separately fitted to the data and the log-likelihood was calculated. The test statistic D is defined as twice the log of the likelihoods ratio:  
formula
(10a)
where the null model represents a special case of the alternative model and the likelihood function is defined by the following:  
formula
(10b)
where M is the sample size. The probability distribution of the test statistic is approximately a chi-squared distribution with degrees of freedom equal to dfaltdfnull, i.e., the number of free parameters of the alternative and the null model, respectively.

For the whole year data set and for all year records, Table 5 reports the comparison of the statistic, D, calculated for the alternative model (Lch) and that calculated for the null models (ID = 2–7). The critical chi-square statistic, at the significant level, p = 0.05, is also reported. The statistic D appears greater than the corresponding critical chi squared, in all the series and in all probability distributions, justifying the use of the three-parameter Lch.

Table 5

Statistic D calculated for the alternative model (Lch) compared to the null models represented by the distributions, ID = 2–7, reported in Table 4, of the selected 26 stations. The critical chi-square statistic for p = 0.05, corresponding to dfalt – dfnull is also reported

Station ID
 
DLch/Poly DLch/Log DLch/Geo DLch/Ext Log DLch/Hur DLch/Zeta 
Alpe Cavalli 363.01 905.63 10,174.71 116.64 725.89 849.32 
Aosta 187.07 193.80 4,784.12 152.20 693.19 1,389.94 
Bra 261.44 383.02 5,342.38 135.69 735.11 1,015.48 
Casale Monferrato 270.65 365.52 5,322.77 152.97 765.52 1,084.49 
Ceresole Reale 295.82 567.64 8,090.65 131.13 705.19 966.14 
Cuneo 128.31 167.28 5,404.24 80.19 566.42 1,192.70 
Domodossola 393.96 823.13 8,964.69 151.60 830.80 996.84 
Formazza 324.69 832.04 11,006.51 102.45 730.87 928.62 
Gressoney-D'Ejola 160.11 371.18 8,892.12 62.86 520.08 945.13 
10 Lago Cignana 211.45 404.17 8,632.62 98.72 610.04 1,025.60 
11 Lago Malciaussia 229.22 517.31 7,543.48 82.49 578.79 792.92 
12 Lago Moncenisio 197.24 253.22 5,883.40 129.67 611.40 1,128.65 
13 Lago Piastra 189.32 282.61 6,304.57 104.36 649.05 1,147.07 
14 Lago Saretto 193.68 334.02 7,377.66 95.83 618.40 1,074.80 
15 Lago Toggia 280.53 849.05 12,580.02 77.58 602.03 765.72 
16 Lombriasco 194.24 300.33 4,725.47 94.64 599.61 868.17 
17 Moncalieri 232.03 367.53 5,680.08 111.35 696.27 1,003.16 
18 Oropa 383.96 1,035.83 11,369.72 107.58 773.40 894.23 
19 Ovada 306.76 457.15 5,375.26 151.93 767.58 969.32 
20 Perreres 188.86 360.69 8,166.55 88.13 578.96 1,016.16 
21 Rosone 249.61 572.28 6,968.64 82.63 589.00 739.10 
22 Torino 263.82 492.53 6,567.87 104.61 716.48 950.44 
23 Ussin 180.54 316.97 6,760.57 88.85 569.42 982.46 
24 Valgrisenche 195.13 321.18 8,302.92 104.86 634.70 1,201.97 
25 Varallo 266.98 794.34 8,796.46 63.09 591.95 706.55 
26 Vercelli 253.49 373.62 5,206.90 131.35 690.74 951.08 
χ2 3.841 5.991 5.991 3.841 3.841 5.991 
dfnulldfalt 
Station ID
 
DLch/Poly DLch/Log DLch/Geo DLch/Ext Log DLch/Hur DLch/Zeta 
Alpe Cavalli 363.01 905.63 10,174.71 116.64 725.89 849.32 
Aosta 187.07 193.80 4,784.12 152.20 693.19 1,389.94 
Bra 261.44 383.02 5,342.38 135.69 735.11 1,015.48 
Casale Monferrato 270.65 365.52 5,322.77 152.97 765.52 1,084.49 
Ceresole Reale 295.82 567.64 8,090.65 131.13 705.19 966.14 
Cuneo 128.31 167.28 5,404.24 80.19 566.42 1,192.70 
Domodossola 393.96 823.13 8,964.69 151.60 830.80 996.84 
Formazza 324.69 832.04 11,006.51 102.45 730.87 928.62 
Gressoney-D'Ejola 160.11 371.18 8,892.12 62.86 520.08 945.13 
10 Lago Cignana 211.45 404.17 8,632.62 98.72 610.04 1,025.60 
11 Lago Malciaussia 229.22 517.31 7,543.48 82.49 578.79 792.92 
12 Lago Moncenisio 197.24 253.22 5,883.40 129.67 611.40 1,128.65 
13 Lago Piastra 189.32 282.61 6,304.57 104.36 649.05 1,147.07 
14 Lago Saretto 193.68 334.02 7,377.66 95.83 618.40 1,074.80 
15 Lago Toggia 280.53 849.05 12,580.02 77.58 602.03 765.72 
16 Lombriasco 194.24 300.33 4,725.47 94.64 599.61 868.17 
17 Moncalieri 232.03 367.53 5,680.08 111.35 696.27 1,003.16 
18 Oropa 383.96 1,035.83 11,369.72 107.58 773.40 894.23 
19 Ovada 306.76 457.15 5,375.26 151.93 767.58 969.32 
20 Perreres 188.86 360.69 8,166.55 88.13 578.96 1,016.16 
21 Rosone 249.61 572.28 6,968.64 82.63 589.00 739.10 
22 Torino 263.82 492.53 6,567.87 104.61 716.48 950.44 
23 Ussin 180.54 316.97 6,760.57 88.85 569.42 982.46 
24 Valgrisenche 195.13 321.18 8,302.92 104.86 634.70 1,201.97 
25 Varallo 266.98 794.34 8,796.46 63.09 591.95 706.55 
26 Vercelli 253.49 373.62 5,206.90 131.35 690.74 951.08 
χ2 3.841 5.991 5.991 3.841 3.841 5.991 
dfnulldfalt 

Table 5 also shows that the extended log (hereafter referred to as Ext log) shows the best performance (lowest D values) among the other distributions, whereas the one-parameter Geo shows the worst performance (greatest D values).

For the purposes of this work, the three-parameter Lch distribution was fitted to the IT data series, which, according to Equations (2) and (3), allows the DS and WS distributions to be derived. Both of the latter could also be derived by directly fitting Poly and Geo distributions to the corresponding data series. Indeed, if θ and s are the parameters of the Poly distribution for DS series and p is the parameter of the Geo distribution for WS series, then the IT distribution can also be derived according to Equation (4):  
formula
(11a)
 
formula
(11b)
The joint use of Equations (11a) and (11b) is actually equivalent to the three-parameter Lch pmf (Equation (5)) and requires the same number of parameters (θ, s, p), with p instead of a. Interestingly, θ and s values associated with the Poly distribution (for DS series, Ferraris et al. 2014) are very close to those associated with the Lch distribution (for IT series).

RESULTS AND DISCUSSION

As can be noticed from the comparison of the elevation data reported in Table 1 and from Figure 1, the selected stations include a wide range of altitudes (from 113 m to 2,170 m a.s.l.), which is representative of the whole region. The comparison between the average IT, <IT>, reported in Table 6, highlights substantial differences occurring between the two seasons. However, a variety of conditions are present among the stations, within the same season, as is illustrated by the minimum and maximum (shown in bold in Table 6) of the average IT values, <IT>. Table 6 also reports the starting and the ending year of the observations, which accounts for the results obtained by the application of the Pettitt test, the average cluster size, <WS>, calculated according to Equation (2), which also characterizes each station.

Table 6

Sample size, average inter-arrival time, IT, average cluster size, <WS>, and p value, for the whole year (from October to September), for the ‘cold season’ (from December to March) and for the ‘warm season’ (from April to November) of the selected 26 stations

ID Station Starting year Ending year From October to September (year)
 
From December to March (‘cold season’)
 
From April to November (‘warm season’)
 
Sample size <IT> (days) <WS> (days) p-value Sample size <IT> (days) <WS> (days) p-value Sample size <IT> (days) <WS> (days) p-value 
Alpe Cavalli 1929 2009 8,693 3.32 2.27 0.158 2,235 4.34 2.10 0.748 6,573 3.00 2.32 0.202 
Aosta 1919 2009 6,427 5.04 1.67 0.112 1,912 5.67 1.72 0.802 4,593 4.88 1.64 0.006 
Bra 1919 2009 5,942 5.39 1.79 0.169 1,615 6.32 1.86 0.530 4,408 5.20 1.75 0.200 
Casale Monferrato 1919 2009 6,071 5.19 1.78 0.104 1,842 5.60 1.87 0.248 4,310 5.10 1.74 0.345 
Ceresole Reale 1929 2009 8,002 3.71 2.05 0.023 2,085 4.75 1.93 0.225 5,974 3.38 2.09 0.121 
Cuneo 1964 2009 3,783 4.34 1.83 0.497 767 6.57 1.62 0.523 2,311 5.07 1.56 0.070 
Domodossola 1919 2009 8,282 3.91 2.10 0.073 2,003 5.34 2.05 0.910 6,385 3.52 2.10 0.059 
Formazza 1919 2008 9,689 3.29 2.22 0.022 2,781 3.82 2.19 0.142 7,029 3.12 2.21 0.375 
Gressoney-D'Ejola 1958 2009 5,553 3.26 2.07 0.123 1,516 4.13 1.88 0.302 4,085 3.08 2.04 0.627 
10 Lago Cignana 1929 2009 8,650 3.38 2.03 0.188 2,599 3.79 2.06 0.888 6,106 3.26 2.02 0.140 
11 Lago Malciaussia 1937 2009 7,167 3.62 2.08 0.041 1,973 4.34 1.93 0.174 5,260 3.41 2.13 0.718 
12 Lago Moncenisio 1949 2009 5,110 4.22 1.80 0.108 1,663 4.31 1.86 0.630 4,245 3.54 1.90 0.163 
13 Lago Piastra 1970 2009 2,848 5.00 1.60 0.221 925 4.96 2.12 0.421 2,732 3.65 2.02 0.425 
14 Lago Saretto 1967 2009 3,506 4.09 1.81 0.283 1,114 4.28 2.01 0.757 3,070 3.47 2.02 0.086 
15 Lago Toggia 1932 2009 10,023 2.75 2.42 0.093 3,135 2.96 2.50 0.457 6,991 2.70 2.37 0.039 
16 Lombriasco 1952 2009 3,822 5.33 1.74 0.002 1,952 5.96 1.90 0.439 2,973 4.79 1.77 0.550 
17 Moncalieri 1919 2009 6,305 5.12 1.80 0.142 1,553 6.44 1.86 0.836 4,826 4.80 1.77 0.082 
18 Oropa 1919 2009 9,299 3.35 2.27 0.021 2,150 4.71 2.02 0.228 7,205 2.99 2.34 0.110 
19 Ovada 1919 2009 5,727 5.25 1.84 0.523 1,918 5.12 1.96 0.402 3,876 5.41 1.77 0.948 
20 Perreres 1929 2009 8,364 3.48 2.00 0.056 2,386 4.09 1.94 0.310 6,110 3.29 2.01 0.080 
21 Rosone 1938 2009 6,549 3.93 2.06 0.085 1,615 5.28 1.91 0.267 5,053 3.52 2.11 0.192 
22 Torino 1919 2009 6,793 4.75 1.89 0.469 1,646 6.21 1.91 0.697 5,234 4.39 1.86 0.685 
23 Ussin 1947 2007 5,621 3.89 1.91 0.226 1,427 5.13 1.74 0.471 3,974 3.73 1.89 0.397 
24 Valgrisenche 1940 2009 6,951 3.55 1.96 0.012 1,889 4.21 1.97 0.433 4,557 3.69 1.80 0.036 
25 Varallo 1919 2009 7,696 3.65 2.18 0.003 1,621 5.48 1.98 0.155 6,172 3.22 2.22 0.037 
26 Vercelli 1929 2009 5,885 4.99 1.83 0.045 1,613 5.81 1.91 0.241 4,360 4.75 1.79 0.405 
ID Station Starting year Ending year From October to September (year)
 
From December to March (‘cold season’)
 
From April to November (‘warm season’)
 
Sample size <IT> (days) <WS> (days) p-value Sample size <IT> (days) <WS> (days) p-value Sample size <IT> (days) <WS> (days) p-value 
Alpe Cavalli 1929 2009 8,693 3.32 2.27 0.158 2,235 4.34 2.10 0.748 6,573 3.00 2.32 0.202 
Aosta 1919 2009 6,427 5.04 1.67 0.112 1,912 5.67 1.72 0.802 4,593 4.88 1.64 0.006 
Bra 1919 2009 5,942 5.39 1.79 0.169 1,615 6.32 1.86 0.530 4,408 5.20 1.75 0.200 
Casale Monferrato 1919 2009 6,071 5.19 1.78 0.104 1,842 5.60 1.87 0.248 4,310 5.10 1.74 0.345 
Ceresole Reale 1929 2009 8,002 3.71 2.05 0.023 2,085 4.75 1.93 0.225 5,974 3.38 2.09 0.121 
Cuneo 1964 2009 3,783 4.34 1.83 0.497 767 6.57 1.62 0.523 2,311 5.07 1.56 0.070 
Domodossola 1919 2009 8,282 3.91 2.10 0.073 2,003 5.34 2.05 0.910 6,385 3.52 2.10 0.059 
Formazza 1919 2008 9,689 3.29 2.22 0.022 2,781 3.82 2.19 0.142 7,029 3.12 2.21 0.375 
Gressoney-D'Ejola 1958 2009 5,553 3.26 2.07 0.123 1,516 4.13 1.88 0.302 4,085 3.08 2.04 0.627 
10 Lago Cignana 1929 2009 8,650 3.38 2.03 0.188 2,599 3.79 2.06 0.888 6,106 3.26 2.02 0.140 
11 Lago Malciaussia 1937 2009 7,167 3.62 2.08 0.041 1,973 4.34 1.93 0.174 5,260 3.41 2.13 0.718 
12 Lago Moncenisio 1949 2009 5,110 4.22 1.80 0.108 1,663 4.31 1.86 0.630 4,245 3.54 1.90 0.163 
13 Lago Piastra 1970 2009 2,848 5.00 1.60 0.221 925 4.96 2.12 0.421 2,732 3.65 2.02 0.425 
14 Lago Saretto 1967 2009 3,506 4.09 1.81 0.283 1,114 4.28 2.01 0.757 3,070 3.47 2.02 0.086 
15 Lago Toggia 1932 2009 10,023 2.75 2.42 0.093 3,135 2.96 2.50 0.457 6,991 2.70 2.37 0.039 
16 Lombriasco 1952 2009 3,822 5.33 1.74 0.002 1,952 5.96 1.90 0.439 2,973 4.79 1.77 0.550 
17 Moncalieri 1919 2009 6,305 5.12 1.80 0.142 1,553 6.44 1.86 0.836 4,826 4.80 1.77 0.082 
18 Oropa 1919 2009 9,299 3.35 2.27 0.021 2,150 4.71 2.02 0.228 7,205 2.99 2.34 0.110 
19 Ovada 1919 2009 5,727 5.25 1.84 0.523 1,918 5.12 1.96 0.402 3,876 5.41 1.77 0.948 
20 Perreres 1929 2009 8,364 3.48 2.00 0.056 2,386 4.09 1.94 0.310 6,110 3.29 2.01 0.080 
21 Rosone 1938 2009 6,549 3.93 2.06 0.085 1,615 5.28 1.91 0.267 5,053 3.52 2.11 0.192 
22 Torino 1919 2009 6,793 4.75 1.89 0.469 1,646 6.21 1.91 0.697 5,234 4.39 1.86 0.685 
23 Ussin 1947 2007 5,621 3.89 1.91 0.226 1,427 5.13 1.74 0.471 3,974 3.73 1.89 0.397 
24 Valgrisenche 1940 2009 6,951 3.55 1.96 0.012 1,889 4.21 1.97 0.433 4,557 3.69 1.80 0.036 
25 Varallo 1919 2009 7,696 3.65 2.18 0.003 1,621 5.48 1.98 0.155 6,172 3.22 2.22 0.037 
26 Vercelli 1929 2009 5,885 4.99 1.83 0.045 1,613 5.81 1.91 0.241 4,360 4.75 1.79 0.405 

The bold and underlined p-values are the ones lower than the assumed significance 0.05 level. Maximum and minimum averages are in bold. The bold and underlined stations are the ones with the null hypothesis of the Pettitt test verified (p = 0.05).

In particular, Table 6 shows a minimum <IT> value of 2.7 days in the ‘warm season’ for Lago Toggia station, one of the more northern stations, and a maximum one of 6.57 days in the ‘cold season’ for Cuneo station, one of the more southern ones. As can be observed, the p-values associated with the χ2 test are greater than 0.05, for the ‘cold season’, for all the IT data sets, whereas, for the ‘warm season’ and for the all-years record, only 22 and 18 data sets satisfy the χ2 test, respectively.

For the 26 considered stations, Figure 5 illustrates the comparison between the observed frequencies and the theoretical probabilities calculated according to the Lch distribution. As can be observed, the dots are close to the line of perfect agreement, especially for high probabilities, showing evidence of the suitability of Lch in describing empirical ITs. In particular, the goodness-of-fit for IT = 1 also emphasized the Lch capability in describing the geometrical distributed WS (Equations (3), (5b) and (11b)).

Figure 5

Comparison between observed frequencies and theoretical probabilities calculated according to the Lch distribution, for the 26 considered stations.

Figure 5

Comparison between observed frequencies and theoretical probabilities calculated according to the Lch distribution, for the 26 considered stations.

Figure 6 reports two examples of IT data fitting with the Lch distribution for Lago Toggia and Torino stations that have different rainfall regimes, as can be observed in Table 6 from the corresponding <IT> and <WS> values. The comparison is performed in terms of cumulative distribution function, derived by the Lch pmf, according to:  
formula
(12)
The cumulative Lch distributions shown in Figure 6 refer to the whole year (from October to September), and are associated with p-values equal to 0.093 and 0.469, for Lago Toggia and Torino, respectively. It is interesting to observe that a good agreement between empirical and theoretical distribution occurs in a wide range of probability, ranging three orders of magnitude. A good fitting also occurs for the low-level probability of exceedance (very high IT values), which is interesting from a practical point of view.
Figure 6

Log–log plot displaying empirical cumulative distributions of the inter-arrival times (IT) and Lerch cumulative distributions (Lago Toggia and Torino stations, whole year data set).

Figure 6

Log–log plot displaying empirical cumulative distributions of the inter-arrival times (IT) and Lerch cumulative distributions (Lago Toggia and Torino stations, whole year data set).

Figure 6 also illustrates the fitting of the more common polylogarithmic (Poly) and of the logarithmic distribution (Log), showing that with respect to the Lch distribution, they are not able to reproduce the observed IT frequencies.

The cumulative Lch distributions for the whole year data set, are compared with those corresponding to the ‘cold’ and to the ‘warm’ seasons, in order to demonstrate the seasonality effect, in Figure 7, for Lago Toggia and Torino stations. For both stations and with respect to the whole year Lch distributions, the figure shows greater differences in probability for the ‘cold season’ than for the ‘warm season’.

Figure 7

Log–log plots displaying the Lerch cumulative distributions of the inter-event times (IT) (Lago Toggia and Torino stations' three data sets: ‘cold season’, ‘warm season’, whole year, Y).

Figure 7

Log–log plots displaying the Lerch cumulative distributions of the inter-event times (IT) (Lago Toggia and Torino stations' three data sets: ‘cold season’, ‘warm season’, whole year, Y).

With reference to the whole year data set, in order to synthetize results of the statistical inference via the Lerch distribution, for the all considered stations, Figure 8(a) compares the observed WS frequency distribution with the geometric distribution derived according to Equation (2), with parameter 1 − p1 ≡ P{IT >1}. Regarding the mentioned series of Lago Toggia and Torino stations, Figure 8(a) shows that the geometric distribution fits the empirical one well. Analogously, Figure 8(b) illustrates a good agreement of the DS empirical distribution with the theoretical polylogarithmic one derived from IT theoretical distribution, according to Equation (3). Thus, Figures 8(a) and 8(b) together, confirm that both WS and DS distributions can be derived from IT distribution fitted with the three-parameter Lerch. The good agreement between the geometric WS distribution and the empirical one, illustrated in Figure 8(a), suggests that the assumption that ITs are independent random variables with identical distribution (i.i.d.) is well founded.

Figure 8

(a) Empirical and theoretical wet spell, WS, (geometric distribution) and (b) empirical and theoretical dry spell, DS, (polylogarithmic distribution) (Lago Toggia and Torino stations).

Figure 8

(a) Empirical and theoretical wet spell, WS, (geometric distribution) and (b) empirical and theoretical dry spell, DS, (polylogarithmic distribution) (Lago Toggia and Torino stations).

The i.i.d. hypothesis is also supported by the agreement between the frequencies of the sum of two observed IT values, randomly sampled, with those derived by the two-fold convolution of the Lch distribution with itself, as follows:  
formula
(13)
As an example, for Bra station and for the whole year data set, this agreement can be observed in Figure 9.
Figure 9

Frequency of the sum of two observed IT values, IT1 + IT2, randomly sampled, and the fitting theoretical distribution (convolution of the Lerch distribution with itself) obtained by Equation (13).

Figure 9

Frequency of the sum of two observed IT values, IT1 + IT2, randomly sampled, and the fitting theoretical distribution (convolution of the Lerch distribution with itself) obtained by Equation (13).

Furthermore, the distribution of sequences of rainy days, interrupted by 1-day DS, was derived from the Lch distribution. Define l a number of days, in which a number n of wet days and a number h of 1-day DS occur. Obviously l = n + h, with the constraints nl ≤ 2 n − 1 and 0 ≤ hn−1. For h = 0, the latter refers to an uninterrupted sequence of wet days, of probability (1 − p1p2), which differs from the WS probability that is equal to (1 − p1). The marginal probability P{n = n0} can be written as:  
formula
(14)
where is the combination of n0 − 1 elements taken h at a time, p1= P{IT = 1}, p2= P{IT = 2} and (1 − p1p2) ≡ P{IT > 2}. As can be observed in Figure 10, Equation (13) is able to reproduce wet days distribution with 1-day DS, which substantially differ from the WS distributions, for both the Lago Toggia and Torino stations.
Figure 10

Rainy days' sequences with 1-day dry spells (with gaps) for (a) Lago Toggia and (b) Torino stations. WS geometric distribution (no gaps), shown in Figure 7, is also reported for comparison.

Figure 10

Rainy days' sequences with 1-day dry spells (with gaps) for (a) Lago Toggia and (b) Torino stations. WS geometric distribution (no gaps), shown in Figure 7, is also reported for comparison.

The above described ‘wet day sequences’ can represent in a broad sense a unique perturbed period, even though just 1-day DS occur in between the same sequences. Equation (13) could have a certain interest from a hydrological point of view, if considering that 1 no-rainy day in between 2 rainy days does not significantly alter the root zone soil moisture status, mainly for clayey soils.

Regarding the parameters of the Lch distribution, for the whole year data set, the θ values varies in the interval (0.84 ÷ 0.92) and, as expected, higher values are associated with the ‘drier’ stations and lower values with ‘wetter’ ones. Moreover, the range of θ values associated with the ‘warm season’ (0.82 ÷ 0.91) is close to that of the whole year, whereas for the ‘cold season’ the range of θ values (0.90 ÷ 0.95) is narrower, if compared to that of the whole year data set. In Figure 11(a), the θ values are plotted vs. <IT>, for the all-year data set, for the ‘warm season’ and for the ‘cold season’. As can be observed, ‘θ behaves somewhat like a scale parameter’ (Kemp 2006) and can be also related to a physical parameter, the elevation, which is used as the independent variable in Figure 11(b). In particular, θ values are negatively correlated with elevation, with coefficient of determination equal to 0.86, 0.86 and 0.78 for the ‘cold season’, the ‘warm season’ and for the whole year data sets, respectively.

Figure 11

θ parameter values of the Lerch pmf (a) vs. average IT and (b) vs. elevation, for the ‘cold season’, for the ‘warm season’ and for the whole year (Y) (all 26 stations).

Figure 11

θ parameter values of the Lerch pmf (a) vs. average IT and (b) vs. elevation, for the ‘cold season’, for the ‘warm season’ and for the whole year (Y) (all 26 stations).

Figure 12 compare parameters θ, s and a, of the ‘warm season’ with those of the ‘cold season’. It is evident that θ values of the ‘warm season’ show a relationship with the corresponding θ values of the ‘cold season’, whereas s and a parameters do not show a clear behaviour, with a similar and narrow range of variability. Furthermore, no relationship of s and a parameters with other physical/climatic variables have been found. In the same figure, the corresponding values of Lch parameters obtained for the Sicily region in a previous work (Agnese et al. 2014) are reported. Similar considerations are valid for the parameters associated with the Sicily region, although the values are different, and their range of variability is wider, compared to that of the Piedmont and Aosta Valley area, likely indicating the different precipitation regime of the two areas. It is interesting that the warm season Sicily's θ, s and a values are in a completely different range. The θ values cold season also show a dfferent range.

Figure 12

Relation between ‘cold season’ and ‘warm season’ Lerch pmf parameters, θ, s and a, estimated for all 26 stations considered in this study, in Piedmont and Aosta Valley regions, and for the 55 Sicilian stations considered in Agnese et al. (2014).

Figure 12

Relation between ‘cold season’ and ‘warm season’ Lerch pmf parameters, θ, s and a, estimated for all 26 stations considered in this study, in Piedmont and Aosta Valley regions, and for the 55 Sicilian stations considered in Agnese et al. (2014).

To conclude the analysis of the results, Figure 13 shows the average cluster size <WS>, which according to the geometric distribution (see Equation (2)) is equal to 1/(1 − p1), as a function of the IT harmonic mean, for the ‘cold season’, for the ‘warm season’ and for the whole year data set. A clear relationship is detected, according to the fact that the harmonic mean enhances low IT values. Moreover, the considered seasons slightly influence this relationship.

Figure 13

Average cluster size <WS> vs. the IT harmonic mean, for the ‘cold season’, for the ‘warm season’ and for the whole year (Y), for the entire period (1919–2009) and for the considered 26 stations.

Figure 13

Average cluster size <WS> vs. the IT harmonic mean, for the ‘cold season’, for the ‘warm season’ and for the whole year (Y), for the entire period (1919–2009) and for the considered 26 stations.

An application of the described Lch pdf at the regional scale was carried out, on the basis of the good agreement between empirical and theoretical distribution. In particular, this application utilizes the good agreement in the wide range of probability values, including extremes with low-level probability of exceedance. Thus, the application was performed for a fixed long dry spell (DS = 45 days) and for a fixed wet spell (WS = 5 days), the values of which, and their associated spatial distribution, can play an important role in farm and environmental management. In fact, high duration of DS influences crops' growth, whereas the probability of occurrence of long precipitation spells is associated with environmental risks. Results are represented in terms of return period, T (years) and are illustrated in Figures 14(a) and 14(b). The expected return period was evaluated by using the known average IT values calculated for the whole year record (Y), reported in Table 4, according to . The probability was calculated for each station by Equation (3). The analysis was carried out for the whole year data set, thus separated ‘cold season’ and ‘warm season’ series (Table 4) were not considered. It is interesting to observe that the high elevation areas (compare Figure 1), are generally characterized by lower return period of WS = 5 days and higher return period of DS = 45 days, if compared with the low elevation areas. It is important to highlight that the expected return period for the above-mentioned was determined under the assumption that both WS and DS are time independent and uncorrelated random variables. In this particular case, the extended return period formulation introduced by Fernández & Salas (1999) and then by Volpi et al. (2015), which differentiated the return period as an average occurrence interval or as an average recurrence interval, according to a not time-dependent or a time-dependent condition, respectively, leads to the same results.

Figure 14

Return period, T (years), of (a) event dry spell DS ≥ 45 days and (b) event wet spell WS ≥ 5 days.

Figure 14

Return period, T (years), of (a) event dry spell DS ≥ 45 days and (b) event wet spell WS ≥ 5 days.

Interestingly, for the case of time-dependent random variables, Volpi et al. (2015) also introduced an equivalent return period (ERP), which resembles the classical definition of return period, but in the case of independence. The ERP, which for this study matches the classical return period formulation considered above, is able to control the probability of the random variable under the time-dependent condition, preserving the virtue of the classical return period formulation that is insensitive to the time-dependent condition.

CONCLUSION

The statistical analysis performed in this work demonstrated the applicability of the discrete three-parameter Lerch probability distribution to the frequency distribution of the inter-arrival time-series in the NW of Italy, characterized by precipitation regimes ranging between that of hilly areas influenced by the Tyrrhenian Sea, to that of the Po Plain and to that of the Alps. In the past, the same distribution has already been successfully tested in the Mediterranean area (Sicily island) (Agnese et al. 2014), which has very different precipitation regimes from those investigated in this study.

The analysis was successfully applied to 70–90 years' long inter-arrival time series derived from 26 daily precipitation stations records and the i.i.d. hypothesis of the IT random variable was then checked. Because of the climatic spatial variability of these regions, and of Sicily, the application of the three-parameter Lerch distribution has demonstrated its wide applicability. The spatial variability of all the three parameters of the distribution, for both ‘cold’ and ‘warm’ seasons, is smaller than in Sicily. In particular, the differences found in the θ parameter, confirm its intrinsic character of scale parameter, whereas both s and a parameters appear to be purely fitting parameters, but with markedly different ranges during the cold season, with respect to Sicily. In this paper, it is shown that the Lerch distribution successfully describes the sequences of rainy days interrupted by 1-day DS, with useful practical applications from a hydrological point of view. Moreover, the equivalence on using a single IT three-parameter Lerch distribution, in place of two distributions, namely, the DS two-parameter polylogarithmic and WS one-parameter geometric ones, is also tested. Finally, an application at the regional scale describing the probability of exceedance of a long wet spell and a long dry spell, has confirmed the value of this distribution in quantifying the wide spatial variability of the precipitation statistics in the considered area.

ACKNOWLEDGEMENT

Research was supported by PRIN 2015 project granted by MIUR (Italian Ministry for University and Research) No. 2015AKR4HX, by the H2020 project ‘ECOPOTENTIAL: Improving Future Ecosystem Benefits Through Earth Observations’, coordinated by CNR-IGG (http://www.ecopotential-project.eu) (it has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 641762), and by the CNR flag project NEXTDATA.

REFERENCES

REFERENCES
Agnese
C.
,
Baiamonte
G.
,
Cammalleri
C.
,
Cat Berro
D.
,
Ferraris
S.
&
Mercalli
L.
2012
Statistical analysis of inter-arrival times of rainfall events for Italian Sub-Alpine and Mediterranean areas
.
Advances in Science and Research
1
,
1
7
.
Doi: 10.5194/asr-1-1-2012
.
Agnese
C.
,
Baiamonte
G.
&
Cammalleri
C.
2014
Modelling the occurrence of rainy days under a typical Mediterranean climate
.
Advances in Water Research
64
,
62
76
.
http://dx.doi.org/10.1016/j.advwatres.2013.12.005
.
Agnese
C.
,
Baiamonte
G.
,
D'Asaro
F.
&
Grillone
G.
2016
Probability distribution of peak discharge at the hillslope scale generated by Hortonian runoff
.
Journal of Irrigation and Drainage E-ASCE
142
,
2
.
Doi: 10.1061/(ASCE)IR.1943-4774.0000973
.,
04015052
.
Aksenov
S. V.
&
Savageau
M. A.
2005
Some Properties of the Lerch Family of Discrete Distributions
. .
Baiamonte
G.
&
D'Asaro
F.
2016
Discussion of 'Analysis of extreme rainfall trends in Sicily for the evaluation of depth-duration-frequency curves in climate change scenarios' by Lorena Liuzzo and Gabriele Freni
.
Journal of Irrigation and Drainage E-ASCE
21
,
6
.
doi: 10.1061/(ASCE)HE.1943-5584.0001381, 07016005
.
Buishand
T. A.
1977
Stochastic Modelling of Daily Rainfall Sequences
.
H. Veenman & Zonen BV, Wageningen
, pp.
212
.
Buishand
T. A.
1978
Some remarks on the use of daily rainfall models
.
Journal of Hydrology
36
,
295
308
.
Canone
D.
,
Previati
M.
,
Bevilacqua
I.
,
Salvai
L.
&
Ferraris
S.
2015
Field measurements based model for surface irrigation efficiency assessment
.
Agricultural Water Management
156
,
30
42
.
Chatfield
C.
1966
Wet and dry spells
.
Weather
21
,
308
310
.
Cox
D.
1970
Renewal Theory
.
Methuen & Co
,
London
, pp.
142
.
de Groen
M. M.
&
Savenije
H. H.
2006
A monthly interception equation based on the statistical characteristics of daily rainfall
.
Water Resources Research
42
,
W12417
.
doi:10.1029/2006WR005013
.
Eichner
J. F.
,
Kantelhardt
J. W.
,
Bunde
A.
,
Havlin
S.
2011
The statistics of return intervals, maxima, and centennial events under the influence of long-term correlations
. In:
In Extremis: Disruptive Events and Trends in Climate and Hydrology
. (
Kropp
J.
&
Schellhuber
H.-J.
, eds),
Springer
,
Berlin, Heidelberg
, pp.
2
43
.
Erdélyi
A.
,
Magnus
W.
,
Oberhettinger
F.
&
Tricomi
F. G.
1981
Higher Transcendental Functions
, Vol.
1
.
Krieger
,
New York
, pp.
30
31
.
Feller
W.
1968
An Introduction to Probability Theory and Its Applications
, Vol.
1
.
John Wiley and Sons
,
New York
,
USA
.
Fernández
B.
&
Salas
J. D.
1999
Return period and risk of hydrologic events. I: mathematical formulation
.
Journal of Hydrologic Engineering
4
(
4
),
297
307
.
Ferraris
S.
,
Baiamonte
G.
,
Cammalleri
C.
,
Agnese
C.
,
Cat Berro
D.
&
Mercalli
L.
2014
Spatial variability of dry spells duration statistical distributions
. In:
Trending Now: Water, 7th International Scientific Conference on the Global Water and Energy Cycle
,
The Hague, The Netherlands
,
July 14–17
.
Foufoula-Georgiou
E.
&
Lettenmaier
D. P.
1986
Continuous-time versus discrete-time point process models for rainfall occurrence series
.
Water Resources Research
22
(
4
),
531
542
.
doi:10.1029/WR022i004p00531
.
Gupta
P. L.
,
Gupta
R. C.
,
Ong
S. H.
&
Srivastava
H. M.
2008
A class of Hurwitz-Lerch Zeta distributions and their applications in reliability
.
Applied Mathematics and Computation
196
,
521
531
.
doi:10.1016/j.amc.2007.06.012
.
Hope
A. C. A.
1968
A simplified Monte Carlo significance test procedure
.
Journal of the Royal Statistical Society B
30
,
582
598
.
Jonquière
A
, .
1889
Note sur la série.
Bulletin de la Société Mathématique de France
.
17
,
142
152
.
Kemp
A. W.
2006
Polylogarithmic distributions
. In:
Encyclopedia of Statistical Science
(
Kotz
S.
,
Banks
D. L.
&
Read
C. B.
eds).
John Wiley & Sons
,
New York
.
Kottegoda
N. T.
&
Rosso
R.
1997
Statistics, Probability, and Reliability for Civil and Environmental Engineers
.
McGraw-Hill
,
New York
.
Manfreda
S.
&
Rodrıguez-Iturbe
I.
2006
On the spatial and temporal sampling of soil moisture fields
.
Water Resources Research
42
,
W05409
.
doi:10.1029/2005WR004548
.
Martínez-Rodríguez
A. M.
,
Sáez-Castillo
A. J.
&
Conde-Sánchez
A.
2011
Modelling using an extended Yule distribution
.
Computational Statistics and Data Analysis
55
(
1
),
863
873
.
doi:10.1016/j.csda.2010.07.014
.
Mercalli
L.
&
Cat Berro
D.
2008
Atlante Climatico Della Valle D'Aosta
.
Società Meteorologica Subalpina
,
Moncalieri, Torino
,
Italy
, pp.
405
.
Pohlert
T.
2018
Non-Parametric Trend Tests and Change-Point Detection
.
Package ‘trend’, Version: 1.1.0, License: GPL-3. http://CRAN.R-project.org/package = trend
.
Rallo
G.
,
Baiamonte
G.
,
Juárez
J.
&
Provenzano
G.
2014
Improvement of FAO-56 model to estimate transpiration fluxes of drought tolerant crops under soil water deficit: application for olive groves
.
Journal of Irrigation and Drainage Engineering
140
(
9
),
1
8
.
10.1061/(ASCE)IR.1943-4774.0000693, A4014001
.
Rodriguez-Iturbe
I.
,
Cox
D. R.
&
Isham
V.
1987
Some models for rainfall based on stochastic point process
.
Proceedings of the Royal Society A
410
,
269
288
.
Serinaldi
F.
,
Kilsby
C. G.
&
Lombardo
F.
2018
Untenable nonstationarity: an assessment of the fitness for purpose of trend tests in hydrology
.
Advances in Water Resources
111
,
132
155
.
Strupczewski
W. G.
,
Kochanek
K.
,
Bogdanowicz
E.
&
Markiewicz
I.
2013
Inundation risk for embanked rivers
.
Hydrology and Earth System Sciences
17
,
3111
3125
.
doi:10.5194/hess-17-3111-2013
.
Volpi
E.
,
Fiori
A.
,
Grimaldi
S.
,
Lombardo
F.
&
Koutsoyiannis
D.
2015
One hundred years of return period: strengths and limitations
.
Water Resources Research
51
(
10
),
8570
8585
.
Wilks
S. S.
1938
The large-sample distribution of the likelihood ratio for testing composite hypotheses
.
The Annals of Mathematical Statistics
9
(
1
),
60
62
.
http://dx.doi.org/10.1214/aoms/1177732360
.
Zhang
W.
,
Yan
Y.
,
Zheng
J.
,
Li
L.
,
Dong
X.
&
Cai
H.
2009
Temporal and spatial variability of annual extreme water level in the Pearl River Delta region, China
.
Global and Planetary Change
69
,
35
47
.
Zörnig
P.
&
Altmann
G.
1995
Unified representation of Zipf distributions
.
Computational Statistics and Data Analysis
19
,
461
473
.
http://dx.doi.org/10.1016/0167-9473(94)00009-8
.