Final Report Hyd
Final Report Hyd
Professor G. Menduni
Ing. A. Raimondi
A.A. 2018-2019
1. Introduction ............................................................................................................................................... 3
2. Rainfall data analysis ................................................................................................................................. 4
2.1. Calculation of main sample moments ................................................................................................... 4
2.2. Frequency analysis................................................................................................................................. 7
2.3. Gumbel Probability Distribution function ........................................................................................... 16
2.4. GEV Probability Distribution function ................................................................................................. 18
2.5. Probability plotting position ................................................................................................................ 21
2.6. Depth-Duration-Frequency curves ...................................................................................................... 25
3. Geomorphological analysis...................................................................................................................... 28
4. Time of concentration ............................................................................................................................. 36
5. SCS – Curve Number Computations ........................................................................................................ 39
6. Rainfall event analysis ............................................................................................................................. 42
6.1. Return period estimation .................................................................................................................... 42
6.2. Net rainfall estimation (SCS-CN model)............................................................................................... 45
7. Hydrological models ................................................................................................................................ 49
7.1. Linear kinematic model ....................................................................................................................... 49
7.2. Linear reservoir model......................................................................................................................... 54
7.3. Nash model .......................................................................................................................................... 57
8. Critical event analysis .............................................................................................................................. 60
8.1. Linear kinematic model ....................................................................................................................... 60
8.2. Linear reservoir model......................................................................................................................... 61
9. Conclusions .............................................................................................................................................. 62
1. Introduction
In this project the aim is to calculate the amount of water that arrives in the closing point of the considered
basin due to a real event. The location is in Toscana region, not so far from the boundaries with Liguria and
Emilia-Romagna regions, actually the studied zone is in the province of Massa-Carrara (MS). The event
studied happened on 25th October 2011 and this territory was particularly interested and damaged.
Before starting all the analysis it’s important to understand the characteristics of the Teglia watershed. One
can update the perimeter of the Teglia watershed on Google Earth and check the general assessment of the
basin and the surroundings. It is possible to observe the region is quite rural and with the 3D visualization
many ridges are evident. Then it’s necessary to notice there is a dam in the middle of the basin, so this task
has to be carefully taken in account.
It’s the centre tendency of a random variable that in this case it’s the depth of the rainfall in the basin.
Variance
It’s the approximate value of how far the real values from the Mean value are.
Standard Deviation
Coefficient of Variation
It’s a standardized measure of the relationship between the mean value and the variance.
Skewness Coefficient
This value it’s a measure of the asymmetry of the distribution of the distributions of depths.
Kurtosis Coefficient
It’s a measure of how spread are the distributions of depths depending on his positive or negative value.
N 69
Hours 1h 3h 6h 12h 24h
h max (mm) 63.4 115.2 188.6 255.2 291.4
h min (mm) 15.6 22 38 56.6 62.2
Mean 35.4 60.9 82.8 111.2 145.4
µ1 (mm) 35.4 60.9 82.8 111.2 145.4
Variance 118.3 417.8 645.3 1058.6 2036.3
µ2 (mm) 118.3 417.8 645.3 1058.6 2036.3
Standard Deviation 11.0 20.6 25.6 32.8 45.5
σ1 (mm) 10.9 20.4 25.4 32.5 45.1
Coefficient of Variation 0.31 0.34 0.31 0.29 0.31
CV 0.31 0.34 0.31 0.29 0.31
µ3 313.2 6822.9 23409.3 48710.6 103294.1
µ4 33884.3 546002.9 2554809.7 7953570.9 18802696.6
Skewness 0.24 0.80 1.43 1.41 1.12
Kurtosis -0.58 0.13 3.14 4.10 1.53
Table – Calculated results from the given data in the previous Table
Maximum Annual Precipitation
1h
350,0
3h
300,0 6h
12h
250,0
24h
200,0
h (mm)
150,0
100,0
50,0
0,0
1930 1940 1950 1960 1970 1980 1990 2000 2010 2020
Year
This analysis is carried out in order to obtain the rainfall frequencies for 1, 3, 6, 12 and 24 hours. Then, each
frequency will be represented in histograms.
k = 2.N0.4
k 10.88
1hr 4.39
3hr 8.57
Δx (mm) 6hr 13.84
12hr 18.26
24hr 21.07
Table – Number of frequency classes and time intervals of each duration
1h 3h 6h 12h 24h I
15.6 22.0 38.0 56.6 62.2 1
17.0 26.0 50.0 61.0 75.6 2
1h 3h 6h 12h 24h I
17.8 29.0 52.0 64.0 84.0 3
20.0 35.0 52.0 69.8 84.6 4
20.2 39.0 54.8 72.6 88.0 5
20.2 40.0 56.0 73.2 97.0 6
20.4 40.4 56.6 73.6 98.0 7
20.4 42.0 57.2 73.8 101.0 8
20.8 43.0 57.2 75.0 102.4 9
22.4 43.6 58.0 75.6 103.0 10
23.2 43.8 59.0 78.2 106.0 11
24.0 44.6 59.4 82.6 107.8 12
24.0 44.8 60.2 83.4 109.2 13
24.6 45.0 61.2 86.8 112.8 14
25.0 45.2 65.0 87.8 114.0 15
25.0 46.0 65.6 87.8 115.0 16
25.2 46.0 67.0 89.4 115.2 17
26.6 46.2 69.4 90.4 115.4 18
27.0 47.0 69.4 90.4 119.0 19
27.8 47.8 70.0 90.4 119.2 20
30.0 48.0 70.0 92.2 120.8 21
30.0 48.0 70.0 96.0 121.0 22
31.0 48.0 70.0 96.0 122.0 23
31.4 49.0 70.2 98.8 123.6 24
31.8 49.0 71.0 99.0 124.0 25
32.0 50.0 72.0 100.0 124.0 26
32.0 50.0 72.0 100.0 124.6 27
32.2 52.0 73.6 100.4 126.0 28
32.2 52.0 74.0 100.8 127.6 29
33.0 52.8 74.0 101.0 128.0 30
33.0 53.0 74.0 102.0 129.2 31
33.4 53.4 74.2 102.0 129.4 32
33.4 53.6 75.0 102.2 133.8 33
34.5 54.8 75.0 103.0 134.4 34
35.0 56.4 76.0 107.6 136.0 35
36.0 57.4 77.0 108.0 137.3 36
36.2 58.4 77.0 108.2 144.2 37
37.0 58.8 77.4 111.8 146.0 38
37.0 59.0 77.6 112.0 146.0 39
37.2 60.0 77.6 113.0 146.4 40
37.4 60.0 78.0 113.4 146.6 41
38.0 60.4 78.2 115.6 146.6 42
38.6 61.4 82.6 117.0 146.8 43
39.0 62.4 84.0 119.0 149.2 44
39.2 63.8 86.0 119.0 150.4 45
39.2 64.2 88.8 119.2 151.6 46
40.0 65.0 92.6 120.0 152.4 47
41.4 65.6 93.0 120.2 156.2 48
1h 3h 6h 12h 24h I
42.0 69.0 93.0 125.2 158.2 49
42.0 69.0 93.0 125.4 159.0 50
42.4 71.4 93.4 126.0 163.0 51
42.8 71.7 93.6 128.2 164.8 52
42.8 72.2 97.8 130.4 167.0 53
42.8 72.4 98.0 134.0 167.2 54
43.2 74.8 98.4 134.8 177.0 55
44.0 76.0 100.0 135.2 182.0 56
44.0 78.0 101.6 136.0 186.4 57
48.8 79.0 101.8 136.0 187.8 58
49.0 80.6 102.0 136.3 188.0 59
50.0 89.4 106.0 137.6 188.0 60
50.0 92.6 116.0 140.2 191.0 61
50.0 93.2 116.4 141.4 193.8 62
50.2 96.0 121.6 146.2 195.6 63
50.4 99.0 122.2 151.4 211.4 64
53.0 99.0 124.0 161.6 220.8 65
54.0 103.0 124.0 162.4 252.0 66
55.8 105.8 135.2 182.0 254.8 67
56.0 110.0 148.0 186.6 281.4 68
63.4 115.2 188.6 255.2 291.4 69
Table – Data sorted from the lowest to the highest values
California = F1
Hazen = F2
Weibull = F3
Chegodayev = F4
Blom = F5
Tukey = F6
Gringorten = F7
I F1 F2 F3 F4 F5 F6 F7
1 0.01 0.01 0.01 0.01 0.01 0.01 0.01
2 0.03 0.02 0.03 0.02 0.02 0.02 0.02
3 0.04 0.04 0.04 0.04 0.04 0.04 0.04
4 0.06 0.05 0.06 0.05 0.05 0.05 0.05
5 0.07 0.07 0.07 0.07 0.07 0.07 0.07
6 0.09 0.08 0.09 0.08 0.08 0.08 0.08
7 0.10 0.09 0.10 0.10 0.10 0.10 0.09
8 0.12 0.11 0.11 0.11 0.11 0.11 0.11
9 0.13 0.12 0.13 0.13 0.12 0.13 0.12
10 0.14 0.14 0.14 0.14 0.14 0.14 0.14
11 0.16 0.15 0.16 0.15 0.15 0.15 0.15
12 0.17 0.17 0.17 0.17 0.17 0.17 0.17
13 0.19 0.18 0.19 0.18 0.18 0.18 0.18
14 0.20 0.20 0.20 0.20 0.20 0.20 0.20
15 0.22 0.21 0.21 0.21 0.21 0.21 0.21
16 0.23 0.22 0.23 0.23 0.23 0.23 0.23
17 0.25 0.24 0.24 0.24 0.24 0.24 0.24
18 0.26 0.25 0.26 0.26 0.25 0.25 0.25
19 0.28 0.27 0.27 0.27 0.27 0.27 0.27
20 0.29 0.28 0.29 0.28 0.28 0.28 0.28
21 0.30 0.30 0.30 0.30 0.30 0.30 0.30
22 0.32 0.31 0.31 0.31 0.31 0.31 0.31
23 0.33 0.33 0.33 0.33 0.33 0.33 0.33
24 0.35 0.34 0.34 0.34 0.34 0.34 0.34
25 0.36 0.36 0.36 0.36 0.36 0.36 0.36
26 0.38 0.37 0.37 0.37 0.37 0.37 0.37
27 0.39 0.38 0.39 0.38 0.38 0.38 0.38
28 0.41 0.40 0.40 0.40 0.40 0.40 0.40
29 0.42 0.41 0.41 0.41 0.41 0.41 0.41
30 0.43 0.43 0.43 0.43 0.43 0.43 0.43
31 0.45 0.44 0.44 0.44 0.44 0.44 0.44
32 0.46 0.46 0.46 0.46 0.46 0.46 0.46
33 0.48 0.47 0.47 0.47 0.47 0.47 0.47
34 0.49 0.49 0.49 0.49 0.49 0.49 0.49
35 0.51 0.50 0.50 0.50 0.50 0.50 0.50
36 0.52 0.51 0.51 0.51 0.51 0.51 0.51
37 0.54 0.53 0.53 0.53 0.53 0.53 0.53
38 0.55 0.54 0.54 0.54 0.54 0.54 0.54
39 0.57 0.56 0.56 0.56 0.56 0.56 0.56
40 0.58 0.57 0.57 0.57 0.57 0.57 0.57
41 0.59 0.59 0.59 0.59 0.59 0.59 0.59
42 0.61 0.60 0.60 0.60 0.60 0.60 0.60
43 0.62 0.62 0.61 0.62 0.62 0.62 0.62
44 0.64 0.63 0.63 0.63 0.63 0.63 0.63
45 0.65 0.64 0.64 0.64 0.64 0.64 0.64
46 0.67 0.66 0.66 0.66 0.66 0.66 0.66
I F1 F2 F3 F4 F5 F6 F7
47 0.68 0.67 0.67 0.67 0.67 0.67 0.67
48 0.70 0.69 0.69 0.69 0.69 0.69 0.69
49 0.71 0.70 0.70 0.70 0.70 0.70 0.70
50 0.72 0.72 0.71 0.72 0.72 0.72 0.72
51 0.74 0.73 0.73 0.73 0.73 0.73 0.73
52 0.75 0.75 0.74 0.74 0.75 0.75 0.75
53 0.77 0.76 0.76 0.76 0.76 0.76 0.76
54 0.78 0.78 0.77 0.77 0.77 0.77 0.77
55 0.80 0.79 0.79 0.79 0.79 0.79 0.79
56 0.81 0.80 0.80 0.80 0.80 0.80 0.80
57 0.83 0.82 0.81 0.82 0.82 0.82 0.82
58 0.84 0.83 0.83 0.83 0.83 0.83 0.83
59 0.86 0.85 0.84 0.85 0.85 0.85 0.85
60 0.87 0.86 0.86 0.86 0.86 0.86 0.86
61 0.88 0.88 0.87 0.87 0.88 0.88 0.88
62 0.90 0.89 0.89 0.89 0.89 0.89 0.89
63 0.91 0.91 0.90 0.90 0.90 0.90 0.91
64 0.93 0.92 0.91 0.92 0.92 0.92 0.92
65 0.94 0.93 0.93 0.93 0.93 0.93 0.93
66 0.96 0.95 0.94 0.95 0.95 0.95 0.95
67 0.97 0.96 0.96 0.96 0.96 0.96 0.96
68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
Table - Calculations of formulae to access the frequency
For 1 hour
Class Interval n N F1 F2 F3 F4 F5 F6 F7
1 15.6 20.0 3 3 0.04 0.04 0.04 0.04 0.04 0.04 0.04
2 20.0 24.4 10 13 0.19 0.18 0.19 0.18 0.18 0.18 0.18
3 24.4 28.8 7 20 0.29 0.28 0.29 0.28 0.28 0.28 0.28
4 28.8 33.2 11 31 0.45 0.44 0.44 0.44 0.44 0.44 0.44
5 33.2 37.6 10 41 0.59 0.59 0.59 0.59 0.59 0.59 0.59
6 37.6 42.0 7 48 0.70 0.69 0.69 0.69 0.69 0.69 0.69
7 42.0 46.4 9 57 0.83 0.82 0.81 0.82 0.82 0.82 0.82
8 46.4 50.8 7 64 0.93 0.92 0.91 0.92 0.92 0.92 0.92
9 50.8 55.1 2 66 0.96 0.95 0.94 0.95 0.95 0.95 0.95
10 55.1 59.5 2 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
11 59.5 63.9 1 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
Table – Frequency analysis for 1 hour
For 3 hour
Class Interval n N F1 F2 F3 F4 F5 F6 F7
1 44.8 53.4 31 31 0.45 0.44 0.44 0.44 0.44 0.44 0.44
2 53.4 61.9 12 43 0.62 0.62 0.61 0.62 0.62 0.62 0.62
3 61.9 70.5 7 50 0.72 0.72 0.71 0.72 0.72 0.72 0.72
4 70.5 79.1 8 58 0.84 0.83 0.83 0.83 0.83 0.83 0.83
5 79.1 87.6 1 59 0.86 0.85 0.84 0.85 0.85 0.85 0.85
6 87.6 96.2 4 63 0.91 0.91 0.90 0.90 0.90 0.90 0.91
7 96.2 104.8 3 66 0.96 0.95 0.94 0.95 0.95 0.95 0.95
8 104.8 113.3 2 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
9 113.3 121.9 1 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
10 121.9 130.5 0 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
11 130.5 139.0 0 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
Table – Frequency analysis for 3 hours
For 6 hour
Class Interval n N F1 F2 F3 F4 F5 F6 F7
1 38.0 51.8 2 2 0.03 0.02 0.03 0.02 0.02 0.02 0.02
2 51.8 65.7 14 16 0.23 0.22 0.23 0.23 0.23 0.23 0.23
3 65.7 79.5 26 42 0.61 0.60 0.60 0.60 0.60 0.60 0.60
4 79.5 93.4 8 50 0.72 0.72 0.71 0.72 0.72 0.72 0.72
5 93.4 107.2 10 60 0.87 0.86 0.86 0.86 0.86 0.86 0.86
6 107.2 121.1 2 62 0.90 0.89 0.89 0.89 0.89 0.89 0.89
7 121.1 134.9 4 66 0.96 0.95 0.94 0.95 0.95 0.95 0.95
8 134.9 148.8 2 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
9 148.8 162.6 0 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
10 162.6 176.4 0 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
11 176.4 190.3 1 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
Table – Frequency analysis for 6 hours
For 12 hour
Class Interval n N F1 F2 F3 F4 F5 F6 F7
1 56.6 74.9 8 8 0.12 0.11 0.11 0.11 0.11 0.11 0.11
2 74.9 93.1 13 21 0.30 0.30 0.30 0.30 0.30 0.30 0.30
3 93.1 111.4 16 37 0.54 0.53 0.53 0.53 0.53 0.53 0.53
4 111.4 129.6 15 52 0.75 0.75 0.74 0.74 0.75 0.75 0.75
5 129.6 147.9 11 63 0.91 0.91 0.90 0.90 0.90 0.90 0.91
6 147.9 166.1 3 66 0.96 0.95 0.94 0.95 0.95 0.95 0.95
7 166.1 184.4 1 67 0.97 0.96 0.96 0.96 0.96 0.96 0.96
8 184.4 202.6 1 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
9 202.6 220.9 0 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
10 220.9 239.2 0 68 0.99 0.98 0.97 0.98 0.98 0.98 0.98
11 239.2 257.4 1 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
Table – Frequency analysis for 12 hours
For 24 hour
Class Interval n N F1 F2 F3 F4 F5 F6 F7
1 62.2 83.3 2 2 0.03 0.02 0.03 0.02 0.02 0.02 0.02
2 83.3 104.3 8 10 0.14 0.14 0.14 0.14 0.14 0.14 0.14
3 104.3 125.4 17 27 0.39 0.38 0.39 0.38 0.38 0.38 0.38
4 125.4 146.5 13 40 0.58 0.57 0.57 0.57 0.57 0.57 0.57
5 146.5 167.5 14 54 0.78 0.78 0.77 0.77 0.77 0.77 0.77
6 167.5 188.6 6 60 0.87 0.86 0.86 0.86 0.86 0.86 0.86
7 188.6 209.7 3 63 0.91 0.91 0.90 0.90 0.90 0.90 0.91
8 209.7 230.8 2 65 0.94 0.93 0.93 0.93 0.93 0.93 0.93
9 230.8 251.8 0 65 0.94 0.93 0.93 0.93 0.93 0.93 0.93
10 251.8 272.9 2 67 0.97 0.96 0.96 0.96 0.96 0.96 0.96
11 272.9 294.0 2 69 1.00 0.99 0.99 0.99 0.99 0.99 0.99
Table – Frequency analysis for 24 hours
Frequency Histogram
35
1h
30
3h
25
6h
20
n (k)
12h
15
24h
10
0
1 2 3 4 5 6 7 8 9 10 11
Class
Cumulative Frequency
80
70
60
50
1h
N (K)
40 3h
6h
30
12h
20
24h
10
0
1 2 3 4 5 6 7 8 9 10 11
Class
The two figures above represent the absolute & cumulative frequency of rainfall data from 1934 to 2011,
classified by an ascending order of rainfall amount into 11 classes. For different classes, a certain duration is
associated
The following graphs represent the frequency peak rainfall for given durations throughout the given years
1,20 Cumulative Frequency of 1h
1,00
0,80
F1 F2
F (h)
0,60
F3 F4
0,40 F5 F6
0,20 F7
0,00
0,0 10,0 20,0 30,0 40,0 50,0 60,0 70,0
h (mm)
Figure – Rainfall frequency by different probability formulae
0,80
F1
F (h)
0,60 F2
F3
F4
0,40 F5
F6
0,20 F7
0,00
0,0 20,0 40,0 60,0 80,0 100,0 120,0 140,0 160,0
h (mm)
Figure – Rainfall frequency by different probability formulae
0,80
F1
F (h)
0,60 F2
F3
F4
0,40 F5
F6
0,20 F7
0,00
0,0 50,0 100,0 150,0 200,0
h (mm)
Figure – Rainfall frequency by different probability formulae
1,20 Cumulative Frequency of 12 hr
1,00
0,80
F1
F (h)
F2
0,60
F3
F4
0,40
F5
F6
0,20 F7
0,00
0,0 50,0 100,0 150,0 200,0 250,0 300,0
h (mm)
Figure – Rainfall frequency by different probability formulae
0,80
F1
F (h)
0,60 F2
F3
F4
0,40 F5
F6
0,20 F7
0,00
0,0 50,0 100,0 150,0 200,0 250,0 300,0 350,0
h (mm)
Figure – Rainfall frequency by different probability formulae
2.3. Gumbel Probability Distribution function
This is a type of distribution which is suitable for extreme events.
Where,
Figure – Gumbel cumulative functions for different durations using reduced variable z
Figure – Gumbel density functions for different durations using reduced variable z
Figure – GEV cumulative functions using reduced variable z for different durations
Figure – GEV density functions using reduced variable z for different durations
This is not a quantitative test as it is just a procedure to visualize the ability (or the inability) of a chosen kind of
probability function to fit a given data sample.
It is for a quick pre-selection of which kind of probability functions look able or unable to fit the data sample.
Afterwards, a quantitative test will select the best fitting probability function among the preselected ones.
Here, we need to calculate the value of a(T) which is given by the equation:
where,
Then, we calculate h and then plot the value of h and ϴ to obtain the DDF curves.
µ1 [mm]
35.43 60.87 82.80 111.19 145.42
σx
10.88 20.44 25.40 32.54 45.13
k 5
Vm
0.311
b 3.596
aµ 36.45
n 0.4447
We can also calculate the IDF curves by plotting the intensity of rainfall and the durations as follows:
• The first step is to upload the DTM 10x10 (Digital Terrain Model) of the whole watershed in ArcGIS
software, so it is possible to choose the best visualization.
• Secondly it is important to check the
Coordinate System is WGS 1984 UTM
Zone 32N (which is EPSG 32632, where
EPSG stays for European Petroleum
Survey Group). One must check that
every datum added is in the same
Coordinate System, otherwise it is to
change in the right one. Table with information about the Coordinate System
• All the steps afterwards are obtained by some specific commands of Spatial Analyst Tools, that could
be activated from the Arc Toolbox.
• To proceed to the hydrological analysis a DTM correction is needed to remove numerical errors by
the process of filling; in this way a depresionless DEM model (Digital Elevation Model) is created.
• Then it is requested to find the flow direction that an hypothetic drop is going to go downstream; as
it is shown in the pictures each colour represents a different flow direction, following the model of
the circular palette.
• The following step has the aim to identify the flow accumulation, so starting from the raster of the
flow direction it is possible to calculate the number of cells that flow into a cell. Where flow
accumulation is very high it is certain that there is a channel, otherwise if the flow accumulation is
zero it is possible to identify a ridge.
Example of how to interpret numbers in the legend of the flow direction map
• Then the following phase is the watershed description, that give some pieces of information about
the geological assessment.
• So, it is requested the slope for each cell using trigonometric formulae, whence in the resulting raster
it is highlighted the steepest downhill descent from the cell.
15,0%
NW NE
10,0%
5,0%
W 0,0% E
SW SE
• After the previous steps of analysis it is possible to continue with the rainfall areal distribution, so
adding to ArcGIS the .csv table with information of pluviometers, one can visualize these on the map
choosing to correlate the field of “Elevation” in the table of contents with the Z field. In this way one
can export data of pluviometers in a punctual .shp format.
Teglia hillshade map with the location of pluviometers employed in the calculations
• Then the necessity is to
correlate the area of the basin
with the average amount of
precipitation over the whole
above-mentioned area: it is
requested the Thiessen
polygon approach, where to
every pluviometer is assigned
the relative influence area. To
carry out this analysis, ArcGIS
has a particular function called
“Create Thiessen Polygons”
which generates a shapefile
Teglia basin subdivided with the rule of Thiessen polygons with these peculiar polygons.
• Afterwards to continue the analysis some parameters of the watershed are needed to proceed with
the next chapters, like the main stream length and the main stream slope.
• To find the main
stream one needs
raster of flow
direction, so with the
function “Flow
length” one can
calculate the
upstream and
downstream
distances along the Mainstream
flow path for every
cell; combining with
the “Raster
Calculator” the
downstream and
upstream maps the result is a raster with an augmented effect because on the cells belonging to the
main stream the value is the same. So, with the function just implemented in the previous lines, it is
possible to change the map in this way: Value 1 on the main stream and No Data in all the other cells.
• To compute the average slope of the main channel, one way
could be the use of the raster of the main stream as a mask
to apply on the raster of the slope (of the whole Teglia
watershed); so after this procedure one can the mean value
for the slope of the main stream consulting statistics section
in the Properties. It is possible to read on the “Mean” row
the value, which is an angle expressed in degrees (°) and
then calculate the tangent of its, figuring the slope.
Mean value of the mainstream slope
• Then the aim is to calculate the length of the main stream, so first it is needed to convert the raster
of the main stream in a polyline vector and then with the “Measure” tool it is possible to find out the
total length of the channel.
√𝐴
Ventura: 𝑇0 = 0.127
√𝑖
𝑟 𝑇 1 𝐿𝑖,𝑗
Most recent calibrations for urban drainage sewers: 𝑇0 = 𝑇𝑒 + 1.5 = 𝑇𝑒 + 1.5 𝑚𝑎𝑥𝑗 {∑𝑗 𝑉𝑟 }
𝑖,𝑗
Li,j , Vri,j : length [m] and full pipe velocity [m/s] of the i-th trunk of the j-th path
First, it’s necessary to study information from the Corine Land Cover Map: one can observe that most of the
territory of Teglia watershed is classified as forest, then some portions of agricultural areas and scrub
vegetation associations. Specific pieces of information about the land use are available in the chapter
4. SCS – Curve number computations. Then the elevation of the basin goes from 706,4 m to 170,4 m, so the
territory is on the hillside and especially in the area farthest from the closing point slope is very pronounced.
First, the formula for urban drainage sewers must be excluded, since this region is mostly rural. All the other
formulae described above are suitable for rural areas, like the Teglia watershed. Giandotti formula has been
studied for basins with areas between 170 and 70.000 Km2, so considering Teglia watershed has an area of
watershed has an area of 38,8 km2, this formula cannot be applied. Then Kirpich formula has been developed
for little basins in Pennsylvania, so it’s suggested to look for a formula that fits also in the territory of the
project. Then from calculations is quite clear that results obtained by Ventura or Pezzoli formulae are not so
different, but the group preferred Pezzoli formula due to its excellent applicability for mountain and hilly
areas. From computations one can observe the right choice to exclude Giandotti and Kirpich formulae
because results are overestimated or underestimated. In the following tables the group provided some useful
information of the basin and calculations with Pezzoli, Ventura, Giandotti and Kirpich formulae.
L[m] 15449,6
α 9,22 °
i[-] 0,162
A[m²] 38840621,0
Hm[m.a.s.l] 706,4
h [m.a.s.l] 170,4
PEZZOLI
T0 [h] 2,1
L[km] 15,4
i[-] 0,16
purposes of this study it’s necessary to classify the hydrologic soil type into 4 classes (SHG=Soil Hydrologic
Group): every class will have some specific characteristics for type of soil and consequently for
needs to join this shapefile with the .csv legend. In this way very detailed classification is provided, but for
this analysis the second level of classification is adequate, so the suggestion is to categorize the view with
the value “LABEL2”. To better understand how to read the Corine Land Cover map, one can start from the
inner part and proceed in this way:
2) second level > arable land, permanent crops, pastures and heterogeneous agricultural areas;
3) third level > non-irrigated arable land, permanently irrigated land, rice fields, vineyards, fruit trees and
berry plantations, olive groves, pastures, annual crops associated with permanent crops, complex cultivation
patterns, land principally occupied by agriculture, with significant areas of natural vegetation and agro-
forestry areas.
It is possible to observe the greater level of technicality when the level gradually increases. So it is possible
to continue and consider the other .csv file to attribute the 4 CN values to every second level class, one for
each Soil Hydrologic Group. After this step is possible to unify map with lithological classes with the one of
the land cover, so with from the ArcToolbox one can select the option “Overlay” and then “Union”.
Afterwards one can work with the Field Calculator, so it’s necessary to add a new field called “CN” where
combining information of the two columns “Soil_Gro_2” and “CN_Attri_i”, the result will be the calculated
CN. For example if the soil group of the j-th row is A the
code will pick value of CN from the column “CN_Attri_2”, and so on. The result is a fragmented map for the
CN, but the aim is to have one polygon for all the areas with the same CN, so it possible proceed with the
tool “Dissolve”. Now the result is a map easy to read, so choosing the categorized view one can observe
different colours for every CN. The last step is to compute the weighted average CN of the total Teglia
watershed, so first it’s necessary to calculate the area of every part of the region adding a new filed in the
Attribute Table of the last shapefile created (“CN_dissolve”); then it is possible to export data and compute
the average CN for the catchment with this formula:
∑𝑖 𝐴𝑟𝑒𝑎𝑖 ∗ 𝐶𝑁
𝐶𝑁 =
∑𝑖 𝐴𝑟𝑒𝑎𝑖
6. Rainfall event analysis
6.1. Return period estimation
What is Return period in hydrology? A return period of x time units, also known as a recurrence interval
(sometimes repeat interval) is an estimate of the likelihood of an event, such as an, flood, rainfall intensity,
a river discharge flow or any observable, to occur (or be overcome) on average every x time units.
After we have obtained CN number of soils in area of study, we can start analysis. The rainfall data for
different station is possible to earn by calculating different stations of measure which work with rain gauges.
Once we clipped the area, we can use tessellation method to weight the area.
How tessellation method works? Thiessen Polygons: They define areas around each data point where each
point internal to the polygon is closer to that data point than to all other data points. Each polygon side
separates two data points
Weight of area is earned by Calculation Geometry from GIS. Once we have sum of all areas, we can divide
each area to sum. So, in this way we can have the percentage of each area which is weight.
From rain gauges we have storm data for every 5 minutes. For calculating total rain height, we need to
multiply every basins 5 minutes storm data with area percentage of that area.
Is needed just to sum every 5 minutes total h until the time period threshold that we want.
P=Cumulated rainfall depths for each duration
Maximum Cumulated
Duration [hours]
rainfall depths
1 38.2
3 100.1
6 186
12 243.5
24 291.3
Estimate maximum values for each duration
According to the subchapter 2.6 ‘’Depth-Duration-frequency Curves’’ the DDF curves and the IDF curves are
like the following figures.
DDF CURVES
400,00
350,00
300,00
h (T=5)
250,00 h (T=10)
h[mm]
h (T=25)
200,00
h (T=50)
150,00 h (T=100)
h (T=250)
100,00
h (T=500)
50,00 h (T=1000)
0,00
0 5 10 15 20 25 30
θ [hours]
140,00
120,00
h (T=5)
100,00 h (T=10)
i[mm/h]
h (T=25)
80,00
h (T=50)
60,00 h (T=100)
h (T=250)
40,00
h (T=500)
20,00 h (T=1000)
0,00
0 5 10 15 20 25 30
θ [hours]
In order to calculate the return period of different rainfall event, each maximum rainfall depth was put into
comparison with the corresponding DDF form. Considering a worse situation, the bigger return period is
chosen rather than a closer one.
For example, the maximum rainfall depth within one hour in the given basin is 38.2mm. Putting it into DDF
form, it is easily known that it located between 5 years and 10 years. Considering a worse situation, ‘10 years’
is chosen as the final return period. This means that in this basin, the return period of rainfall depth reaching
38.2 mm within 1 hour is 10 years.
With the same processes, all the return periods are calculated for the corresponding duration.
h
38,2 100,1 186,0 243,5 291,3
max(mm)
Therefore, since the worse is chosen for case scenario, the return period is T=250 years.
6.2. Net rainfall estimation (SCS-CN model)
Before estimating net rainfall, a typical rainfall event was selected from the data series obtained previously.
Considering hydrological losses during the rainfall event, the net rainfall must be analyzed before rainfall-
runoff analysis.
𝑄 = 𝑃 − (𝐼𝑎 + 𝐹)
Where
• Q: Direct runoff
• P: Rainfall
• Ia: sum of all losses before the runoff begins
• F: retention after runoff begins
Combining water balance equation and proportionality assumption, the final SCS-CN equation is
obtained:
[𝑃(𝑡) − 𝐼𝑎]2
𝑄(𝑡) =
𝑃(𝑡) − 𝐼𝑎 + 𝑆
In which 𝐼𝑎 is usually assumed as:
𝐼𝑎 = 0.2 ∗ 𝑆 = 22,4
Using this function, the data series based on our selected rainfall event is obtained. Values of P, Q and F
were evaluated in the following form:
Date h[mm] P[mm] Pn[mm] F[mm]
As It is observed in the graph, due to the relatively small value of CN, which is refering to a more pervious
soil, the cumulated runoff volume tends to be much small. Most rainfall has become losses through
infiltration effect.
By considering the intensity of rainfall(I) and runoff (In), the following form is obtained further and graph by
applying the formula:
∆𝑃 ∆𝑄
𝑖= 𝑖𝑛 =
∆𝑡 ∆𝑡
In this case ∆t = 0.08 hour
25/10/11 0.00 0 0 0
60
50
40
intesity [mm/h]
30 I [mm/h]
In [mm/h]
20 f [mm/h]
10
0
0:00 4:48 9:36 14:24 19:12 0:00 4:48
Time [Hours]
Rainfall/Runoff intensity
7. Hydrological models
There are two main types of these models, which are distributed or lumped models. Three models are used
to simulate, Linear Kinematic, Linear Reservoir, and Nash Model. All of them are Lumped models.
IUH is the “Instantaneous Unit Hydrograph” which by multiplying it with the net rainfall obtained from the
runoff that will reach the downstream of the basin, which shows the response of our basin to this rainfall. And
it has two base times finite and infinite, and its unit is 1/time, which it is going to be used as [1/h].
First using the net rainfall intensity from the SCS-CN Model, it is calculated the Pn (the discharge); the net
intensity multiplied by the area at the exact time, which allow to build a Matrix according to the time, such
that at each time had all the discharge values until this t(i) in one row, which will allow to add them all. And
since the model is using all the catchment at once, A(total) is going to be used at each time in the matrix.
Pn = in * A total
Equation of Discharge in linear kinematic model
t [h] in [mm/h]
0,0 0,0000
1,0 0,0000
2,0 0,0000
3,0 0,0000
4,0 0,0000
5,0 0,0000
6,0 2,3831
7,0 1,4466
8,0 2,5637
9,0 5,8676
10,0 8,4093
11,0 11,7034
t [h] in [mm/h]
12,0 36,0028
13,0 15,0517
14,0 25,0515
15,0 20,1742
16,0 15,0367
17,0 17,6968
18,0 21,4299
19,0 0,7922
20,0 3,2318
21,0 0,0148
22,0 8,1399
23,0 2,7779
Table of Net discharge values for kinematic model
And using
Pn(r, c) = Pn(r − c + 1)
if r≥c, where r and c refer to row and column respectively. Other than that Pn(r,c)=0 if r<c. Therefore, the
Matrix is obtained, below is the part of it, and then all of it.
1 2 3 4 5 6 7 8
In this model the time-area curve is a straight line, which gives a constant IUH, which is IUH=1/T0. And below
are the graph and part of the IUH data.
t [h] IUH
0 0,47
1 0,47
2 0,47
3 0,0
4 0,0
5 0,0
6 0,0
7 0,0
8 0,0
9 0,0
10 0,0
11 0,0
12 0,0
13 0,0
14 0,0
15 0,0
16 0,0
17 0,0
18 0,0
19 0,0
20 0,0
21 0,0
22 0,0
23 0,0
Table IUH for kinematic model
IUH for kinematic model - graph
Now to get the runoff discharge at the downstream of the catchment we have the multiply the IUH by the
matrix.
This model is assuming the basin as one reservoir that has an input of net rainfall and an output of:
𝑁𝑒𝑡 𝑟𝑎𝑖𝑛𝑓𝑎𝑙𝑙
𝑄=
𝑘
Equation of Runoff Discharge in linear reservoir
𝑄 = (0.5~1.3) ∗ 𝑇𝑜
Equation of Linear reservoir routing constant
basin is a mountain area, therefore the coefficient 0,8 for the k, and T0 is 2.1h, which gives k=1,7
As for the Matrix, it is going to be used the same one as Linear Kinematic, since it has the same concept of
one Area and a cumulative net rainfall intensity.
In this model, the base time is infinite, so the IUH is not constant like before, it is exponential like this model,
with a formula of:
1 𝑡
𝐼𝑈𝐻 = ∗ 𝑒− ⁄𝑘
𝑘
IUH equation for linear reservoir
0,60
0,50
IUH [1/h]
0,40
0,30 IUH
0,20
0,10
0,00
0,0 5,0 10,0 15,0 20,0 25,0
t [hours]
Now by using the IUH instead of using the main formula of Linear Reservoir Model (Equation 80), Q0 won’t
be consider and IUH will be used. Then, the process remains the same as Linear Kinematic model, multiplying
IUH by the Matrix to calculate Q, and getting the results shown below.
𝑡 𝑡
𝑄 = 𝑃𝑛∗ (1 − 𝑒− ⁄𝑘) + 𝑄0∗ 𝑒− ⁄𝑘
In this Model then IUH is also Infinite base time, with a formula of:
Using n=3, because n=2~3, which means assuming our basin as a series of three linear reservoirs, we get the
following IUH data.
Now by doing the Same as before in the two Models, the IUH is multiplied by the Matrix to get Q. Part of the
results are shown below, and the graph.
Duration (h)
IUH Q[m3/s]
0,0 0,0000 0,0
1,0 0,1942 0,0
2,0 0,2147 0,0
3,0 0,1781 0,0
4,0 0,1313 0,0
5,0 0,0907 0,0
6,0 0,0602 0,4
7,0 0,0388 5,6
8,0 0,0245 10,5
9,0 0,0153 14,8
10,0 0,0094 24,1
11,0 0,0057 45,7
12,0 0,0034 70,5
13,0 0,0021 109,9
14,0 0,0012 148,8
15,0 0,0007 190,9
16,0 0,0004 212,6
17,0 0,0003 216,1
18,0 0,0001 215,3
19,0 0,0001 193,5
20,0 0,0000 151,3
21,0 0,0000 113,1
22,0 0,0000 82,1
23,0 0,0000 64,4
Runoff discharge for nash model
The maximum runoff discharge is 217.3 m3/s.
• φ: is the ratio between the cumulative net rainfall and the cumulative rainfall, which is from going
back to the SCS-CN is the sum of ∑Pn[mm]=22100.92mm and ∑P[mm]=40192.65mm.
• Then φ=∑Pn/∑P, which gives φ=0.550.
• A(total) of the catchment in [m2] which is as mentioned before. Therefore, the A(total) is equal to
38’840.621m2.
• The parameters a(T) and n, for a certain return period.
To know the return period, the return period is chosen for a Δt that includes our T0(time of concentration),
which in this case is as mentioned before T0=2.1h. Then, according to the table mentioned before for
choosing the return period, the return period is taken from the first Δt higher than T0 which is Δt=3.
Δt[hours] 1 3 6 12 24
Maximum water level (mm) for the Basin.
Therefore, the return period used to choose the parameters is T=25years. Then going back to choosing a(T)
at T=25years, it is obtained:
Therefore, a(T) is 51.21mm/h, or in the proper units of a Critical event for Linear kinematic model is 1.70*1e-
5m/s. The n value is 0.4447 which is constant for all return periods and having T0=2.1h.
Replacing the previous values, the Qmax for Kinematic is 233.6 m3/s, which is r close to the Qmax obtained
in the subchapter “Linear Kinematic” of the present section (283.73 m3/s).
Thus, in this case, critical event method works for Linear Kinematic Model.
8.2. Linear reservoir model
For this case the equation will be:
• The common parameters with the equation for the linear kinematic model are the same.
• D(n)=0,6 which is recommended to use.
• Where k=1,7 as calculated before. This gives us: Qmax (Reservoir)=167.2 m3/s.
Θc is 1.73h.
But we notice that Qmax from the critical event method is different from the one before in the Linear
Reservoir Model Qmax (Reservoir)=270.3 m3/s. This is because Linear reservoir model works on linear
catchments, and since the basin is a mountain area, Linear Reservoir model will not give a good estimation,
also because in a mountain the routing coefficient cannot be constant like k.
9. Conclusions
From the analysis of the selected region, it has been observed that in this study case there are not big urban
centres. So, on one hand the population is not huge, but on the other hand villages are very limited and quite
far one from another. Moreover, roads are just few and some of these are very narrow, which is obviously a
problem in case of emergency.
When the event of October 25th, 2011 happened, it was immediately noticeable the exceptionality of this
event, but the catastrophic consequences were stressed by many other factors: modification on the land use,
so many themes correlated to excessive urban concreting and environmental protection, increasing intensity
of the rainfall (more extreme events). On the Teglia basin (due to its majority of rural environment) problems
could be correlated with low maintenance of rivers and channels, carelessness of the woods and in the end
the presence of new types of invasive non-autochthonous plants. Then from Google Earth one can observe
in normal situation on the Teglia watershed the river is almost dried up due the dam regulation, but during
a huge rainfall event the situation can change rapidly, and the river suddenly increases the flow rate.
Moreover from the map one can observe that some buildings in the Castagnetoli village are very close to the
riverbed (actually some of these are no more than 25-30 metres away from the riverbed); it's important to
analyze the whole context: even if there is pretty recent plan for the hydrogeologic assessment - Piano di
assetto idrogeologico – some buildings are 50 or 60 years old.
Finally, it’s very important to reason on the Rocchetta dam, which is located approximately in the centre of
the catchment and the medium elevation is 420 m a.s.l. This dam is used for the underneath plant which
produces energy by water obliged in pressure pipes. In particular for the damages inducted by the dam an
inquiry was opened, and results were that some buildings were where they must not be.
Flood warning signals
• Based on Digital Terrain Model of basin which mapped in software ArcMap, it obtained the Lithology
Unit in catchment area contained Clay deposits of fluvio-lacustrine or marine origin, with
intercalation of sands, gravels and other materials which are materials with low and moderate
permeability.
• It can be observed in CLC in catchment area which obtained from mapping in ArcMap mostly the 85%
of the catchment correspond to natural vegetation (Forest), 10% correspond to Heterogeneous
agricultural areas and 5% are areas with arable land and Scrub or Herbaceous Vegetation
associations.
• It is clear the huge importance of the DDF curves and the IDF curves. moreover, it is easy to see his
relation because in the DDF the highest values of depth match with the longest durations of the rain,
nevertheless in the IDF curves the highest values of intensity match with the shortest durations of
the rain.
• The choice of time of concentration to be used needs high carefulness and good understanding of
the impact of the parameter in other calculation. It is better to be more conservative in design than
underestimating the real value. Between the three formulas both Pezzoli and Ventura results
between 1.96-2.1 hours of time of concentration, meanwhile Giandotti offers 2.60 hours. Lower time
of concentration means more conservative design as we compress the rainfall depth to a fewer
amount of time which also means higher discharge flow.
• By using certain calculation to get rainfall-runoff models, it will be derived a simulated of the routing
of runoff to downstream of basin section, which modelled by Linear Kinematic, Linear Reservoir, and
Nash Model. Then to get the runoff discharge at the downstream of the catchment, it has to multiply
the IUH. So, it can be obtained the maximum results of each model (Qmax) in some duration of
rainfall. For Linear Kinematic = 37.68 m3/s; Linear Reservoir = 56.63 m3/s, and last is Nash Model =
16.18 m3/s, they all are peaked between the fifteen and seventeen hours.
• Method of Critical Event used to estimate the maximum value of downstream runoff, and it already
calculated for two models Linear Kinematic Model and Linear Reservoir Model. For each model are
included Time of Concentration (T0) and Parameter (a(T)), thus it derived Qmax for Linear Kinematic
Model 233.6 m3/s and Qmax for Linear Reservoir Model 167.2 m3/s.
• The comparison of distinction for all Qmax explained that Linear Kinematic is more effective for this
case because of morphology of the catchment area. Linear Reservoir Model works on flat
catchments, and since our basin is a mountain area, Linear Reservoir Model is not effective to give
good estimations.