You are on page 1of 16

CIVE 646 - Probabilistic Methods in Hydroscience

Fall 2015
Homework 1
Mohammad Ashifur Rahman
CLID: mxr2467

Problem 1.2
Mean annual maximum flow = 5407.9 m3/s
Standard deviation = 1741.9 m3/s
Histogram:

Cumulative relative frequency diagram:

1st quartile = 4002.5 m3/s


Median = 5400 m3/s
3rd quartile = 6815 m3/s
Boxplot:

Probability that flow will exceeded 5000 m3/s during a 12-month period =
5000 2
)
61

1(

= 1 0.59022
= 0.6517

Problem 1.3

Mean annual maximum flow = 2837.5 m3/s


Standard deviation = 1314.1 m3/s
Histogram:
25

20

frequency

15

10

0
1000

2000

3000

4000

5000
bin

6000

7000

4000

5000

6000

7000

8000

9000

Cumulative relative frequency diagram:

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1000

2000

3000

8000

9000

1st quartile = 1980 m3/s


Median = 2410 m3/s
3rd quartile = 3245 m3/s
Boxplot:

8000
7000
6000
5000
4000
3000
2000
1000
1

The distribution in problem 1.2 is almost symmetric, but distribution in problem 1.3 is skewed to the
left.

Problem 1.9
Histogram:

140

120

frequency

100

80

60

40

20

0
40

60

80

100
120
mean time

140

160

180

The distribution is skewed to the right.


There are on primary peak at around 70 minute and another secondary peak at around 90 mins.
Probably due to the perception of the speed limit or required speed, the two peaks occurred. The
reason of second peak could also be from rough weather (for example, fog), where many cars traveled
in low speed.
The difference between two peaks is about 20 mins. It took 20 minutes more if the drivers had to lower
the speed due to any disturbance like foggy weather.

Problem 1.11
Stem and leaf diagram:
Stem Leaf
80 7
81
82
83
84 6
85
86
87 6
88 5 6
89 6
90 9
91 3
92 2
93
94 0 8
95 0 9
96 9
97 8
98 6 7
99 3 5 7 9
100
101 1 5 7
102 6 8 9
103
104 6 6
105 1
106
107
108
109 0 6
110 0 0
111 2
112 8
113 3 3

114 2
115 9
116
117 1
118
119 6 7
120
121 5 8
122 8
123
124
125 9
126 4
127
128
129 0
130
131 8
132 3
133
134 5 9
135 6
136 2
137
138
139
140
141
142 2
143
144
145
146
147
148
149 6
150 1
151

152 9
153
154
155
156 4
157
158
159
160
161
162
163
164
165 4

Boxplot:

1600
1500
1400
1300
1200
1100
1000
900
800
1

The distribution is skewed to the right.

Problem 1.12
Number of class width, nc = 1 + 3.3 log10 n
N = 123, class width (Sturges) nc = 7.9 8
1

Number of class width,

3
2

r = 60, n = 123, iqr = 5.76

60

30

50

25

40

20
Frequency

Frequency

Class width (Freedman and Diaconis) = 25.9 26

30

15

20

10

10

10

20

30
40
bins (Sturges)

50

60

70

10

20
30
40
50
bins (Freedman and Diaconis)

60

Histograms showed that the distribution is skewed to the right. Second histogram gives better
illustration of the data.
Boxplot:
60

50

The mean (12.4) and median (11.3)


are very close. If only 5 outliers are
removed, it becomes a good
distribution. Therefore, the
contractors claim is not justified.

40

30

20

10

0
1

70

Problem 1.15
Mean, mg/L
Standard Deviation, mg/L
Coefficient of variation

Chloride
65.2
3.3363
0.05

Phosphate
1.823
0.1992
0.11

From the coefficient of variation value, it can be seen that Phosphate has more variability in
concentration than Chloride.

Scatter plot:

2.3
2.2
2.1

Phosphate

2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
54

56

58

60

62

64
Chloride

66

68

70

72

74

Correlation coefficient, r = 0.0271, indicates very poor correlation. It is very difficult to predict from this
poor correlation but further analysis of variance could be checked whether other relationships between
these two have any influence on prediction.

10

Problem 1.18
Line diagram:
20
18
16
14

frequency

12
10
8
6
4
2
0

4
5
6
Number of hurricanes

The pattern of the two line diagram is not very different. Hurricane occurrences are more than floods.
There was an increase in 6 number of floods.

11

Problem 1.19

Mean
Standard Deviation
Coefficient of variation

NO2
122.25 g/m3
16.0779 g/ m3
0.1315

CO
4.5125 mg/ m3
1.4317 mg/ m3
0.3173

From the coefficient of variation value, it can be seen that CO has more variability in concentration than
NO2.

Scatter plot:

7.5

CO concentration (milligram per cubic meter)

7
6.5
6
5.5
5
4.5
4
3.5
3
2.5
90

100
110
120
130
140
NO2 concentration (microgram per cubic meter)

150

Correlation coefficient, r = -0.1522, does not indicate a very strong correlation.

12

Problem 1.20
Minutes
Mean
Standard Deviation
Coefficient of skewness, g1 =
3 ( )

5
13.48
5.94
1.23

10
20.44
9.15
0.93

20
31.95
16.11
0.78

30
38.66
20.44
1.15

40
45.51
26
1.38

50
52.47
31.74
1.36

60
57.9
36.83
1.45

120
74.81
46.28
1.19

180
83.66
46.18
1.20

Although mean and median depth of rainfall show continuous increase with time, coefficient of
skewness of depth shows initial decrease, increase and then decrease.

90
80
70

mean

60
50
40
30
20
10

20

40

60

80
100
120
duration (minutes)

140

160

180

13

50
45

Standard Deviation

40
35
30
25
20
15
10
5

20

40

60

80
100
120
duration (minutes)

140

160

180

20

40

60

80
100
120
duration (minutes)

140

160

180

1.5
1.4

Coefficient of Skewness

1.3
1.2
1.1
1
0.9
0.8
0.7

14

Short storms are localized and depths are predictable but long storms are less predictable. Therefore
mean and standard deviations are increasing with durations but it is difficult to analyze the pattern of
coefficient of skewness with duration since it involves the median as well.

Problem 1.23

9
Observed
Calibration 1
Calibration 2

8.5
8
7.5

Period (s)

7
6.5
6
5.5
5
4.5
4
1.5

2.5

3.5
Height (m)

4.5

5.5

Coefficient correlation of
Observed values = 0.6962
Calibration 1 = 0.9946
Calibration 2 = 0.9847

Equation for calibration 1: y = 1.1358x + 3.4508


Equation for calibration 2: y = 1.0938x + 3.6048

15

The table has been prepared using those equations.


Observed Observed Calibration Calibration Deviation Deviation
Height
Period
1 period
1 period
1 (s)
2 (s)
(m)
(s)
(s)
(s)
2.26
6.1
6.017708
6.076788 0.082292 0.023212
3.1
3.22
3.84
2.56
2.74
2.28
3.88
2.49
4.22
2.01
2.77
3.61
3.51
2.52
2.12
2.73
3.3

4.3
5.7
7.7
5.3
5.7
4.9
6.7
5
6.9
5
5.9
6.5
7.4
5
5.1
6.5
5.4

6.97178
7.108076
7.812272
6.358448
6.562892
6.040424
7.857704
6.278942
8.243876
5.733758
6.596966
7.551038
7.437458
6.313016
5.858696
6.551534
7.19894

Mean of deviations
Standard Deviation of
deviations
Coefficient of variation of
deviations

6.99558
7.126836
7.804992
6.404928
6.601812
6.098664
7.848744
6.328362
8.220636
5.803338
6.634626
7.553418
7.444038
6.361176
5.923656
6.590874
7.21434

2.67178
1.408076
0.112272
1.058448
0.862892
1.140424
1.157704
1.278942
1.343876
0.733758
0.696966
1.051038
0.037458
1.313016
0.758696
0.051534
1.79894

2.69558
1.426836
0.104992
1.104928
0.901812
1.198664
1.148744
1.328362
1.320636
0.803338
0.734626
1.053418
0.044038
1.361176
0.823656
0.090874
1.81434

Calibration 1
0.9755
0.6699

Calibration 2
0.9988
0.6766

0.6868

0.6774

The calibration 1 is better than calibration two since calibration 1 has lower Mean of deviations and
Standard Deviation of deviations from the observed value than calibration 2, although Coefficient of
variation of deviations is bigger for calibration 1. Correlation coefficient of calibration 1 is closer to 1
than calibration 2. It will probably be wise to go for calibration 1.

16

You might also like