3.5.

Hydro Climatic Data Availability and Its Quality

3.5.1. Rainfall Data

The data includes the record of daily rainfall precipitation throughout the year ranging from 2004
to 2024 for 20 consecutive years data is only available for my study area (Ayertena). The
collected data from the Ethiopian meteorology agency has missing data for the consecutive years
starting from 2004 up to 2006 and from 2016 up to 2018.

B. Estimating missing rainfall data

Due to the absence of observer or instrumental failure rainfall data record occasionally are
incomplete. In such a case one can estimate the missing data by using the nearest station rainfall
data. There are different approaches for estimating missing rainfall data varying with and based
on the effect of orography on rainfall, distance between the rainfall stations and the variation of
rainfall amount recorded on the stations. Among different method Normal ratio method was one
of them which was recommended to estimate missing rainfall data in regions where annual
rainfall between stations differ by more than 10%. (Silesh, 2011).

Where: Pz-missing rainfall data (daily, monthly or yearly) P1, P2 and P3 – rainfall data at nearest
different station (daily, monthly or yearly) Nz - mean annual rainfall at missed station N1, N2,
and N3- mean annual rainfall at different nearest station.

C. Check the quality of data

Check on outliers has been undertaken on the recoded rainfall data to identify any low or high outliers.
Outliers are data points, which depart significantly from the trend of the remaining data. The outlier test is
done to check whether the adopted data is within the limited range or not. The daily heaviest rainfall data
of Ayertena meteorological station from 2004 to 2023 was taken for the design. Hence, 20 years of daily
heaviest rainfall data is available. These data was also checked for its consistency by higher and lower
outlier testes using Z –Score.

The Z-score, also known as standard score, is a statistical measure that indicates how many standard
deviations a data point deviates from the mean. It is calculated using the formula

Where Z:
 X is the observed value,

µ is the mean of the dataset, and

σ is the standard deviation of the dataset

In this study, the Z-score was used to standardize the rainfall data, enabling a comparative analysis of
extreme value. Data points were classified as outliers based on their Z- score values:

 Higher Outlier: If the value is significantly higher than the mean.

 Lower Outlier: If the value is significantly lower than the mean.

The detailed Z - score results are presented in the appendex

3.8 Design Rainfall Analysis

3.8.1 Estimation of average depth of Rainfall

4. RESULTS AND DISCUSSIONS

4.1. Rainfall Analysis Result

Rainfall data of the area has been collected from the Ayertena gauging station which is
controlled by the National Meteorology Service Agency (NMSA)…………………………...

Figures 4.1 below shows Ayertena time series rainfall data.

4.1.1. Rainfall Data Quality Result

Test for Outliers

This test helps to avoid those data lie out of the range in between the lowest and the highest. In
this study, the Z-score was used to standardize the rainfall data, enabling a comparative analysis of
extreme value. Data points were classified as outliers based on their Z- score values:

 Higher Outlier: If the value is significantly higher than the mean.

 Lower Outlier: If the value is significantly lower than the mean.
Annual Max Mean(X) Xi-Xm (Xi-Xm)2 STDV Z-SCORE
Rainfall(Xi)
37.4 46.57 -9.17 84.00 13.212 -0.694
37.5 46.57 -9.07 82.17 13.212 -0.686
60.1 46.57 13.54 183.20 13.212 1.024
32.4 46.57 -14.17 200.65 13.212 -1.072
34.7 46.57 -11.87 140.78 13.212 -0.898
34.2 46.57 -12.37 152.89 13.212 -0.936
59.4 46.57 12.84 164.74 13.212 0.971
38.1 46.57 -8.47 71.66 13.212 -0.641
43.4 46.57 -3.17 10.02 13.212 -0.240
42.6 46.57 -3.97 15.72 13.212 -0.300
39.6 46.57 -6.97 48.51 13.212 -0.527
51.2 46.57 4.63 21.48 13.212 0.351
69.4 46.57 22.84 521.44 13.212 1.728
49.6 46.57 3.03 9.21 13.212 0.230
34.1 46.57 -12.47 155.38 13.212 -0.943
41.2 46.57 -5.37 28.78 13.212 -0.406
47.7 46.57 1.13 1.29 13.212 0.086
52.6 46.57 6.03 36.42 13.212 0.457
83.6 46.57 37.04 1371.59 13.212 2.803
42.5 46.57 -4.07 16.52 13.212 -0.308
 Chacking Outlier
90
83.6
80
76
70
Annual Max Rainfall

60

50

40 Annal MaxRainfall
Higher Outlier
30
Upper Limit
20

10

0
2000 2004 2008 2012 2016 2020 2024
Year

Figure 4.2 Graphical comparison of maximum rainfall prior to test for outlier

As observed from the table and graph, there is one data point (83.6 mm) that exceeds the upper
limit of 76 mm or a Z-score of 2. However, there is no value below the lower limit. Therefore,
the value of 83.6 mm is excluded as it exceeds the upper limit.

4.1.2. Rainfall Frequency Analysis

The rainfall frequency analysis is done using both Gumble and log Pearson type III methods as
recommended by ERA manual 2013. The result obtained is tabulated in the following table
Table 4.2: Yearly Extreme series frequency analysis

Return Period (Tr) Extreme Rainfall depth (mm)

Gumble Log Person Type III

2
42.65 42.35

5
51.52 51.27

10
57.39 57.49

25
64.81 65.67

50
70.31 72.11

100
75.76 78.34

In order to identify which distribution fits to the theoretical probability distribution, goodness of
fit test conducted using Easy Fit 5.6 professional software and the log Pearson Type-III
distribution fits for the statistical value is lesser than that of the Gumble values the result is
tabulated below. That is, Log Pearson Type-III method have proved to be good fit in all the three
tests compared to the Gumble‘s Method. Accordingly, the Log Pearson-III is chosen for further
analysis. The statistics for both methods are calculated and the ranking is given below table 4.3.

Table 4.3: Goodness of Log-person-III and Gumbel methods

Distribution Kolmogorov-Smirnov Anderson-Darling Chi-squired

Statistics Rank Statistics Rank Statistics Rank

Log-person type III 0.152 4.98 4.12

Gumble 0.173 5.63 4.87

Table 4.4: Log pearson type-III daily heaviest rainfall analysis for Ayertena

Return Exceedance Skew Frequency Standard YT XT(mm)

Period Probability Coefficient(Cs) factor, K(T,n) deviation, Sy

2 0.5 0.692 -0.11 0.0921 1.627 42.355

5 0.2 0.692 0.792 0.0921 1.710 51.274

10 0.1 0.692 1.331 0.0921 1.760 57.491

25 0.04 0.692 1.959 0.0921 1.817 65.675

50 0.02 0.692 2.4 0.0921 1.858 72.111

100 0.01 0.692 2.79 0.0921 1.894 78.343

4.1.3. Intensity Duration Frequency Curves

The IDF curve is developed from a 24-hour rainfall data of 20 years duration obtained from
Ethiopian Meteorological Agency - gauge located Tikur Anbesa Hospital. Reduction equation as
depicted in the methodology section has been applied. Consequently, the following IDF curve
has been produced. The data obtained for production of IDF curve is the result of calculations
using reduction formula and it is tabulated below. Then, using the data in the table the IDF curve
has been produced as shown below.

Figure: Duration of rainfall against their corresponding average intensities

Duration Intensity for given return periods(mm/hr)

(minutes) T= 2 years T=5 years T=10 years T=25 years T=50 years T=100yrs

5 75.32 91.12 101.88 116.2 127.4 138.05

10 64.2 77.62 86.95 99.64 109.5 118.8

15 55.84 67.52 75.58 86.61 95.22 103.32

20 49.28 59.62 66.85 76.65 84.26 91.42

25 44.02 53.3 59.8 68.6 75.44 81.92

30 39.68 48.1 53.98 62 68.2 74.08

35 36 43.72 49.08 56.4 62 67.32

40 32.84 39.98 44.82 51.5 56.54 61.32

45 30.08 36.7 41.05 47.2 51.7 56.02

50 27.64 33.82 37.7 43.4 47.4 51.32

55 25.46 31.3 34.7 40 43.6 47.1

60 23.5 29.05 32 36.8 40.2 43.28

90 17.48 21.6 23.82 27.4 29.9 32.22

120 14.02 17.38 19.15 22 23.98 25.88

 Intensity-Duration-Frequency (IDF) Curve (Ayertena)
150

120
Rainfall Intensity (mm/hr)

90 T= 2 year
T=5 yrs
T=10 yrs
60
T=25 yrs
T= 50yrs
30 T=100yrs

0
0 20 40 60 80 100 120
Duration (min)

Figure 4.3: IDF Curve for Assosa from Assosa Observatory