Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al

*Present address National Institute of Hydrology, Roorkee, Corresponding author E mail: [email protected]




STATISTICAL AND PROBABILITY ANALYSIS OF RAINFALL FOR CROP PLANNING
IN A CANAL COMMAND
P. K. MISHRA*, DEEPAK KHARE, ARUN MONDAL, SANANDA KUNDU AND RITURAJ SHUKLA

Department of Water Resources Development &Management, Indian Institute of Technology Roorkee, Roorke, Uttarakhand -247 667, India
Received: 09.10.2013 Revised accepted: 23.11.2013
ABSTRACT

In the present study, statistical and probability analysis of 40 years daily rainfall data for the period 1971-2010
were carried out on weekly, monthly and annual basis. Rainfall pattern in the area is also analyzed using thiessen
polygon and Histogram technique. Probability analysis was done using two and three parameters probability
distribution functions. The best distribution representing the command was identified using Chi-squared (
2
) test
at different probability level. Amongst the nine distributions fitted in the Tawa canal command, Log-normal
distribution and Weibull distribution (2) are found as the best fit distribution for annual and monthly rainfall data
respectively. When the distributions depicting standard monsoon and non-monsoon weeks in the region are
studied, it has been found that Beta and Weibull represents best in comparison to other distributions. Finally, the
best fit distribution is employed for computing the minimum assured amount of rainfall at different probability
levels for the command area.
Keywords:
Goodness-of-fit,
Chi-squarer (
2
) test,
Probability
distribution,
rainfall,
statistical analysis


INTRODUCTION
India is an agrarian country with about 60% of the
cultivated area under rain-fed conditions and the remaining
40% under irrigated area. Rain-fed agriculture supports 40% of
the population and 60% of the livestock in the country (NRAA,
2011). Rainfall and its distribution in the country are erratic in
nature. There is high temporal and regional variation of rainfall
in the country. India receives rainfall mainly from two fronts
viz. south-east monsoon and south-west monsoon concentrating
in the months of June to September. Almost 80% of the total
rainfall is being received within 100 hours distributed in those
months. DES, 2011 report suggests a highest overall rainfall
variation of 18.6% in the year 2002-03. During the same period
monsoon rainfall declines to 19.2% in comparison to rainfall
received during the last two decades. Therefore, among other
factors, rainfall (monsoon) plays an important role on nations
growth and economy in particular. Also, rainfall is one of the
most important inputs to crop production. Hence,
understanding the underlying processes describing the rainfall
pattern, distribution and availability in a region is the first step
before planning further. The inference from the analyzed
rainfall data can be utilized in number of ways such as
identifying dry spell, developing drought indices, estimating
flood, etc.
It is the rainfall distribution which affects the yield in a
region rather than the volume of rainfall received (Fisher,
1924). Probability and frequency analysis of rainfall data
enables us to determine the expected rainfall at different
probability level. Rainfall at 80 per cent probability can be
taken as assured rainfall, while the median value (50 per cent)
as the maximum limit for taking any risks (Gupta et al., 1975).
At below 50 per cent, if considered, chances of crop failure will
be high. Many similar studies by different researchers
worldwide have undertaken for regional crop planning through
analyzing rainfall data. Nemichandrappa et al. (2010) analyzed
30 years rainfall data in the semi-arid zone of Raichur,
Karnataka. Probability analysis of daily rainfall data using 2-
parameters and 3-parameters probability distribution and
concluded that Gumbel distribution suits the best for the region.
Sharma and Singh (2010) analyzed daily rainfall data for 37
years collected from the IMD approved observatory at GB Pant
University of Agriculture and Technology, Pantnagar, India.
The data were analyzed fitting distributions and Goodness-of-
fit for those distributions were performed using three statistical
test viz. Kolmogorov-Smirnov test, Anderson-Darling test and
Chi-Squared Test. Mokashi et al. (2008) analyzed 30 years
daily rainfall data collected from Dry Farming Research
Station, Solapur, India and studied the probability distribution
of rainfall for the monsoon weeks for crop system planning
based on the rainfall distribution. Bhakar et al. (2008) made a
detailed statistical analysis of weekly and monthly rainfall data
using 35 years daily rainfall data. In the study different density
functions were tried to identify the best fit distribution
Agriculture for Sustainable Development
http://www.sustainableagriculture.in
Agriculture for Sustainable Development 1(1):95-102, 2013/Review ISSN 2347-5358 (Print)





Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al


generalizing the Kota region in India. Baskar et al., (2006)
observed the frequency analysis of consecutive days peaked
rainfall at Banswara, Rajasthan, India, and found gamma
distribution as the best fit as compared to other distributions
using by Chi-squared test. NIH (1998) extensively studied the
rainfall series to identify any trend and persistence in the Sagar
division, Madhya Pradesh. The study indicated a poor serial
correlation with no persistence. The rainfall series described a
normal distribution in the region. Upadhaya and Singh (1998)
predicted rainfall using various probability distributions for
certain returns period. Biswas and Khambete (1989) computed
rainfall at different probability level by fitting gamma
distribution considering weekly data from 80 stations in the
state of Maharashtra, India. Ramasastri and Nirupama (1986),
carried out statistical analysis of monthly and annual rainfall
data of Belgaum district to identify the presence of trend and to
study low rainfall pattern in the region. Kulandaivelu (1984)
undertaken daily rainfall data for 70 years for the Coimbatore
region in the state of Tamilnadu and analysed the daily
precipitation data by fitting Gamma distribution model.
Virmani et al. (1982) carried out analysis of the rainfall
probabilities for 77 semi-arid locations of the Indian sub-
continent. The analysis was done on weekly basis utilizing
Markov Chain probabilities. RamaRao et al., (1975) analyzed
the daily rainfall data for the period 1921 to 1970 at Bijapur.
Asian summer monsoon has been investigated for
variability using Climate System Model, version 1.4.
Simulation analysis indicated that a stronger monsoon is being
favoured by the large-scale warm temperature. Ashkar and
Mahdi (2003) investigated generalized probability weighted
moments (GPWM) and maximum likelihood (ML) fitting
methods in the two-parameter log-logistic (LL) model.
Probable Maximum Precipitation (PMP) often need in planning
and designing multi-purpose reservoirs. Papalexiou and
Koutsoyiannis (2006) estimated the PMP and analyzed the
results using probabilistic approach for the data collected from
4 stations. Durbude (2008) estimated PMP for planning soil
and water conservation structures in the Banswara district of
Rajasthan, India. Weerasinghe (1989) analyzed 35 years
consecutive rainfall data using Markov Chain probabilities for
weekly rainfall. Rainfall availability in the region was assessed
for crop planning.

MATERIALS AND METHODS
Study area
Tawa command is distributed in an area of about 5273.12 km
2

falling in the district of Hoshangabad, Madhya Pradesh. It lies
between north latitude from 2254 to 2300

N and longitude
from 76457

to 7845

E. The area is characterized by a hot
summer and evenly distributed rainfall during the southwest
monsoon period. The temperature starts rising from beginning
of February and peak is reached in the month of May 42C. The
winter season commences with November and December is the
coolest month with temperature 7.2 C. The relative humidity
during summer is low in the month of April i.e. about 18.1%
and is maximum in August i.e., 86.7%. The average annual
rainfall based on observations recorded during 1971 to 2010 is
1174.78 mm. The location map of the study area is shown in
Fig.1.


Data used
The daily rainfall data were collected from India
Meteorological Department (IMD), Pune for 40 years from
1971 to 2010. Daily rainfall data were also collected from the
fourteen local stations spread over in and around the Tawa
command area from the State Data Centre, Bhopal, Madhya
Pradesh. The daily data were converted to weekly and monthly
rainfall.
Probability distributions for rainfall data analysis
The probability distribution of both 2-parameters and 3-
parameters were considered to indentify the best fit probability
distribution for the region. The description of each distribution
type with its range and shape factors is presented in Table 1.

RESULTS AND DISCUSSION

Percentage variation in actual and normal rainfall received
in area
The study started with analyzing rainfall data during the
period 1992-93 to 2010-11 to recognize any variation in the
overall annual and monsoon rainfall in the region. As presented
in Table 2 that a large variation (both increasing and
decreasing) is seen both during monsoon rainfall (Jun-Sept)
and overall rainfall (Jun-May). The percentage variation is
highest in the year 2002-03 when the rainfall received in the
district declined considerably. Tawa command falls in the
central part of India. It is also revealed from Table 2 that the
command area also witnesses considerable percentage variation
particularly during monsoon. The year-wise percentage
variation during 1992-93 to 2010-11 is also shown graphically
in Fig.2.



Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al




Table 1: Description of Probability Density Functions considered in the study
Distribution type Probability Density Function Range Parameters
Normal


= Mean
= Standard deviation (>0)
Log-normal


= Mean
= Standard deviation (>0)
Gumbel Min


= Mean
= Standard deviation (>0)
Gumbel Max


= Mean
= Standard deviation (>0)
Weibull (2)


= Shape parameter (>0)
= Scale parameter (>0)
Gamma (2)


= Shape parameter
= Scale parameter
= Gamma function
Beta


1 = Shape parameter
2 = Shape parameter
B = Beta function
Log- Pearson (3)

(<0)
(>0)
= Shape parameter (>0)
= Scale parameter (>0)
= Location parameter
Gamma (3)


= Shape parameter
= Scale parameter
= Gamma function
Weibull (3)


= Shape parameter (>0)
= Scale parameter (>0)
= Location parameter

Table 2: Rainfall variation during 1992-93 to 2010-11 for Monsoon, Overall and Central part of India

Year
Monsoon rainfall (mm) (June-Sept) Overall rainfall (mm) (June-May) Monsoon rainfall in Central India (mm) (June-Sept)
Actual Normal % variation Actual Normal % variation Actual Normal % variation
1992-93 830.7 899.2 -7.6 1091.6 1175.6 -7.1
1993-94 902.1 908.9 -0.7 1184.3 1192.6 -0.7
1994-95 999.2 906.8 10.2 1297.3 1190.7 9.0
1995-96 904.5 904.7 0.0 1154.6 1189.3 -2.9
1996-97 927.6 905.7 2.4 1195.5 1190.3 0.4
1997-98 927.4 908.6 2.1 1291.5 1198.3 7.8
1998-99 945.2 903.6 4.6 1275.5 1198.8 6.4
1999-00 866.9 903.2 -4.0 1183.5 1197.0 -1.1
2000-01 833.7 902.3 -7.6 1043.7 1195.5 -12.7
2001-02 826.0 901.1 -8.3 1120.2 1196.0 -6.3
2002-03 737.1 911.7 -19.2 981.4 1205.4 -18.6
2003-04 947.3 902.7 4.9 1278.0 1196.5 6.8
2004-05 779.6 893.3 -12.7 1085.9 1197.3 -9.3 1094.9 993.2 10
2005-06 879.3 892.5 -1.0 1185.4 1196.8 -1.0 1152.2 993.9 16
2006-07 886.6 892.2 -0.6 1133.0 1195.5 -5.2 1073.8 993.9 8
2007-08 936.9 892.2 5.0 1180.2 1194.8 -1.2 956.9 993.9 -4
2008-09 873.2 892.2 -2.1 1075.0 1196.4 -10.1 794.8 993.9 -20
2009-10 689.8 892.2 -22.7 972.8 1195.6 -18.6 1027.9 991.5 4
2010-11 912.8 893.2 2.2 1212.3 1191.7 1.7
Source: DES, 2011
Thiessen polygon showing spatial pattern of rainfall received
in area
Thiessen polygon (Voronoi diagram) provides the area
of influence of each gauge station in and around the study area.
The perpendicular bisector of each line joining two stations
generates the area of influence (polygon). Once the area of
influence (thiessen weight) is assessed for individual stations,
average rainfall received in the area is estimated using the
following equation:

(1)
Where,
= Average rainfall; P
N
= Rainfall for individual station; A
N
=
Area of influence (Thiessen weight) for individual station; A =
Total area. The average rainfall received in the area using
Thiessen polygon is 1036.18 mm as presented in Table 3.




Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al


Table 3: Station-wise average rainfall and corresponding Thiessen
weight
Sl.
No.
Station Latitude Longitude
Aavg.
rainfall
Thiessen
weight
(km
2
)
Thiessen
average
1 Babai 22.704167 77.931667 923.97 554.934 512742.37
2 Bankhedi 23.775000 78.541667 1176.15 41.1951 48451.62
3 Harda 22.345833 77.093333 1051.66 972.866 1023124.26
4 Hoshangabad 22.750556 77.720833 1197.48 332.578 398255.50
5 Itarsi 22.611111 77.775000 1148.30 435.273 499823.99
6 Khirkia 22.170833 76.860000 828.35 714.403 591775.73
7 Pipariya 22.751667 78.357222 1054.50 235.634 248476.05
8 Seoni Malwa 22.454167 77.470000 1172.91 862.903 1012107.56
9 Sohagpur 22.705556 78.202778 987.84 368.478 363997.31
10 Timarani 22.377778 77.230556 1013.62 754.863 765144.23
10554.78 5273.1271 5463898.61




Fig. 2 Thissen polygon drawn utilizing the local stations is shown.


Fig.3. Thiessen polygon over Tawa command

Fig.2. Rainfall variation during the period 1992-93 to 2010-11
Histogram showing range of rainfall with corresponding
events
Histogram is a representation of frequency of
occurrence of any event in an experiment. When the annual
mean rainfall data is analyzed to assess the frequency of
occurrence of a particular rainfall amount, it is found that 75%
of the rainfall events are in the range of 751 mm to 1500 mm.
Extreme rainfall with more than 1501 mm is also witnessed in
17.5% of the rainfall events occurring in the command area.
The histogram of rainfall is shown in Fig.4.

Fig.4. Histogram of rainfall for the period 1971-2010

Statistical analysis of rainfall for the period 1971-2010
The summery of statistical mean (), standard deviation
(), variance (
2
), coefficient of variation (C
v
) and coefficient
of skewness (Ck) for weekly, monthly, and annual rainfall for
40 years (1971-2010) is presented in Table 4, Table 5 and
Table 6 respectively. The mean rainfall in the Tawa command
for all years annually is 1174.78 mm, monthly is 97.98 mm and
weekly is 22.59 mm. Mean monthly rainfall varies from 0.83
mm in the month of April to 404.27 mm in the month of July.
Whereas, mean weekly rainfall varies from 0.06 mm in the SW
16 to 97.94 mm in the SW 33. The weekly coefficient of
variation suggested that the variation is maximum during the
standard monsoon weeks (SW 26 to SW 37) indicating
fluctuating rainfall pattern in the region. Similar trend is also
seen for the four monsoon months viz. June, July, August and
September indicating atmospheric instability during monsoon.
Minimum value of standard deviation is 2.74 during April and
0.16 during SW 16. It indicates better weather stability in the
period.





Agriculture for Sustainable Development 1(1):95-102, 2013/Review Mishra et al



Table 4: Statistical analysis of weekly rainfall for the period 1971-2010
Weekly rainfall analysis for the period 1971-2010 (Tawa region)
SW Total Mean () SD () Variance (
2
) Coefficient of variation (Cv) Coefficient of skewness (Ck)
1 7.23 0.18 0.55 0.30 0.02 3.50
2 101.49 2.54 7.71 59.50 0.34 3.03
3 18.04 0.45 1.36 1.84 0.06 3.56
4 71.95 1.80 10.14 102.82 0.45 6.28
5 40.01 1.00 3.75 14.04 0.17 4.60
6 110.06 2.75 7.14 51.00 0.32 3.09
7 14.36 0.36 0.84 0.71 0.04 2.38
8 34.00 0.85 4.10 16.78 0.18 5.46
9 54.28 1.36 4.14 17.16 0.18 3.79
10 99.25 2.48 8.86 78.52 0.39 5.00
11 55.54 1.39 5.01 25.10 0.22 3.96
12 1.66 0.04 0.26 0.07 0.01 6.32
13 21.01 0.53 2.22 4.93 0.10 5.17
14 5.99 0.15 0.60 0.36 0.03 4.75
15 18.77 0.47 2.66 7.06 0.12 6.21
16 1.00 0.03 0.14 0.02 0.01 6.20
17 3.94 0.10 0.39 0.15 0.02 4.34
18 5.95 0.15 0.59 0.35 0.03 5.56
19 10.51 0.26 0.68 0.46 0.03 2.95
20 97.88 2.45 7.96 63.39 0.35 4.32
21 45.06 1.13 3.86 14.89 0.17 5.56
22 142.82 3.57 8.44 71.17 0.37 3.23
23 480.28 12.01 17.74 314.65 0.79 1.60
24 1194.18 29.85 48.14 2317.15 2.13 2.39
25 1806.12 45.15 44.31 1963.17 1.96 1.68
26 3665.50 91.64 126.64 16036.79 5.61 3.70
27 2943.26 73.58 78.63 6182.68 3.48 1.44
28 3422.48 85.56 69.12 4777.35 3.06 0.89
29 2342.16 58.55 85.52 7313.38 3.79 3.08
30 3257.04 81.43 93.94 8824.06 4.16 1.99
31 3131.03 78.28 75.55 5707.74 3.34 1.60
32 3511.01 87.78 64.06 4103.79 2.84 1.34
33 3917.67 97.94 106.68 11380.01 4.72 1.83
34 3896.47 97.41 75.76 5739.51 3.35 0.74
35 3714.33 92.86 97.35 9476.05 4.31 1.47
36 2613.57 65.34 72.23 5217.77 3.20 1.22
37 2190.53 54.76 95.58 9135.68 4.23 2.83
38 1247.00 31.17 41.25 1701.93 1.83 2.12
39 842.18 21.05 31.53 994.09 1.40 1.47
40 470.38 11.76 19.73 389.17 0.87 2.20
41 267.30 6.68 18.16 329.79 0.80 4.61
42 110.43 2.76 9.04 81.76 0.40 4.48
43 129.44 3.24 18.13 328.66 0.80 6.25
44 93.91 2.35 11.04 121.89 0.49 5.80
45 118.56 2.96 11.27 126.97 0.50 4.33
46 118.13 2.95 10.69 114.29 0.47 4.14
47 76.39 1.91 9.15 83.65 0.40 5.95
48 167.24 4.18 20.11 404.53 0.89 6.06
49 184.07 4.60 25.07 628.69 1.11 6.26
50 12.53 0.31 1.27 1.61 0.06 4.77
51 55.05 1.38 6.51 42.40 0.29 5.73
52 51.99 1.30 6.38 40.72 0.28 5.67
Weekly analysis
Mean Max Min SDmax SDmin Cvmax Cvmin Ckmax Ckmin
22.59

97.94 (SW33) 0.03 (SW16) 126.64 (SW26) 0.14 (SW16) 5.61 (SW26) 0.01 (SW12) 6.32 (SW12) 0.74 (SW34)






Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al



Table 5: Statistical analysis of weekly rainfall for the period 1971-2010
Monthly rainfall analysis for the period 1971-2010 (Tawa region)
Months Total Mean () SD () Variance (
2
) Coefficient of variation (Cv)
Coefficient of skewness
(Ck)
Jan 230.89 5.77 15.59 243.00 0.16 3.92
Feb 181.08 4.53 8.29 68.78 0.08 2.11
Mar 213.55 5.34 13.94 194.30 0.14 3.88
Apr 33.07 0.83 2.74 7.53 0.03 5.30
May 250.75 6.27 14.79 218.88 0.15 3.67
Jun 6798.40 169.96 147.32 21703.55 1.50 2.15
Jul 13237.06 330.93 180.34 32521.08 1.84 0.83
Aug 16170.77 404.27 179.48 32214.34 1.83 0.93
Sep 8022.83 200.57 179.20 32114.32 1.83 1.41
Oct 1061.14 26.53 35.73 1276.79 0.36 1.64
Nov 467.12 11.68 27.12 735.72 0.28 2.88
Dec 324.34 8.11 26.65 710.47 0.27 5.22
Monthly analysis
Mean Max Min SDmax SDmin Cvmax Cvmin Ckmax Ckmin
97.90 404.27 (Aug) 0.83 (Apr) 180.34 (Jul) 2.74 (Apr) 1.84 (Jul) 0.03 (Apr) 5.30 (Apr) 0.83 (Jul)

Table 6: Statistical analysis of annual rainfall for the period 1971-2010
Yearly rainfall analysis for the period 1971-2010 (Tawa region)
Mean Max Min SDmax Cv Ck
1174.78 1897.67 (2003) 651.81 (2000) 302.04 0.26 0.59

Table 7: Probability analysis of monthly rainfall series for the period 1971-2010 using Chi-squared (
2
) test

Month
Chi-Squared values {(Oi-Ei)^2/Ei}
Normal Log normal Gumbel Min Gumbel Max Weibull (2) Gamma Beta Log-Pearson (3) Gamma (3) Weibull (3)
Jan Inf 84.07 Inf Inf 11.76 572.61 Ext 28.34 8.59 472.98
Feb Inf 89.32 Inf Inf 65.20 3806.66 545.55 NA 247.29 Ext
Mar Inf 212.29 Inf Inf 201.19 Ext Ext NA 481.78 Ext
Apr Inf 8.41 Inf Inf 22.74 23.76 Ext NA 73.11 Ext
May Inf 307.29 Inf Inf 510.95 Ext Ext NA 546.11 Ext
Jun Inf 151.68 Inf 50.40 130.11 99.94 204.58 201.01 Ext Ext
Jul 10.36 86.00 36.84 38.43 40.24 35.19 10.44 1.12 Ext Ext
Aug 9.18 36.16 30.69 31.44 28.91 21.72 28.2 1237.83 Ext Ext
Sep Inf 291.64 Inf 301.53 211.23 195.34 272.57 189.29 Ext Ext
Oct Inf 802.19 Inf Inf 336.71 549.81 Ext NA 3957.40 Ext
Nov Inf 1029.58 Inf Inf 820.60 Ext NA NA 2155.60 Ext
Dec Inf 560.37 Inf Inf 472.82 NA Ext NA 1236.61 Ext
Total Inf 3659 Inf Inf 2852.46 Ext Ext NA Ext Ext
Inf: Infinity; Ext: Extremely high; NA: Distribution not fitted
Table 8: Probability analysis of Annual, Standard Non-monsoon and Monsoon weeks of rainfall series for the period 1971-2010 using Chi-squared test
Period
Chi-Square values {(Oi-Ei)^2/Ei}
Normal Log normal Gumbel Min Gumbel Max Weibull (2) Gamma Beta Log-Pearson (3) Gamma (3) Weibull (3)
Yearly 13.95 6.87 40.68 9.31 21.18 8.09 18.32 7.34 7.81 8.96
SNMW Inf 261.28 Inf Inf 178.39 185.84 1294.73 NA 179.44 405.2
SMW 9.06 16.52 31.05 31.15 11.77 13.50 2.36 65.07 14.32 8.25
Inf: Infinity; NA: Distribution not fitted; SNMW: Standard non-monsoon week; SMW: Standard monsoon week

Annual rainfall of Tawa command indicates minimum
of 651.81 mm during the year 2000 and maximum of 1897.67
mm in the year 2003. The standard deviation of annual series of
rainfall is 302.04 mm. Annual coefficient of variation and
coefficient of skewness in the region are 0.26 and 0.59
respectively.



Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al



Probability analysis of rainfall for the period 1971-2010
Probability analysis of rainfall series is done to know
the assured rainfall in a region at different probability level, so
that a risk free irrigation crop planning can be formulated.
Probability analysis in the Tawa command is performed
considering 40 years of daily rainfall data for the period 1971-
2010. Data were converted into weekly, monthly and annual
series before finding the best fit distribution in the region in a
spread sheet programme.
In the study, a total of 9 distributions, both 2-
parameters and 3-parameters, were fitted in the weekly,
monthly and yearly series arranged earlier. All the distributions
have been discussed in Table 1 with their Probability Density
Function (PDF), Range and Parameters.
It is mandatory to test the goodness-of-fit of a
probability distribution for identifying the best distribution
which represent the region, to extrapolate a data series. Here,
Chi-squared (
2
) test was carried out to test the goodness of fit
of the probability distributions considered in the study. The
relationship of Chi-squared (
2
) test statistic is given below:
(2)
Where,
n = No. of observations; O
i
= The observed values in the i
th

observation; E
i
= The expected (computed) values in the i
th

observation.
Probability analysis of standard monsoon weeks,
non-monsoon weeks, monthly and annual series of rainfall was
made using nine probability distributions namely Normal, Log
normal, Gumbel Min, Gumbel Max, Weibull (2), Gamma,
Beta, Log Pearson (3), Gamma (3) and Weibull (3) distribution.
Goodness-of-fit of these distributions was tested by Chi-
squared (
2
) test and values are presented in Table 7 and Table
8.As revealed from the Tables, the annual series of the rainfall
data is best represented by Log normal distribution with
minimum
2
-value of 6.87, where as the standard monsoon and
non-monsoon weeks are represented by Beta distribution and
Weibull (2) distribution respectively. Probability analysis of
monthly series reveals that Weibull (2) distribution is the best
fit distribution representing the Tawa command. The monthly
rainfall at different probability level by two parameter Weibull
distribution is shown in Fig.5. Rainfall at 70% probability can
be taken as assured in the crop planning in the Tawa command.
CONCLUSIONS
Proper crop planning in an irrigated area requires
quality analysis of rainfall data as well as its correct
interpretation. A poor, faulty analysis affects the sowing date,
rainwater availability, irrigation need, and ultimately the net
returns from the command. In the present study, 40 years daily
rainfall data were analyzed for the Tawa irrigation project on
weekly, monthly and annual basis, to arrive at analytical
conclusion for better irrigation planning. The analysis of data in
the region confirmed the variability in the mean rainfall
received; hence the impact of climate change cannot be totally
ignored. The command area is on high use since its inception
from 1978. A lot of changes have occurred in the mean time in
its cropping pattern and water utilization.The statistical analysis
of rainfall data in the region confirms a instable monsoon
system with mean annual rainfall of 1174.78 mm, mean
monthly rainfall variation of 0.83 mm in the month of April to
404.27 mm in the month of July and mean weekly rainfall
variation from 0.06 mm in the SW 16 to 97.94 mm in the SW
33.
The probability analysis of weekly, monthly and
annual series of rainfall data was carried out by employing nine
probability distributions namely Normal, Log normal, Gumbel
Min, Gumbel Max, Weibull (2), Gamma, Beta, Log Pearson
(3), Gamma (3) and Weibull (3) distribution. Goodness-of-fit
of these distributions was tested by Chi-squared (
2
) test. The
annual series of the rainfall data is best represented by Log
normal distribution with minimum
2
-value of 6.87, where as
the standard monsoon and non-monsoon weeks are represented
by Beta distribution and Weibull (2) distribution respectively.
Probability analysis of monthly series reveals that Weibull (2)
distribution is the best fit distribution representing the Tawa
command.

Fig.5. Monthly rainfall at different probability level by Log-
normal distribution

ACKNOWLEDGEMENT
The authors, thankfully, acknowledges the Tawa Command
Authority, Hoshangabad for providing rainfall station data due
to which quality analyses of rainfall in the study have been
made possible. Lastly, we acknowledge our thankfulness to
State Data Centre, Bhopal, India for supporting during the
study.









Agriculture for Sustainable Development 1(1): 95-102, 2013/Review Mishra et al


REFERENCES
Ashkar, F. and Mahdi, S. 2003. Comparison of two fitting
methods for the log-logistic distribution. Water
Resources Research. 39 (8): 1217-1225.
Bhakar, S.R., Bansal, A.N., Chhajed, N., and Purohit, R.C.
2006. Frequency analysis of consecutive days
maximum rainfall at Banswara, Rajasthan, India. ARPN
Journal of Engineering and Applied Sciences. 1(3): 64-
67.
Biswas, B.C. and Khambeta, N.K. 1989. Distribution of short
period rainfall over dry farming tract of Maharashtra.
Journal of Maharashtra Agricultural University.
Directorate of Economic and Statistics, Department of
Agriculture and Cooperation, Ministry of Agriculture,
GoI, 2011.
Durbude, D.G. 2008. Estimation of probable maximum
precipitation for planning of soil and water conservation
structures. Journal of Soil and Water Conservation. 7
(3): 31-35.
Fan, F. and Mann, M.E. 2008. Understanding changes in the
Asian summer monsoon over the past millennium:
Insights from a long-term coupled simulation. Journal
of Climate. 22: 1736-1748.
Fisher, R.A. 1924. The influence of the rainfall on the yield of
wheat at Rothamsted. Philosophical transaction of the
Royal Society of London, Series B, Vol. 213.
Kulandaivelu, R. 1984. Probability analysis of rainfall and
evolving cropping system for Coimbatore. Mausam, 5
(3): 257-258.
Mokashi, D.D., Deshpande, A.N., Khadtare, S.V., and Jadhav,
J.D. 2008. Rain:Its probability based forecast at
Solapur, India. Internatational Journal of Agricultural
Sciences. 4(1): 390-393.
National Institute of Hydrology, NIH 1998. Statistical analysis
of rainfall in Sagar division. CS (AR) 5/98-99, pp 1-96.
National Rainfed Area Authority, NRAA, GoI 2011.
Challenges of food security and its management.
Position paper, 5: 1-45.
Nemichandrappa, M., Balakrishnan, P. and Senthilvel, S.
(2010) Probability and confidence limit analysis of
rainfall in Raichur region. Karnataka Journal of
Agricultural Sciences, 23 (5): 737-741.
Papalexiou, S.M. and Koutsoyiannis, D. (2006) A probabilistic
approach to the concept of probable maximum
precipitation. Advances in Geosciences, 7, pp 51-54.
RamaRao, B.V., Kavi, P.S. and Sridharan, P.C. (1975) Study of
rainy days and wet spells at Bijapur. Annals of Arid
Zone, 14 (4): 371-372.
Ramasastri, K.S. and Nirupama, P. 1987. Statistical analysis of
rainfall in Belgaum district, Karnataka, National
Institute of Hydrology, Technical report, TR-4.
Sharma, M.A. and Singh, J.B. 2010. Use of probability
distribution in rainfall analysis. New York Science
Journal, 3(9): 40-49.
Upadhaya, A., and Singh, S.R. 1998. Estimation of consecutive
days maximum rainfall by various methods and their
comparison. Indian Journal of Soil Conservation, 26
(2): 193-200.
Virmani, S.M., Sivakumar, M.V.K. and Reddy, S.J. 1982.
Rainfall probability estimates for selected locations of
semi-arid India. Research BulletinNo.1, 2
nd
Edit.
enlarge., International Crops Research Institute for the
Semi-Arid (ICRISAT).
Weerasinghe, K.D.N. 1989. The rainfall probability analysis of
Mapalana and its application to agricultural production
in the area. Journal of the National Science
Council of Sri Lanka, 17 (2): 176-183.