Statistical Analysis of Rainfall Event in Seonath River Basin Chhattisgarh

To understand the mechanism of potential hydrologic impacts of climate and land use land cover changes. In this study identified to the significant temporal trend was carried out monthly, seasonal timescales. Using the rainfall data of 39 Meteorological stations under entire seonath basin which is subdivided into five weighted stations with the help of by creating Thiessen polygon over Seonath River, Chhattisgarh state of 32 years for a period of 1980-2012. Hydro metrological variables are analysed by using a combined Mann-Kendall/Thiel-Sen slope estimator trend detection approach. The results reveal a significant decreasing trend for the month of January, February, March, and may, September, October and December for all the five gauging stations similarly the month of April, June, July, August and November show the significant increasing trend. For all the five gauging stations excluded month of January in simga gauging station shows having no trend. Whereas the trend in different seasons are follows (1) season of winter and pre-monsoon shows significant decreasing trend for all the five gauging stations and post-monsoon season andhyakore, pathridih and simga gauging stations also represents the decreasing trend and (2) the season of monsoon shows increasing trend for all the five gauging stations and seasons of winter and pre-monsoon of ghatora and simga stations also shows the increasing trend.


INTRODUCTION
Change in climate is a long-term phenomenon and most alarming issues for the entire world, Therefore; quantification of climate changes has become preliminary. Identification of temporal trend of hydrometrological variables is important to sustainable management and development of water resources in future. Rainfall is one of a most important event of the hydrological cycle.
[1] This depends on the changes in the variations of atmospheric concentration of greenhouse gaseous. A number of studies done on rainfall trend analysis in various regions. Most of the paper used nonparametric approaches for detecting trend and few of the paper have used a linear regression test. In this study we have to use non-parametric methods because they are distributed free and mostly used non-parametric method which is recommended for detection of significant monotonic upward and downward trend for different climatologic and hydrologic time scale by World Meteorological Organization (WMO) and Thiel-Sen slope estimator is mostly used for estimating magnitude of linear trend and it has been also most commonly used for detecting the magnitude of linear trend in hydrometeorological time series. [2]

II.
STUDY AREA AND DATA USED The study area is the seonath river basin of Chhattisgarh state, India.It is a major tributary of Mahanadi river which is situated between 20֯ 16'N to 22֯ 41'N Latitude and 80֯ 25'E to 82֯ 35'E Longitude it consists a large portion of the upper Mahanadi valley and its traverse length of 380 kilometres.The area of the basin is 30560 square kilometres. The Monthly precipitation data of 39 Meteorological stations for whole seonath river basin for a period of 32 years i.e. 1980-2012 is collected from Department of state data centre Water Resources, Raipur (Chhattisgarh) this data is then analysed for detecting if the trend is monotonic increasing or decreasing. The latitude, longitude and area (Thiessen polygon) for each station is shown in table Moreover, the location of each station in Seonath river basin map is represented in figure

STUDY OF TIME SERIES
A time series is a combination of statistics, usually collected at a regular interval ( average monthly rainfall, annual rainfall temperature) and it occurs naturally in many application areas ( precipitation, rainfall, temperature).The method of time series analysis pre-date those for general stochastic processes and Markov chains.
[3] The objective of time series analysis is to describe and summarise the time series data and assembles to low dimensional model and to predict the future forecast.
3.2 TREND ANALYSIS It is a statistical method which is most commonly used for studying the temporal trend of hydroclimatic timescales. In the present study, trend detection analysis has been done by using non-parametric (Mann-Kendall test and Thiel Sen Slope estimator test).A non-parametric test approach is taken into account against the parametric one because it is distribution-free, robustness against outliers. [4]

IV. METHODS FOR TREND DETECTION ANALYSIS 4.1 Testing significance of trend by MK test
The purpose of the Mann-Kendall (MK) test is to statistically assess if there is a monotonic upward or downward trend of the variable of interest over time series. A monotonic upward (downward) trend means that the variables consistently increase (decreases) through time, but the trend sometimes may or may not be linear. An assumption not required by the MK test, that is the MK test is a nonparametric (distribution-free) test. In MK test statics we have to use Signum function and correlate the variables if X1<X2 =-1, ifX1=X2, =0 and if X1>X2,=1and by adding all this correlation values to detect whether significant increasing or decreasing trend in the given series In MK test statics the sample size should not be less than 4. It is the most widely used for the analysis of detecting a trend in climatologic and hydrologic time scales.
[5] The test is based on statistic S defined as follows: Where N is the number of data point in the given time series, xi and xj are the data values at time scale I and j (j>i) respectively. This test statistics represent a number of positive differences minus negative of differences for considering all the differences. And denoted by, For sample (N>4), the distribution of S is assumed to follow the normally distributed with variance and zero mean.
Where N is the number of tied (the difference between compared values is zero) group and tk is the number of data points in the kth group of tied. The Z-statistics or standard normal deviate is computed by using the equation: Here, if the value of │Z│>Z' then the null hypothesis of no trend is rejected at 0.05% of significance level in a two-tailed test. (Trend is significant), in this study, a positive value of Z represents an upward or increasing trend and negative value of Z represents the downward or decreasing trend. [5]

THIEL SEN'S SLOPE ESTIMATOR TEST
Thiel-Sen's slope estimator is useful for estimating the changes in the amount of trend and it has been most commonly used for detecting the magnitude of a linear trend in hydro-meteorological time series. Here, the Slope (Ti) of the entire data group is computed as follows.    Vol-5, Issue-1, Jan-2018]  https://dx.doi.org/10.22161/ijaers.5.1.20  ISSN: 2349-6495(P) | 2456-1908(O) www.ijaers.com Page | 141

VI. CONCLUSION
Change of climate is very preliminary to cope up with ever-changing conditions. The trend analysis is done for seonath river basin of Chhattisgarh state for monthly rainfall data for the period of 1980-2012 is analysed by using a non-parametric Mann-Kendall and Theil-Sen Slope Estimator test. The results reveal a downward or decreasing trend for most of the months and seasons of a year for the period of 1980-2012 under the analysis. In this study, we noticed that the rainfall events in the whole seonath river basin are continuous decreases. Both from Mann-Kendall's and Sen Slope estimator test. The adverse effect of the observed decreasing trend in rainfall event may be expected for different water-related sectors, primarily rain-fed agriculture and freshwater availability in the region. This study suggests that the knowledge of changes in rainfall pattern and its periodicity estimation could be useful for the hydrologist and management of irrigation planners for more efficient utilisation of water in the region and to make an appropriate decision on cropping pattern.