Thumbnail
Access Restriction
Open

Researcher Girish, J.
Advisor Munirajappa, R.
Source KrishiKosh-Indian National Agricultural Research System
Content type Text
Educational Degree Master of Science (M.Sc.)
Publisher University of Agricultural Sciences
File Format PDF
Language English
Subject Domain (in DDC) Technology ♦ Agriculture & related technologies
Subject Keyword Agricultural Statistics and Informatics ♦ Prediction of Seasonal Rainfall Using Statistical Model In Eastern Dry Zone
Abstract The most important climatic factor for agricultural production is based on the distribution and the amount of rainfall. Accurate and timely forecast of rainfall distribution and its quantum is vital in planning policy making. Hence an attempt is made to predict the possible amount of rainfall and studying the seasonal distribution of rainfall for selected areas of eastern dry zone of Karnataka. The data collected for the study is amount of monthly rainfall from ZARS, GKVK, Bangalore district and ARS, Chinthamani, Kolar district for 30 years from 1978 to 2007. The gamma distribution is used to model the distribution o f seasonal rainfall and rainfall of individual months o f the South-west monsoon season. Fitting o f gamma distribution is done by estimating the scale and shape parameters by method of moments. Chi-square test results imply that data in all the cases fallows the gamma distribution. In all the cases values o f a > 1, this implies that the distribution is less skewed and the value of scale parameter ( /? ) are less than one and small values of /3 implies that the less spread o f values and squeezing distribution the In prediction of South-west monsoon rainfall, the amount of South-west monsoon rainfall of current year (June to September) is the response variable and regressor variables are, amount of rainfall in North-east monsoon of previous year (October to January), amount of rainfall in summer of current year (February to May) and amount of rainfall in South-west monsoon of previous year (June to September). In both the cases intercept and coefficient of amount of rainfall in North-east monsoon of previous year were found to be significant. Further the R2 values for ZARS, GKVK and ARS, Chinthamani found to be 0.7392 and 0.7873 respectively. ARIMA model (2 0 2) was identified as suitable and the predictability of this ARIMA model is compared with developed regression model using MSE and AIC. Regression model found to be better than ARIMA model for data o f both the locations under study.
Educational Use Research
Education Level UG and PG
Learning Resource Type Thesis
Publisher Place Bangalore
Size (in Bytes) 847.09 kB
Page Count 120