PROBABILITY ANALYSIS OF DRY AND WET SPELLS AT HYDERABAD AND MAXIMUM RAINFALL DISTRIBUTION

被引：0

作者：

Sharma, G. C. ^{[1
]}

机构：

[1] Project Directorate Farming Syst Res, Meerut 250110, Uttar Pradesh, India

来源：

INTERNATIONAL JOURNAL OF AGRICULTURAL AND STATISTICAL SCIENCES | 2011年 / 7卷 / 02期

关键词：

Rainfall probability distribution function; Markov chain model; Transitional probability; Log-Normal; Log Pearson III and Gumbel distributions;

D O I：

暂无

中图分类号：

S [农业科学];

学科分类号：

09 ;

摘要：

A study was carried out to develop forecast model for rainfall dry and wet spells. The study was carried out for weekly rainfall for the period from 1983 to 2008 at ANGRAU, Hyderabad. Markov chain probability model has been used extensively to determine the long-term frequency behaviour of Wet and dry weather spells: The probabilities occurrence of two or more dry/wet weeks preceded by dry/wet weeks has been evaluated. The probabilities of two or more consecutive dry/ wet weeks have been worked out for agricultural policy makers and also forward and backward accumulations of rain water suitable for crop production. Further, annual maximum weekly rainfall at different probability levels were predicted using three different probability distribution functions, viz., Log-Normal, Log Pearson III and Gumbel. The Chi-square test statistic for goodness of fit showed Log Pearson III is the best in predicting annual maximum weekly rainfall.

引用

页码：517 / 525

页数：9