Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Titre  : Stochastic Modelling Of hydrological Drought In Pennar Basin

Auteur : Radha Krishna Murthy.B

Université de soutenance : Sri Venkateswara University

Grade : Doctor of Philosophy (PhD) 2000

Conclusions
per Pennar Project stream flows. 9.2 Conclusions Conclusions drawn from the rainfall-runoff relationship study are, i. There exists a strong correlation between present month runoff and present month rainfall. ii. Study revealed that number of rainy days can be excluded from the regression analysis and iii. Results obtained from co-axial graphical correlation are comparable with that obtained from regression analysis. The following conclusions are drawn from the modeling of annual rainfall in the basin, i. Autoregressive models preserve mean, standard deviation of historical series. ii. The choice of the model depends on the degree of complexity. As ARI model satisfy all the criteria, it is used in simulating the drought in the basin. The following paragraphs enunciate the conclusions drawn from the modeling of drought using annual rainfall series. i. The rainfall in the basin is positively skewed. The negative first correlation coefficient indicates high rainfall years followed by low rainfall years and vice versa. ii. All stations have identical coefficient of variation indicating the homogeneity of region with respect to rainfall. iii. Wide range of variation of coefficient of variation, skew and first correlation coefficient signifies that rainfall series are not independent normal. iv. Entire basin is drought prone from mild to severe intensity. v. Duration of mild drought varied from 3 to 10 years, whereas for moderate drought it varied from 1 to 7 years and for severe drought is one year only. vi. The study reveals the suitability of transformation functions to normalize the series (Box-Cox and TSPT). For obtaining the truly normal series it is necessary to use both transformations. vii. Gumbel distiribution fits well for rainfalllstream flow drought severity and duration. Log nonnal distribution fits well for the case of drought severity only. The identified droughts by modified procedure of Herbst et.al (1966) are concurrent with historicilly realized droughts, thus providing the viability of the methodology in identification of drought conditions. It is hoped that the present study can be applied with confidence to characterize the drought in the long term perspective of water resources planning and management of Pennar basin.

Présentation et version intégrale