Using the step linear regression at the analysis of hydrological conditions of the Fryšávka drainage basin

key words: runoff curve number, correlation index, landscape cover, DesQ–MAXQ model, runoff, regression and correlation analysis, retention

Summary:

Problems of the retention capacity of the landscape and related runoff con-ditions are at present, at the time of the occurrence of frequent storm floods, a topical problem. Our research was aimed at the evaluation of effects of physical-geographical factors on hydrological conditions of the Fryšávka River drainage area occurring in the southern part of the Žďárské vrchy CHKO (Žďárské Hills Protected Landscape Area) depending on the current position of the landscape cover. Preparation and analyses of source data were carried out in ArcGIS 9.2 in the ArcView software product by means of the set of integrated software applications ArcMap, ArcCatalog and ArcToolbox. To determine hydrological conditions at the absence of hydrometric observations a method was used of numbers of runoff curves CN in a modification of the DesQ-MAXQ deterministic episode model. Based on hydrological characteristics, potential retention was monitored as well as the volume of flood wave, concentration time and peak discharge. Hydrological modelling was carried out on a design storm precipitation with the return period of 100 years. The significance of physical-geographical factors affecting the landscape retention capacity and runoff conditions was quantified by the statistical method of multiple regression and correlation analysis on the set of 95 partial drainage areas of the studied region. The analyses were carried out using the UNISTAT system. The method makes possible to select independent variables by gradual steps from most important down to least important. Finally, using the step linear regression, a certain number of independent variables was selected mostly affecting the size of a respective dependent variable together with the quantification of their operation through regression coefficients. Results achieved could be used at the implementation of preventive measures resulting in increasing the retention capacity of partial drainage areas, which served for the registration of major runoffs from a unit area.

 

 

Citation:

Mašíček T., Toman F., Palát M. 2011, vol. 8. Using the step linear regression at the analysis of hydrological conditions of the Fryšávka drainage basin. Infrastruktura i Ekologia Terenów Wiejskich. Nr 2011, vol. 8/ 11