Return to search

Regionalization of average annual runoff models for ungaged watersheds in arid and semiarid regions

A prevailing problem in applied hydrology is the estimation of runoff from ungaged small watersheds and drainge basins. In this study, arid and semiarid regions were Grouped according to their climatic, geomorphologic, and soil characteristics, disregarding their geographic position. Eighty watersheds were used in this study from three countries: U.S., Brazil, and Australia. Twenty-two climatic, geomorphologic and soil variables were used for the delineation of homogeneous Groups in the cluster analysis, and two major Groups were defined. The results suggest that homogeneous Groups can be delimited independently of their geographic position. Cluster analysis and Andrews' plot were used for regionalization of the watersheds. The variables used for development of the models for each Group were selected by stepwise multiple regression analysis. The Andrews' plot further examination reinforce the statement that hydrologically similar watersheds are independent of their geographic position. In a preliminary study 60 watersheds were used to determine the most important variables. For Group I, the stepwise multiple regression analysis reduced the available 21 independent variables to three variables: rainfall, soil permeability index, and temperature. For Group II, only two variables were statistically significant (rainfall and watershed form factor). Once the most significant variables were selected, 20 additional watersheds with data were also included in the final study. Upon evaluation of the regression statistics, Group II responded better than Group I. The equations were: UNFORMATTED EQUATION FOLLOWS: Group I "Dry" Q = -68.476 + 0.0784 P + 4.131Temp -3.950Slpr n = 29 R² = 70% SE = 11.16 mm/yr Group II "Wet" Q = 1.29*10⁻¹¹* P⁴·⁴¹* Rf⁻⁰·⁰⁹³ n = 37 R² = 79% SE = 30.52 mm/yr UNFORMATTED EQUATION ENDS Computed annual runoff values for Group II showed a good agreement with observed values, suggesting that the developed equation is good for prediction of the annual runoff water yield. In contrast, predicted values for the Group I showed poor agreement with the observed values, suggesting that the model should be treated with caution.

Identiferoai:union.ndltd.org:arizona.edu/oai:arizona.openrepository.com:10150/282559
Date January 1997
CreatorsAndrade, Eunice Maia de, 1956-
ContributorsHawkins, Richard H.
PublisherThe University of Arizona.
Source SetsUniversity of Arizona
Languageen_US
Detected LanguageEnglish
Typetext, Dissertation-Reproduction (electronic)
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Page generated in 0.0088 seconds