Abstract

CAMPOS-ARANDA, Daniel Francisco. Estimation of Monthly Runoff in Humid Climates Using Regression Models. Tecnol. cienc. agua [online]. 2015, vol.6, n.2, pp.113-130. ISSN 2007-2422.

Mathematical modeling of the rainfall-runoff relationship (RRR) is indispensable when temporal and spatial data are scarce. Ungauged basins is one example of a lack of data at sites of interest. And future records of induced or natural hydrological changes in a basin is an example of data that cannot be measured. In both cases, the use of a regional RRR model makes it possible to perform the needed evaluations. The simplest model for estimating monthly runoff volume is a monthly polynomial regression, which can model a linear or curved RRR. In addition, this method can include the delay in monthly runoff by averaging antecedent precipitation. The present study fitted a monthly regression model to the joint set of precipitation and runoff data from the Tancuilin and El Cardon hydrometric stations in Partial Hydrological Region 26 (Lower Panuco River), with records containing 33 an 37 years respectively. The study found that the monthly coefficients of the regression models can be regionalized based on the average runoff coefficient. The comparisons performed show that regionalized regression models provide an excellent estimation of monthly runoff, accurately reproducing average monthly values. They also provide a good approximation of the dispersion in small and medium basins located in humid climates.

Keywords : Linear regression; linear correlation coefficient; transport factor; monthly average runoff coefficient; statistical parameters; confidence interval of prediction; determination coefficient of prediction.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )