A Bayesian Approach for the Rainfall-Runoff Problem in Multiple Basins

X Congreso latinoamericano de probabilidad y estadística matemática. CLAPEM. Lima-Perú.

We present an analysis of runoff and rainfall data from Rio Grande, a basin located in the northeast of Brazil. The main problems we face here are:

1. to model runoff and rainfall jointly, taking into account their different spatial units,
2. to use stochastic models where all the parameters have physical interpretations, and (iii) to model multiple runoff series simultaneously.

The intrinsically uncertain nature of those hydrological processes makes Bayesian analysis natural within this field. Our approach is based in dynamic models. For run-off, we use transfer functions, and for rainfall, we use a change of support from a multisite spatial model. Besides the computational effort to implement the proposed models, it is worth pointing out that some methodological novelties are also implemented.

The data consist of monthly recorded series from January 1984 to September 2004, at three runoff stations and nine rainfall stations irregularly located in an area of drainage of 37522.48 km2. Model assessment, spatial interpolation and temporal predictions were part of our analysis. Results has shown that our approach is a promising tool for the runoff-rainfall analysis.