Regional streamflow estimation by standard regional dependence function approach

The main purpose of this paper is to propose a standard regional dependence function (SRDF) based on concepts of semivariogram and especially point cumulative semivariogram for regional streamflow estimation. The SRDFs are obtained from available spatial data and show regional dependence, which decreases with distance from a given site. These functions present quantitatively the regional dependence of the streamflow phenomenon recorded at irregular sites over a drainage basin and provide a unique opportunity for the establishment of regional objective estimation method based on weighed averages. The weightings are obtained by means of the SRDF given the distance between any two sites. The implementation of the proposed methodology is presented for some streamflow records from the Lower and Upper Mississippi River watershed in the United States. For the application, the experimental SRDF forms are first obtained from the available data, and these are then employed directly in the regional estimation procedure. The study indicated that the use of all the stations in a region for the estimation at any particular station is rather naive because far away stations (more than 1,000 km away) are taken into consideration. The final conclusion is that discharge at any particular station is better described as a function of discharge at several (3-5) closest stations. The reliability of the method is measured through the cross validation procedure, and it is observed that the procedure yields streamflow predictions with less than 10% relative error.