DESIGN OF A SAMPLING NETWORK FOR AN ESTUARY IN THE COLOMBIAN CARIBBEAN

A network for monitoring physical chemistry and biological variables in the Ciénaga Grande de Santa Marta estuary, in the Caribbean coast of Colombia, was designed. Jnitially, through systematic sampling on a square grid. a set of 115 sampling points was chosen to measure the variables considered. Based on the data provided, a spati al auto-correlation structure for each variable was estimated through the semivariance function. Later, for different size networks, the kriging prediction variances were calculated, taking the adjusted semivariogram models as a basis. The compare son among the prediction variances for the different networks and their associated costs allowed establishing a set of sampling sites, that ata reasonable cost, substantially diminishes the prediction error for the variables of interest.