Longitudinal analysis of spatially correlated data

摘要

Nowadays, the detection of areas with high nitrate concentration in groundwater is of public concern because an excess of these nitrogen compounds in any area represents serious environmental pollution. Hence, governmental agencies require geostatistical researchers to develop methods for assessing nitrate concentration. In this paper, we propose a linear mixed model which incorporates both spatial as well as longitudinal information with the particular objective of detecting areas with excess of nitrate in groundwater. The model extends in a natural way the traditional kriging approach, incorporating the longitudinal dimension. We illustrate the results using real data taken from wells pouring into the Ebro river and its tributaries in the province of Navarra, Spain.