Reduction of a Groundwater Level Network 'Source Title' Methods and Instrumentation for the Investigation of Groundwater Systems
Bibliographic details
Summary
A practical prediction algorithm has been developed to discriminate between independent or primary locations and dependent for secondary locations of a groundwaterlevel observation network. The selected criterium is an objective measure: a value of 15 cm for the root mean squard erro of the prediction erro in the model validation period. the algorithm is based on the kalman filtering technique. After and introduction of the theoretical model thelink will be made between the theoretical model and its practical use for network reductio. Four applications directed to network density and observation frequency will be described. This paper is solely concerned with reduction of a network. Netork reinforecement at unmeasured sites will be discussed in the paper: spatial uncertainty in piezonetric haad