An automated decision support system for aided assessment of variogram models

Emanuele Barca, Emilio Porcu, Delia Bruno, Giuseppe Passarella

Research output: Contribution to journalArticlepeer-review

28 Scopus citations

Abstract

In the present paper, an extensive cross-validation procedure, based on the analysis of numerical indices and graphical tools, is described and discussed. The procedure has been implemented in a software application designed to support practitioners in the variogram model assessment. It provides an extensive report, which summarizes a large post-processing stage and suggests how to interpret the performed analysis to rate the model to be validated. Besides classical accuracy indices, two new integrated tools based on the variogram of residuals are introduced, which take the spatial nature of the dataset into account. Finally, inspecting the summary report, the user can decide whether the considered model is satisfactory for his/her goals or it needs to be improved. Finally, a case study is presented related to the variogram assessment of groundwater level measured in a porous shallow aquifer of the Apulia Region (South-Italy).

Original languageBritish English
Pages (from-to)72-83
Number of pages12
JournalEnvironmental Modelling and Software
Volume87
DOIs
StatePublished - 1 Jan 2017

Keywords

  • Decision-support system
  • Extensive cross-validation
  • Geostatistics
  • Rank coefficient of spatial correlation

Fingerprint

Dive into the research topics of 'An automated decision support system for aided assessment of variogram models'. Together they form a unique fingerprint.

Cite this