A local correlation integral method for outlier detection in spatially correlated functional data

Jorge Sosa*, Paula Moraga, Miguel Flores, Jorge Mateu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


This paper proposes a new methodology for detecting outliers in spatially correlated functional data. We use a Local Correlation Integral (LOCI) algorithm substituting the Euclidean distance calculation by the Hilbert space L2 distance weighted by the semivariogram, obtaining a weighted dissimilarity metric among the geo-referenced curves, which takes into account the spatial correlation structure. In addition, we also consider the distance proposed in Romano et al. (2020), which optimizes the distance calculation for spatially dependent functional data. A simulation study is conducted to evaluate the performance of the proposed methodology. We analyze the role of a threshold value appearing as an hyperparameter in our approach, and show that our distance weighted by the semivariogram is overall superior to the other types of distances considered in the study. We analyze time series of Land Surface Temperature (LST) data in the region of Andalusia (Spain), detecting significant outliers that would have not been detected using other procedures.

Original languageEnglish (US)
Pages (from-to)1197-1211
Number of pages15
JournalStochastic Environmental Research and Risk Assessment
Issue number3
StateAccepted/In press - 2023

Bibliographical note

Publisher Copyright:
© 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.


  • Functional data analysis
  • LOCI
  • Outlier detection
  • Random fields
  • Spatial correlation
  • Spatial statistics

ASJC Scopus subject areas

  • Environmental Engineering
  • Environmental Chemistry
  • Water Science and Technology
  • Safety, Risk, Reliability and Quality
  • General Environmental Science


Dive into the research topics of 'A local correlation integral method for outlier detection in spatially correlated functional data'. Together they form a unique fingerprint.

Cite this