Empirical Orthogonal Function analysis of the palmer drought indices

TATLI H., Turkes M.

AGRICULTURAL AND FOREST METEOROLOGY, vol.151, no.7, pp.981-991, 2011 (Peer-Reviewed Journal) identifier identifier

  • Publication Type: Article / Article
  • Volume: 151 Issue: 7
  • Publication Date: 2011
  • Doi Number: 10.1016/j.agrformet.2011.03.004
  • Journal Indexes: Science Citation Index Expanded, Scopus
  • Page Numbers: pp.981-991


The aim of the paper is to employ Empirical Orthogonal Function (EOF) analysis to drought indices including Palmer Drought Severity Index (PDSI), Palmer Hydrological Drought Index (PHDI), Palmer Moisture Anomaly (Z) Index (Z-Index), Weighted Palmer Drought Severity Index (WPDSI), and Water Deficit (P-PET) Index and to compare their resultant spatial patterns across Turkey. In this respect, the PDSI, PHDI, Z, WPDSI, Aridity Index (AI) and P-PET Index values are calculated based on observed monthly temperature and precipitation values of the 96 Turkish meteorological stations for the 1929-2009 period, and gridded available water holding capacities (AWHC) in 1-m soil depth are extracted from the soil data sets of the Oak Ridge National Laboratory Distributed Active Archive Centre (ORNL DAAC). By considering the strong match among the PDSI, PHDI, and WPDSI based on the results of the patterns of the first EOFs (EOF1) with the highest eigenvalue, we suggest that using and applying one of these drought indices could be adequate for a drought analysis in Turkey. On the other hand, it is evident that the significant results from the EOF2, EOF3 and EOF4 loadings of the PHDI and WPDSI indices have a similar pattern across Turkey. Therefore, with respect to the geographical autocorrelation and magnitudes of the loading values, it is explained that the PDSI and WPDSI in particular could have originated from the same physical process, and one may have a preference for the effective usage of a combination of PDSI, PHDI and Z-Index. Consequently, in a multi-purpose medium- and long-term drought management plan for Turkey or in any country having similar climatic and physical geographical conditions, using one of the three drought indices along with the Z-Index could be acceptable for successful applications. (C) 2011 Elsevier B.V. All rights reserved.