The effect of the length of record on the standardized precipitation index calculation

Hong Wu, Michael J. Hayes, Donald A. Wilhite, Mark D. Svoboda

Research output: Contribution to journalArticlepeer-review

Abstract

The effect of the length of record on the standardized precipitation index (SPI) calculation was investigated by examining correlation coefficients, the index of agreement, and the consistency of dry/wet event categories between SPI values derived from different precipitation record lengths. The effect was also illustrated by comparing SPI values derived from different lengths of record in some severe drought and flood years, and by comparing drought intensities based on SPI values temporally and spatially. Furthermore, the reason for the SPI value discrepancy was explored by investigating the changes of the shape and scale parameters of the gamma distribution when different lengths of record were involved. The gamma distribution is a frequency distribution of climatological precipitation time series and is the basis of the SPI calculation used in this analysis. The results show that SPI values computed from different lengths of record are highly correlated and consistent when the gamma distributions of precipitation over the different time periods are similar. However, the SPI values are significantly discrepant when the distributions are different. It is recommended that the SPI user should be aware of the numerical difference of the SPI values if different lengths of record are used in interpreting and making decisions based on SPI values.

Original languageEnglish (US)
Pages (from-to)505-520
Number of pages16
JournalInternational Journal of Climatology
Volume25
Issue number4
DOIs
StatePublished - Mar 30 2005

Keywords

  • Gamma distribution
  • Length of precipitation record
  • Nebraska
  • Standardized precipitation index (SPI)
  • Time scale

ASJC Scopus subject areas

  • Atmospheric Science

Fingerprint

Dive into the research topics of 'The effect of the length of record on the standardized precipitation index calculation'. Together they form a unique fingerprint.

Cite this