Verification of abrupt and gradual shifts in Iranian precipitation and temperature data with statistical methods and stations metadata

Environ Monit Assess. 2021 Feb 22;193(3):139. doi: 10.1007/s10661-021-08925-2.


Climate time series may exhibit abrupt or gradual shift, due to non-climatic changes (e.g., the station relocation) or actual climate change of a region. This study presented a step-by-step methodology for detecting the climatic and non-climatic changes in annual precipitation ([Formula: see text]) and maximum ([Formula: see text]) and minimum ([Formula: see text]) air temperature data related to 37 weather stations across Iran, using the statistical methods and stations metadata. All data cover the common period of 1961-2014. Abrupt changes in climate data were detected using the non-parametric Pettitt test and the piecewise linear regression model, the gradual changes using the non-parametric Mann-Kendall (MK) test, and the magnitude of trends using the Sen's slope estimator. In addition, a two-sample t-test was used to consider whether means of the climate data have been significantly changed in the presence of change points recorded in stations metadata. Results indicated overall increasing trends in [Formula: see text] and[Formula: see text], with more increasing rate for [Formula: see text]. In case of precipitation, most stations indicated non-significant decreasing/increasing trends while six of them showed significant decreasing trends. The detected breaking points, mainly in [Formula: see text] were concurrent with the years of change in the original locations of 6 out of 37 stations. It was specified that the unreliable stations' data intensified the trend orientation and magnitude of climatic variables compared to the reliable ones. In addition, increasing rates in [Formula: see text] and [Formula: see text] (decreasing rate in[Formula: see text]) for the stations located in the urban areas were larger (smaller) than those in the non-urban areas. This research revealed the necessity of metadata for accurate interpretation of results obtained from the statistical methods. The study suggests to the Iranian climate researchers to employ with caution a homogeneous length of series rather than total inhomogeneous length of series.

Keywords: Change point; Iran; Metadata; Precipitation; Temperature; Trend.

MeSH terms

  • Climate Change
  • Environmental Monitoring*
  • Iran
  • Metadata*
  • Temperature