Complexity of Forces Driving Trend of Reference Evapotranspiration and Signals of Climate Change

TLDR

Understanding trends in reference evapotranspiration (ETo) and its meteorological drivers is essential for hydrology, vegetation restoration, and agriculture, yet prior studies are limited by short time series, narrow variable scope, regional similarity, serial correlation bias, and lack of extremum analysis. This study aims to analyze 50‑year trends of ETo and 12 meteorological variables across 18 Iranian sites, remove serial correlation effects, identify key drivers, and assess extremum coincidences. Using a trend‑free pre‑whitening approach on 50‑year annual data from 18 contrasting climate sites, the authors identified the most influential meteorological variables and examined the alignment of their extreme values with ETo extremes. The analysis revealed significant 95% confidence trends in ETo and key meteorological variables at over half of the sites, with wind speed driving most ETo trends and its extremes coinciding with ETo extremes in 83.3% of sites, underscoring climate change impacts and the need to adjust cropping patterns in Iran.

Abstract

Understanding the trends of reference evapotranspiration (ETo) and its influential meteorological variables due to climate change is required for studying the hydrological cycle, vegetation restoration, and regional agricultural production. Although several studies have evaluated these trends, they suffer from a number of drawbacks: (1) they used data series of less than 50 years; (2) they evaluated the individual impact of a few climatic variables on ETo, and thus could not represent the interactive effects of all forces driving trends of ETo; (3) they mostly studied trends of ETo and meteorological variables in similar climate regions; (4) they often did not eliminate the impact of serial correlations on the trends of ETo and meteorological variables; and finally (5) they did not study the extremum values of meteorological variables and ETo. This study overcame the abovementioned shortcomings by (1) analyzing the 50-year (1961–2010) annual trends of ETo and 12 meteorological variables from 18 study sites in contrasting climate types in Iran, (2) removing the effect of serial correlations on the trends analysis via the trend-free pre-whitening approach, (3) determining the most important meteorological variables that control the variations of ETo, and (4) evaluating the coincidence of annual extremum values of meteorological variables and ETo. The results showed that ETo and several meteorological variables (namely wind speed, vapor pressure deficit, cloudy days, minimum relative humidity, and mean, maximum and minimum air temperature) had significant trends at the confidence level of 95% in more than 50% of the study sites. These significant trends were indicative of climate change in many regions of Iran. It was also found that the wind speed (WS) had the most significant influence on the trend of ETo in most of the study sites, especially in the years with extremum values of ETo. In 83.3% of the study sites (i.e., all arid, Mediterranean and humid regions and 66.7% of semiarid regions), both ETo and WS reached their extremum values in the same year. The significant changes in ETo due to WS and other meteorological variables have made it necessary to optimize cropping patterns in Iran.