Influence of the accuracy of reference crop evapotranspiration on drought monitoring using standardized precipitation evapotranspiration index in mainland China

Abstract

Reference crop evapotranspiration (ET0) estimation accuracy strongly affects the accuracy of drought monitoring using the standardized precipitation evapotranspiration index (SPEI). The FAO56 Penman–Monteith (the reference method here), Irmak, Berti, Priestley–Taylor, and Valiantzas methods were used to compute ET0 and subsequently SPEI at 1‐, 3‐, and 12‐month timescales during 1961–2016 at 763 weather stations in China. The spatiotemporal variations in the climatic variables and ET0, D (=P − ET0) and SPEI with five ET0 estimators (hereafter SPEIPM, SPEIIRA, SPEIPT, SPEIMHS, and SPEIVal) were investigated. The cumulative frequency and periods (using the Morlet wavelet analysis) of the annual SPEIs were analyzed. The results showed that, corresponding to the P, ET0, and D, drought severity was characterized by large seasonal and regional differences using SPEIPM, SPEIIRA, SPEIPT, SPEIMHS, or SPEIVal, especially in northwestern China, the Inner Mongolia region, and the Qinghai–Tibetan Plateau (P < 500 mm/yr). Additionally, the effects of the different ET0 equations on the minimum SPEI and frequency values were large in these drier subregions but generally small in other regions. Meanwhile, an increased ET0 prediction accuracy did not automatically lead to an improved SPEI accuracy in regions P > 500 mm/yr. With respect to the averaged values for Subregion 1 (P < 134 mm/yr), although the vibration intensities and spectra were visually similar for different SPEIs, the periodical signals differed. Regional climatic conditions have a significant impact on the SPEI estimates in regions characterized by P (<500 mm/yr); we therefore suggest that the ET0 equations for drought analysis with SPEI should be carefully selected.

Spatiotemporal variability of standardized precipitation evapotranspiration index in mainland China over 1961–2016

“…The Penman-Monteith equation (Allen et al, 1998) which has been proven suitable in many regions or locations (Beguería et al, 2013;Li et al, 2017b) was applied here for estimating ET 0 , which may have increased the accuracy of SPEI estimation. In addition, the three-parameter loglogistic probability distribution, which was proven applicable at 763 sites of China by Yao et al (2019c) and, was used for computing SPEI.…”

Future changes of drought characteristics in Coupled Model Intercomparison Project phase 6 Shared Socioeconomic Pathway scenarios over Central Asia

“…It is also claimed that the Penman–Monteith‐based drought index may provide better estimates of the global drought trend (Sheffield et al ., 2012; Trenberth et al ., 2014). Several studies have confirmed Penman–Monteith's superiority in drought monitoring over temperature‐only based PET methods (e.g., Thorthwaite) (Chen and Sun, 2015; Yao et al ., 2020). Therefore, the Penman–Monteith method is used to calculate PET based on the “penman” function in R package “SPEI” (Begueria and Vicente‐Serrano, 2013) in this study.…”

Drought Events in Western Part of Timor Island Indonesia

“…These results indicated that SPEI detected a higher frequency in other classes compared to SPI. The higher drought proportion in SPEI implied the lower standardized index value, which could be due to the influence of the evapotranspiration variable included (Polong et al, 2019;Yao et al, 2020). Furthermore, SPEI was also indirectly influenced by the air temperature variable (Fung et al, 2020;Qutbudin et al, 2019;Sein et al, 2021), since it was used as input data to calculate evapotranspiration.…”