Performance evaluation of different potential evapotranspiration models and application of optimal models for drought monitoring using the RDI index across different climate regimes of Iran

Introduction

Drought is a climatic anomaly that results from long-term disruptions in components of the water balance (Wang et al., 2021; Portner et al., 2022; Zhang et al., 2023; Kartal, 2024; Tareke, 2025). This phenomenon has both direct and indirect adverse impacts, with water resources being the most significantly affected (Balooei et al., 2024). Water scarcity and its associated challenges are recognized as among the most critical and urgent global crises (Zarei et al., 2019a). Although drought cannot be prevented, understanding its nature and characteristics enhances the potential for partial prediction and, through preparedness and planning, helps reduce—and, where possible, control—its detrimental effects (Rezaei et al., 2024). Therefore, greater attention must be paid to drought and to identifying the key factors influencing it across different regions, particularly in vulnerable countries such as Iran, which face growing water scarcity. The increasing need to understand drought and its consequences has motivated extensive global research aimed at developing various drought indices. Among these, the Reconnaissance Drought Index (RDI), introduced by Tsakiris et al. (2007), is one of the most notable.



Materials and Methods

In this study, daily observational data from nine synoptic stations covering a 30-year period (1991–2020) were obtained from the Iran Meteorological Organization to estimate potential evapotranspiration and the RDI index. These stations were selected so that each represents one of Iran’s major climatic groups. In this research, the performance of six temperature-based models and three radiation-based models for estimating potential evapotranspiration was evaluated. The primary goal of the analysis is to identify which of these simplified approaches provides results most consistent with the FAO Penman–Monteith (FAO-56 PM) model, which is widely recognized as the standard reference method (Allen et al., 1998).



Results and Discussion

Evaluation of Daily Potential Evapotranspiration Models

The results indicated that the performance of evapotranspiration models is strongly influenced by the climatic conditions of each region. For example, the Blaney–Criddle model performed best in certain climates, while the same model showed lower accuracy in others. This finding is consistent with previous studies, including Eghtedarnezhad et al. (2016), which emphasized the role of climatic factors in drought monitoring. Moreover, the variability in model performance across different regions further underscores the need to evaluate and select models that are appropriate for the specific climatic conditions of each area. This observation also aligns with the findings of Lehner et al. (2020) and Beobide-Arsuaga et al. (2021), who stressed the importance of model adaptability to local conditions. Overall, it can be concluded that choosing the appropriate model for estimating evapotranspiration is a crucial step in drought studies and water resource management, and must be carried out with careful consideration of each region’s climatic characteristics.

Evaluation of Monthly Potential Evapotranspiration Models

Overall, the results demonstrate that, similar to the daily scale, temperature-based models do not exhibit uniform behavior across different climates at the monthly scale. A model may perform exceptionally well in one climate while ranking among the weakest in another. These differences highlight the necessity of considering climate type, geographical characteristics, the study period, and careful model selection in climatological research (Latrech et al., 2024).

Assessment of Drought Using the 6- and 12-Month RDI Indices for Optimal Models at Selected Stations

The overall findings of this study indicate that at the 6-month timescale, the frequency of drought and wet periods is higher, whereas at the 12-month timescale, their frequency decreases but their persistence increases. This result is in agreement with the study by Ahrari and Raja (2025), who examined meteorological, agricultural, and hydrological drought indices in the Mahabad plain. Furthermore, based on the results of the present study, the 12-month RDI was identified as a more suitable timescale for monitoring drought and wet periods, which is consistent with the findings of Nouri and Homaee (2020), Torabinejad et al. (2023), and Rezaei et al. (2024).

Conclusion

Frequent drought events and the significant damages they cause in various sectors, including agriculture, the environment, socio-economic, and other areas, have made this phenomenon one of the fundamental challenges in different regions of the world. The present study aimed to evaluate the performance of different potential evapotranspiration models and their influence on the RDI drought index across diverse climatic zones, using 30 years (1991–2020) of data from nine synoptic stations representing nine distinct climate types in Iran (BSh, BSk, BWh, BWk, Cfa, Csa, Csb, Dsa, Dsb). The evaluation of nine evapotranspiration models showed that their performance is strongly affected by the climatic characteristics of each region. For example, the Droogers–Allen model performed best in cold semi-arid (BSk), Mediterranean temperate with warm summers (Csb), and cold climates with dry, warm summers (Dsb), whereas the same model exhibited poor performance in hot semi-arid climates (BSh) and several other regions. In the drought analysis section, comparison of the 6- and 12-month RDI indices revealed that although both timescales confirm the occurrence of frequent droughts during most of the study period, the 12-month RDI was identified as the more suitable scale for drought monitoring and analysis in Iran. This is due to its ability to filter out short-term fluctuations and better reflect the persistence of drought periods.