Document Type : Research Article
Authors
Department of Geography, Faculty of Literature and Humanities, Urmia University, Urmia, Iran
Abstract
Drought is one of the most important climatic phenomena affecting water resources and environmental conditions in the Lake Urmia Basin, making the assessment of its spatiotemporal variability particularly important. This study aimed to investigate the spatiotemporal variations in drought using the Standardized Precipitation Evapotranspiration Index (SPEI) and the Crossing Empirical Trend Analysis (CETA) method over the period 1987–2024. Precipitation and temperature data from eight selected stations across the basin were used. The results showed substantial temporal fluctuations in SPEI, with a non-uniform spatial pattern of variability across the basin. CETA analysis indicated that the upper slope (Us) and pivot slope (Ps) were decreasing at most stations, whereas the behavior of the lower slope (Ls) varied among stations. At Mahabad, Saqez, Takab, and Urmia, us showed a decreasing trend while Ls showed an increasing trend, indicating opposing changes in the two tails of the SPEI distribution. In contrast, both slopes showed decreasing trends at Maragheh, Sahand, and Sarab. Sen’s slope (Ss) was also decreasing at most stations, while differences in its magnitude indicated spatial heterogeneity in the rate of temporal changes in SPEI. Overall, the findings indicate a predominant decreasing tendency in SPEI across most parts of the basin, along with spatial differences in the behavior of different segments of the index distribution
Introduction
Drought is one of the most important climatic phenomena affecting water resources, agricultural systems, and environmental conditions, particularly in arid and semi-arid regions. Its temporal evolution and spatial distribution are influenced by variations in precipitation and atmospheric water demand, making the identification of changes in drought conditions important for sustainable water resources management. The Lake Urmia Basin in northwestern Iran has experienced substantial fluctuations in moisture conditions during recent decades, and changes in drought characteristics have important implications for the management of water resources in the basin. Previous studies in the Lake Urmia Basin have mainly focused on the temporal behavior of drought indices and their overall trends. However, an overall trend may not fully represent changes occurring across different portions of the distribution of a climatic index. In particular, the upper and lower parts of a drought-index distribution may exhibit different temporal behaviors. The Crossing Empirical Trend Analysis (CETA), introduced by Şen (2022), provides an approach for examining trend behavior at different portions of a time-series distribution through the upper slope (Us), lower slope (Ls), and pivot slope (Ps). In addition, Sen’s slope (Ss) can be used to characterize the overall direction and magnitude of temporal change. Therefore, this study aimed to investigate the spatiotemporal variations in drought in the Lake Urmia Basin using the 12-month Standardized Precipitation Evapotranspiration Index (SPEI-12) and the CETA method during 1987–2024.
Material and Methods
The Lake Urmia Basin, with an area of approximately 51,876 km², is located in northwestern Iran. Daily precipitation and air temperature data from eight synoptic stations, including Urmia, Tabriz, Sahand, Sarab, Maragheh, Mahabad, Saqez, and Takab, were used for the period 1987–2024. The selected stations represent different geographical and climatic conditions within the basin. Missing observations in the original meteorological records were reconstructed and completed before the analysis. The SPEI was calculated at a 12-month timescale based on the climatic water balance, defined as the difference between precipitation and potential evapotranspiration. Potential evapotranspiration was estimated using the Hargreaves–Samani method based on the available daily temperature data. The resulting daily potential evapotranspiration values were aggregated to monthly values and used together with monthly precipitation to calculate SPEI-12. The temporal behavior of SPEI was subsequently examined using the Crossing Empirical Trend Analysis (CETA). In this method, the time series is divided into two temporal sections and the crossing behavior of the series is used to derive trend lines representing different portions of the distribution. The upper slope (Us), lower slope (Ls), and pivot slope (Ps) were calculated to characterize changes in the upper, lower, and central portions of the SPEI distribution, respectively. Sen’s slope (Ss) was also calculated to provide an overall measure of the direction and magnitude of temporal change. Because CETA does not provide a statistical significance test for the estimated slopes in the present implementation, the direction of change was determined according to the sign of each slope. Positive values were therefore interpreted as increasing and negative values as decreasing. The results were summarized in tables and spatial distribution maps to identify differences among the stations.
Results and Discussion
The temporal variations in SPEI-12 showed considerable fluctuations between dry and wet conditions across the eight stations, with noticeable spatial differences in the magnitude and temporal behavior of these variations. The CETA results indicated that the upper slope (Us) was decreasing at all stations, with values ranging from −0.00640 at Maragheh to −0.00013 at Tabriz, indicating differences in the rate of change across the basin. The pivot slope (Ps) was also decreasing at seven stations, whereas Tabriz showed a positive value. In contrast, the lower slope (Ls) exhibited a heterogeneous spatial pattern. Ls was positive at Mahabad, Saqez, Tabriz, Takab, and Urmia, and negative at Maragheh, Sahand, and Sarab. The highest positive Ls occurred at Urmia (0.010571), highlighting the marked spatial contrast in the lower portion of the SPEI distribution. Accordingly, Us and Ls showed opposite directions at Mahabad, Saqez, Tabriz, Tekab, and Urmia, while both were decreasing at Maragheh, Sahand, and Sarab. Sen’s slope (Ss) was negative at seven stations, with the strongest overall decrease occurring at Maragheh (−0.00320), whereas Tabriz showed a positive slope (0.000863). The differences between Ss and the CETA parameters indicate that the overall trend did not always represent the same behavior across different portions of the SPEI distribution. The Pivot Point and pivot trend line further helped distinguish the behavior of the central portion from the upper and lower segments. Overall, the spatial distribution maps confirmed considerable heterogeneity in drought trends across the basin. These findings are consistent with previous studies reporting spatial and temporal heterogeneity in drought conditions in Iran and with applications of CETA to climatic time series. The results highlight the value of considering different portions of the SPEI distribution alongside the overall trend when examining drought variability.
Conclusion
This study investigated the spatiotemporal variations in drought across the Lake Urmia Basin using SPEI-12 and the Crossing Empirical Trend Analysis (CETA) method during 1987–2024. The results demonstrated substantial spatial and temporal heterogeneity in SPEI behavior across the eight selected stations. The upper slope (Us) was decreasing at all stations, while the lower slope (Ls) showed a heterogeneous spatial pattern, with increasing values at Mahabad, Saqez, Tabriz, Takab, and Urmia and decreasing values at Maragheh, Sahand, and Sarab. This resulted in opposing Us and Ls directions at five stations, whereas both slopes were decreasing at the remaining three stations. The pivot slope (Ps) was decreasing at seven stations and increasing at Tabriz. Sen’s slope was also decreasing at seven stations, with the largest overall decrease occurring at Maragheh, while Tabriz showed a positive overall slope. These results indicate that the predominant decreasing direction of SPEI across the basin was accompanied by substantial spatial differences in the magnitude and distributional behavior of the changes. The findings also demonstrate that the overall trend of SPEI does not necessarily describe the behavior of all portions of its distribution. Considering Us, Ls, Ps, and Ss together therefore provides a more detailed description of the spatial and temporal characteristics of drought conditions. The identified spatial differences can contribute to drought monitoring, the identification of areas with greater changes, and improved water resources management in the Lake Urmia Basin.
Keywords
Subjects
