عنوان مقاله [English]
Introduction: Evapotranspiration is one of the key elements of hydrological cycle. This parameter plays a crucial role in different water related studies such as agricultural water management, environmental energy budget, water balance of watersheds, water reservoirs and water conveyance structures (such as channels, dams, barriers and so on). Increasing greenhouse gases has led to increased atmosphere temperature. Such changes in air temperature and other atmospheric parameters caused some natural hazards in many regions. One of the important parameter impacted by climate change is potential evapotranspiration. Different studies conducted in the recent decade to detect the monotonic trends and abrupt changes in meteorological parameters. Most of them are on trend analysis of meteorological and hydrological parameters. In the recent years, monotonic trend analysis of reference crop evapotranspiration (ET0) has interested many investigators around the globe. Many investigators attempted to find the possible reasons of trends in ET0. In many cases, this is accomplished by sensitivity analysis of ET0 to different meteorological parameters. Other investigators attempted to model ET0 using the hydrologic time series modeling. Detection of sudden change point in different time series including ET0 is very important in changing climate. However, in spite of tremendous studies on monotonic trend analysis, it seems that no serious work has been conducted to detect abrupt changes in ET0 in Iran, especially in west and northwest of Iran. This region has fertile soils and produce an important portion of cereal yields of Iran, thus providing water to agricultural section is crucial under climate change. Therefore, the main objectives of this study were i) estimation of ET0 values in the selected stations in west and northwest of Iran using the FAO-Penman Monteith method, and ii) detection of significant change points in ET0 time series using the nonparametric Pettit test.
Materials and Methods: The 32 synoptic stations were selected in this area for analysis. Data needed for this study were gathered from IRIMO. Meteorological parameters were daily records of maximum air temperature, minimum air temperature, sunshine hour duration, wind speed, and relative humidity. The ET0 values were estimated using FAO-56 Penman-Monteith model. In order to detect the significant change point the non-parametric, Pettitt test was used. Both monthly and annual time scales were used in analysis. The null hypothesis of test is there is no sudden change point in the time series. We calculated the p-values for time series under test and compared it with significance level (5%). If the calculated p-value was less than the significance level (0.05), then the null hypothesis is rejected, and the alternate hypothesis (i.e. there is a significant sudden change point in the time series) will be accepted.
Results and Discussion: The results showed that around 60% of the monthly time series had significant sudden change points. For instance, Urmia showed significant abrupt changes in ET0 for all months. Specifically, more than 86 and 78 % of the stations experienced sudden change in ET0 in March and August, respectively. The strongest abrupt change observed at Maragheh, in which the difference in monthly ET0 before and after the change point date reached to about 45 mm. It is worth to mention that all detected sudden changes had upward direction. In annual time scale, more than 80 % of the stations showed significant abrupt changes in ET0. Among all stations, Sararoud- Kermanshah showed a large difference in mean annual ET0 for the subseries of before and after the change point date which was approximately 235 mm. In annual scale, all sites (except Sahand and Parsabad) experienced upward ET0 abrupt changes. In order to inspect the reason this change, we plotted different meteorological parameters time series. The results indicated that the wind speed showed negative trends (except for two stations) leading to ET0 increase. Furthermore, it was found that almost all stations exhibited increasing trends in air temperature. These changes caused an increase in ET0. The most prominent abrupt change date in ET0 time series was found for the years from 1995 to 1998. For example, in February, April, May, and June, monthly ET0 time series suddenly increased in 1998, which were statistically significant (p < 0.05). Following the year of 1998, some other monthly ET0 series showed abrupt change point in 1995 (p < 0.05).
Conclusions: The sudden change in ET0 was confirmed in west and northwest of Iran. According to the results, ET0 time series (in monthly or annual time scales) exhibited upward sudden changes. Such changes in ET0 time series ring the alarms and decision makers should be, therefore, cautious in management of water resources.