Application of Mann-Kendall trend test and Normalized Difference Vegetation Index (NDVI) in hydroclimatological change detection – A Case Study of Urmia Lake watershed, Iran

Urmia Lake watershed, which is located at the northwest of Iran has gone through crucial hydroclimatological changes that resulted in Urmia Lake Desiccation. Long term average temperature and precipitation trends, precipitation pattern transition and changes in precipitation-snow timing are some of the hydroclimatological changes that have occurred in this watershed. Aforementioned changes are provoked by anthropogenic activities such as land cover changes, booming urbanization, unauthorized exploitations of Urmia Lake and inefficient crop management, followed by natural factors that could be caused by climate change.

In this study we aim to address contributing hydroclimatological factors and vegetation changes that resulted in Urmia Lake desiccation. In order to examine the vegetation changes in this watershed, we analyzed remote sensing data. In this regard, normalized difference vegetation index (NDVI)-based models for two sub-basins in East Azerbaijan province and West Azerbaijan -located at two sides of Urmia Lake watershed- are developed by an ensemble of satellite data from 1990 to 2019. Data of Landsat 5 TM satellite from 1990 to 2013 and Landsat 8 OLI/TIRS satellite from 2013 to 2019, are collected and analyzed to consider vegetation changes. Hydrological data for precipitation, temperature and Urmia Lake water level elevation are also considered for analyzing hydroclimatic impacts. The adequacy of NDVI-based models and long term hydrological time series are checked by Mann-Kendall trend test.

The evaluation of NDVI-based models shows an increasing trend in vegetation. In comparison, the studied sub-basin in West Azerbaijan province has a higher level of increasing trend than the sub-basin in East Azerbaijan province. The evaluation of precipitation time series shows a decreasing trend and temperature data exhibit an increasing trend. The trend pattern changes validates the hypothesis that increasing trend of vegetation in Urmia Lake watershed is in proportion to the escalating agricultural activities. Furthermore, the escalation of land use is higher in West Azerbaijan province where there exists more agricultural activities.