Date Published: February 28, 2019
Publisher: Public Library of Science
Author(s): Roohollah Noori, Fuqiang Tian, Ronny Berndtsson, Mahmud Reza Abbasi, Mohammadreza Vesali Naseh, Anahita Modabberi, Ali Soltani, Bjørn Kløve, João Miguel Dias.
Climate change’s effect on sea surface temperature (SST) at the regional scale vary due to driving forces that include potential changes in ocean circulation and internal climate variability, ice cover, thermal stability, and ocean mixing layer depth. For a better understanding of future effects, it is important to analyze historical changes in SST at regional scales and test prediction techniques. In this study, the variation in SST across the Persian Gulf and Gulf of Oman (PG&GO) during the past four decades was analyzed and predicted to the end of 21st century using a proper orthogonal decomposition (POD) model. As input, daily optimum interpolation SST anomaly (DOISSTA) data, available from the National Oceanic and Atmospheric Administration of the United States, were used. Descriptive analyses and POD results demonstrated a gradually increasing trend in DOISSTA in the PG&GO over the past four decades. The spatial distribution of DOISSTA indicated: (1) that shallow parts of the Persian Gulf have experienced minimum and maximum values of DOISSTA and (2) high variability in DOISSTA in shallow parts of the Persian Gulf, including some parts of southern and northwestern coasts. Prediction of future SST using the POD model revealed the highest warming during summer in the entire PG&GO by 2100 and the lowest warming during fall and winter in the Persian Gulf and Gulf of Oman, respectively. The model indicated that monthly SST in the Persian Gulf may increase by up to 4.3 °C in August by the turn of the century. Similarly, mean annual changes in SST across the PG&GO may increase by about 2.2 °C by 2100.
Sea surface temperature (SST) variations under climate change influence species in the marine environment and may thus threaten sensitive ocean corals, alter the intensity and frequency of blooms, reduce the nutrient flux from the deep to surface waters, raise sea level, change the global food chain, and create health-related problems for humanity by providing a more suitable environment for pathogenic microbes [1–6]. The variations in SST can also have a significant impact on climate components [7–10]. Several studies have demonstrated that SST has increased at global scale during the 20th century [11–13], while regional studies in some parts of the world have found that shallow waters such as gulfs may display a larger variation in SST increase compared with deep water areas . This may be a result of the relatively lower water depth in shallow water bodies [15–17]. Gulfs are mainly affected by their surrounding land mass and therefore the SST in such water bodies is also affected by air temperature, which consequently leads to more variability. For example, the decadal rate of SST increase in Narragansett Bay, USA, which is about 1.1 °C, is four times greater than that of the main ocean .
The Persian Gulf and Gulf of Oman, a climate change hotspot , borders Iran in the north, Iraq and Pakistan in the northwest and northeast, respectively, and Oman, United Arab Emirates, Qatar, Kuwait, Bahrain, and Saudi Arabia in the south (Fig 1). The water mass is connected to the Arabian Sea and ocean water in the east. The climate of surrounding land is dry and subtropical. Air temperature exceeds 50 °C in summer and evaporation rate exceeds rainfall in PG&GO. Mean annual rainfall along the southern and northern coasts is less than 50 and 200 mm, respectively. Freshwater discharge into PG&GO is mainly provided by the rivers of Iran, among which Arvand Rud provides the largest share. On the south coast, small amounts of freshwater flow into PG&GO. Higher salinity concentration in PG&GO than in the Indian Ocean, especially in western parts including the Persian Gulf, is the main factor influencing water exchange between PG&GO and the ocean [37,38].
Previous studies have investigated SST variations using monthly data [43,44], while Shaltout and Omstedt , Hobday et al. , and Noori et al.  used daily SST data to study marine extreme temperatures. Generally speaking, the variability is higher for daily SST data than for data recorded at monthly time scales, reaching about 1.5 times higher in some parts of the ocean . As monthly SST datasets are averaged from daily data, they are smoothed and do not properly show extreme high and lows of SST. Therefore, in this study we used DOISSTA data to reveal the high and low extremes of SST across PG&GO. The DOISTA data available from the NOAA website have a spatial resolution of 0.25 degree on a daily basis. The data comprise multiple sensor observations (satellites, ships, and marine buoys). More details on the DOISSTA database are presented in Reynolds et al.  and Reynolds .
This study evaluated variations in DOISSTA across PG&GO during the past 34 years using descriptive statistics and predicted SST by 2100 using a POD model. Investigations of DOISSTA based on statistical indices revealed an increasing trend in mean, minimum, and maximum DOISSTA of about 1 °C during the 34-year study period, likely due to the impacts of climate change. In the study period, the Persian Gulf experienced smaller and larger DOISSTA than the Gulf of Oman. The spatial distribution of the first mode calculated by POD revealed maximum variation in DOISSTA in the Persian Gulf, especially along its southern coasts between Bahrain and Qatar coastlines and in the northernmost area, where the river Arvand Rud discharges into the gulf.