Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf

MPB-07248; No of Pages 19 Marine Pollution Bulletin xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf Abubaker Elhakeem a, Walid Elshorbagy b,⁎ a b

Environment Department, Dubai Municipality, Dubai, United Arab Emirates MWH Global-Middle East, Dubai, United Arab Emirates

a r t i c l e

i n f o

Article history: Received 24 March 2015 Received in revised form 10 October 2015 Accepted 16 October 2015 Available online xxxx Keywords: Hydrodynamic evaluation Climate change Coastal effluents Salinity Temperature Arabian Gulf

a b s t r a c t A comprehensive basin wide hydrodynamic evaluation has been carried out to assess the long term impacts of climate change and coastal effluents on the salinity and seawater temperature of the Arabian Gulf (AG) using Delft3D-Flow model. The long term impacts of climate change scenarios A2 and B1 of the IPCC-AR4 on the AG hydrodynamics were evaluated. Using the current capacity and production rates of coastal desalination, power, and refinery plants, two projection scenarios until the year 2080 with 30 year intervals were developed namely the realistic and the optimistic discharge scenarios. Simulations of the individual climate change scenarios ascertained overall increase of the AG salinity and temperature and decrease of precipitation. The changes varied spatially with different scenarios as per the depth, proximity to exchange with ocean water, flushing, vertical mixing, and flow restriction. The individual tested scenarios of coastal projected discharges showed significant effects but within 10–20 km from the outfalls. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction The Arabian Gulf (AG) is a semi-enclosed extension of the Indian Ocean located in the subtropical, hyper-arid region between 24 and 30° latitudes occupying an approximate surface area of 240,000 km2 and connecting directly with the deep Gulf of Oman through the Strait of Hormuz which is a narrow 60 km wide passage. The AG is bordered by 6 countries i.e. Kuwait, Saudi Arabia, Bahrain, Qatar, United Arab Emirates, and Iran. The widest section of the Gulf spans over 340 km between the coasts of the United Arab Emirates (UAE) and Iran, and the bathymetry is shallow with an average depth of 36 m asymmetric along its axis with a deeper zone close to the Iranian coast and a broad, shallower shelf off the UAE coast. The AG forms an inverse estuary that experiences salinities higher than the adjacent Indian Ocean. The circulation of the AG is driven principally by the tides (Elshorbagy et al., 2008; Elshorbagy et al., 2006; Azam et al., 2006; Reynolds, 1993). The tides in the AG are complex standing waves and the dominant pattern varies from primarily semi-diurnal to diurnal. The tidal range has values greater than 1 m (Lehr, 1984). The residual circulation in the AG is density-dominated in the central and southern regions while it shows frictional-balanced, wind-dominated circulation in the

⁎ Corresponding author. E-mail address: [email protected] (W. Elshorbagy).

NW Region (Hunter, 1983). The open waters of the Arabian Gulf experience evaporation rates of 2 m/yr (Privett, 1959; Meshal and Hassan, 1986; Ahmad and Sultan, 1991) which considerably surpass the net freshwater input by precipitation (0.15 m/yr) (Johns et al., 2003) and the river discharges of 0.15–0.19 m/yr (Johns et al., 2003; Reynolds, 1993). The temperature and salinity of the AG was best described by the Mt. Mitchell expedition (Reynolds, 1993). The data showed that the southern shallow areas are more saline, with values up to 43 psu in winter. Reynolds (1993) showed that the surface salinity near the Straits of Hormuz was lower, with values of about 36–38 psu in both summer and winter due to the effect of the Indian Ocean surface water (IOSW). A stratified bottom layer with higher salinity and colder temperature was observed during both summer and winter. The lower summer salinity is attributed to the intensification of the Indian Ocean Surface Water (IOSW) inflow along the Iranian coastline in spring. In autumn and winter, together with a weakening of the IOSW inflow, the low-salinity surface signature partially disappears (Kämpf and Sadrinasab, 2006). In most Gulf waters, salinity greater than 39 psu occurs (Alessi et al., 1999) and salinities of over 70 psu are observed at low flushing major embayment in its central and southern parts (Kämpf and Sadrinasab, 2006). The rate of exchange between the AG and the Gulf of Oman via the Strait of Hormuz that determines the residence time for the AG basin has been studied by a number of scientists. Older studies showed that the residence time falls in the range of 2–5 years (Hughes and Hunter 1979; Hunter 1983).

http://dx.doi.org/10.1016/j.marpolbul.2015.10.032 0025-326X/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

2

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx Table 2 AOGCM in order of their skill for air temperature.

Fig. 1. Multi-model averages and assessed ranges for surface warming (source: IPCC, 2007).

1.1. Climate variability in the AG The region is expected to be under higher stresses due to the anticipated climate change impacts. The Intergovernmental Panel for Climate Change Fourth Assessment Report (IPCC, 2007) mentioned that by the middle of the 21st century, the Middle East region is expected to get warmer across all seasons. It also stated that “For the next two decades a warming of about 0.2 °C per decade is the average projected according to the special report on emission scenarios (SRES). Even if the concentrations of all greenhouse gases (GHG) and aerosols had been kept constant at year 2000 levels, a further warming of about 0.1 °C per decade would be expected.” Climate modeling results for the Middle East and Gulf region predicts an increase between 2.5 to 3.7 °C in summer, and 2.0 to 3.1 °C in winter by 2100 (Hemming et al., 2007). Higher temperatures will increase the vapor pressure which along with the changes in atmospheric circulation patterns will likely have a considerable effect on the intensity, frequency and spatial variability of precipitation. With less than 66% agreement of the climate model results on the sign of the change, the region will likely get drier, with significant rainfall declines in the wet season outweighing slight increases during the drier summer months (IPCC, 2007). The weather is also likely to become more unpredictable,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Model

Total rank

MIROC-HI MPIECH-5 CCSM—30 MRI-232A UKHADCM3 GISS—ER MIROCMED ECHO—G CCCMA-31 GFDLCM20 GFDLCM21 GISS—EH CNRM-CM3 UKHADGEM BCCRBCM2 CSIR0-30 FGOALS1G INMCM-30 IPSL_CM4 NCARPCM1

8 8 7 6 6 4 4 3 2 2 2 2 1 1 0 0 0 0 0 0

with the region experiencing an increase in extreme rainfall events (Tolba and Saab, 2009). Using long-term historical climate records for the Arabian Peninsula Nasrallah and Balling (1996) indicated a linear increase of temperature of 0.63 °C and an insignificant decrease in precipitation over the last 100 years. More recently, Al Sarmi and Washington (2011) examined trends in temperature and precipitation parameters for the Arabian Peninsula (AP) during the last 2 to 3 decades using recorded measurements from 21 stations and concluded a general warming pattern of the Arabian Peninsula mean annual temperature with the UAE (Dubai) showing an increasing rate of 0.81 °C per decade. In their study they showed that the mean annual precipitation changes were insignificant showing a decreasing trend. Most recently Elhakeem et al. (2015a,b) used statistical downscaling (SD) and proposed an improved systematic approach to select influential predictors to downscale the Hadley Model (HadCM3) projections using local observations at two stations representing the dominating bioclimatic zones in the UAE. The tested climate change scenarios revealed a range of increase of the annual mean maximum temperature of 2.79–3.80 °C and a range of reduction of annual precipitation between 16.80–37.00% by 2080 at the considered stations.

Table 1 Statistical analysis of Middle-East and global modeled and observed temperature. Model

BCCRBCM2 CCCMA-31 CCSM–30 CNRM-CM3 CSIR0-30 ECHO—G FGOALS1G GFDLCM20 GFDLCM21 GISS–EH GISS–ER INMCM-30 IPSL_CM4 MIROC-HI MIROCMED MPIECH-5 MRI-232A NCARPCM1 UKHADCM3 UKHADGEM

CORR (ME)

CORR (G)

RMSE (ME)

RMSE (G)

BIAS (ME)

BIAS (G)

°C

°C

°C

°C

°C

°C

0.954 0.957 0.973 0.961 0.941 0.929 0.931 0.963 0.938 0.929 0.956 0.906 0.949 0.976 0.944 0.977 0.938 0.919 0.98 0.934

0.988 0.99 0.995 0.99 0.991 0.99 0.973 0.989 0.992 0.983 0.988 0.987 0.989 0.994 0.991 0.996 0.995 0.99 0.994 0.992

4.233 3.794 1.943 4.36 4.541 2.199 3.285 4.273 3.244 3.033 1.637 5.221 3.563 1.222 2.109 1.218 2.01 7.007 2.159 5.533

3.274 3.011 1.396 2.68 2.649 2.029 4.393 3.12 2.299 2.71 2.296 3.019 2.782 1.665 2.198 1.473 1.889 2.977 2.051 2.9

−3.862 −3.426 −1.415 −3.961 −4.131 −0.715 −2.407 −3.974 −2.61 2.223 0.069 −4.638 −3.1 −0.057 0.159 −0.231 0.201 −6.641 −1.849 −5.159

−2.216 −1.805 −0.294 −1.756 −1.772 0.307 −1.994 −2.278 −1.47 0.62 −0.499 −1.969 −1.789 −0.536 −1.059 −0.257 −0.811 −2.138 −0.901 −2.109

CORR-RMSE (ME)

CORR-RMSE (G)

1.732 1.631 1.332 1.823 1.886 2.08 2.235 1.572 1.927 2.063 1.636 2.397 1.755 1.22 2.104 1.196 2 2.234 1.115 2

2.411 2.411 1.364 2.025 1.969 2.006 3.915 2.132 1.767 2.638 2.241 2.288 2.13 1.576 1.926 1.45 1.706 2.071 1.842 1.991

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

3

Fig. 2. Evaporation record at BIA, adjusted using a 0.7 pan coefficient.

The effects of the warm water spikes that occurred in 1996 and 1998 especially, are increasingly well documented (Sheppard and Loughland, 2002; Purkis and Riegl, 2005). Sheppard et al. (2010) provided a thorough review of the substantial changes that have taken place in marine habitats and resources of the Gulf over the past decade and concluded that major impacts come from numerous industrial, infrastructurebased, and residential and tourism development activities beside the witnessed impacts of climate change. 1.2. Coastal effluent impacts on seawater salinity and temperature: The phenomenal rapid development has dictated the need of vast fresh water and electricity supplies in response to the increasing demands. In the absence of other reliable alternatives, the AG is serving as a source and sink for mega seawater desalination and power plants. The 2010 report of the GCC water experts committee (GCC, 2010) has indicated that the planned desalination projects and those under construction are intended to add almost 6 million m3/day of fresh water to the current capacity. For example 90% of Kuwait's potable water is derived from desalinated seawater (Darwish et al., 2008), and this level of dependence is seen throughout the region. Currently the combined production is estimated as 25 million m3/day (GCC, 2010; Lattemann et al., 2013) representing about 44% of the total world fresh water capacity.UAE, Saudi Arabia, and Kuwait have the largest installed desalination capacity in the Gulf with a production of about 17 Mm3/day (Lattemann et al., 2013). Brine water rejects from desalination plants is commonly hot by 10 to 15 °C. These can cause elevations of 5 °C and 3 psu above ambient into waters which are already warm and highly saline (Linden et al., 1988). Moreover, 660 GWh of electricity is produced in the Gulf with an estimated 200 m3 of seawater used as cooling water for producing 1 MWh of electricity(DOE, 2006). Beside the conventional thermoelectric power plants, nuclear plants for electricity production have come into the picture in the United Arab Emirates where two units have been licensed for production by 2017 to secure the countries' 9% annual increase in electricity demands. Furthermore, the energy exploration, processing, storage and transporting facilities are currently handling a daily production of 25.5 million barrels and using the same water source for cooling and routine operations imposing additional stresses on the seawater quality. Crude oil refineries in the AG hold an estimated capacity of 4.2 Mbbl/day. The USGS report on “Water Requirements for Petroleum Refining Industry” (OTTS, 1963) estimated 345 gal/barrel of crude as cooling water needed by the refineries. Lattemann and Höpner(2008), and Areiqat and Mohamed (2005) examined the biological and ecological impacts on the Gulf's marine life

Fig. 3. Evaporation by season at BIA.

from thermal plume generated by the power and desalination plants. However, environmental impact studies are generally focused on local ecosystem effects and have difficulty projecting an assessment on the Gulf as a whole with a long term perspective (Kim and Jeong 2013). Recent studies in the Arabian Gulf (Purnama et al., 2005; Smith et al., 2007) modeled the change in salinity due to the desalination process considering a peak salinity of 40 ppt suggested an increase of salinity of the order of 0.06 ppt due to the Al Jubail desalination plant in Saudi Arabia. The observation was based on simplistic modeled assumptions. In addressing the issue of the salinity changes in the coastal waters of Kuwait due to large scale power and desalination activities Al-Dousari (2009) reported upon using systematic sampling campaigns at the uptake and outfall a difference of 4 ppt increase at the outfall compared to the uptake zone. Elshorbagy et al. (2013) reported in a study at Ruwais area in the western region of AbuDhabi emirate in the UAE a persistent increase over ambient of 0.2 and 0.9 °C of the seawater temperature in summer and winter seasons respectively near the outfall. The study indicated the prominent effect of industrial discharges in winter as surface temperature and salinity was significantly higher than in other surrounding locations. 2. Method and data Atmospheric Ocean Global Climate models (AOGCM) are the only available tool to scientifically project different future climate changes (Samuelsson, 2010). The use of hydrodynamic models provides the ability to simulate future water quality by using the climate model outputs (Perroud et al., 2009), but this possibility has not been widely used so far. Studies using the climate model outputs as input to the hydrodynamic models are almost non-existent nor is there any standardized methodology to incorporate the projected climate model anomalies in the used hydrodynamic models. This study aims to conduct the needed evaluation of climate change and coastal effluent impacts using Delft3DFlow by adopting a long term and a comprehensive basin wide approach. The modeling approach and data manipulation of climate change and coastal effluents were described and detailed in the technical note paper Elhakeem and Elshorbagy (2013). The paper described the approach adopted here for evaluating the long-term variability of seawater salinity and temperature in response to natural and anthropogenic stressors in the Arabian Gulf. The used hydrodynamic model was presented in a previous work (Elhakeem et al., 2015b). Using the Delft3D-Flow a 3-D prognostic baroclinic hydrodynamic model of the Arabian Gulf (AG) was developed with 10 km rectilinear grid arrangement and 5 sigma layers in the vertical. The model was forced with long-term time averaged climatological data over the computational domain and long-term salinity and temperature boundary conditions applied at its tidal open boundary. The k-Ɛ 3D-turbulence closure was applied with a background typical of horizontal viscosity and horizontal diffusivity coefficients for the AG specified of 15 m2/s (Azam et al., 2006; Al-Osairi et al., 2011; Kämpf and Sadrinasab, 2006). Heat flux plays a major role in the AG circulation. The “Ocean” model following Gill (1982) and Lane (1989) was used to force the heat flux at the air water interface. The model applies Chézy friction formulation

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

4

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

Table 3 Statistical analysis of Middle-East and global modeled and observed precipitation. Model

BCCRBCM2 CCCMA-31 CCSM–30 CNRM-CM3 CSIR0-30 ECHO—G FGOALS1G GFDLCM20 GFDLCM21 GISS–EH GISS–ER INMCM-30 IPSL_CM4 MIROC-HI MIROCMED MPIECH-5 MRI-232A NCARPCM1 UKHADCM3 UKHADGEM

CORR (ME)

CORR (G)

RMSE (ME)

RMSE (G)

BIAS (ME)

BIAS (G)

mm/day

mm/day

mm/day

mm/day

mm/day

mm/day

0.629 0.815 0.284 0.434 0.908 0.46 0.484 0.931 0.823 0.615 0.859 0.874 0.876 0.788 0.76 0.907 0.895 0.604 0.949 0.885

0.793 0.888 0.797 0.772 0.814 0.91 0.816 0.868 0.857 0.733 0.774 0.7 0.808 0.8 0.833 0.808 0.886 0.665 0.858 0.797

0.454 0.215 0.478 0.591 0.221 0.577 0.406 0.244 0.253 0.389 0.196 0.202 0.328 0.481 0.245 0.311 0.235 0.383 0.199 0.388

1.311 0.949 1.327 1.438 1.209 0.864 1.226 1.099 1.149 1.512 1.43 1.606 1.269 1.34 1.162 1.351 0.967 1.715 1.256 1.614

0.145 −0.033 0.103 0.388 −0.145 0.142 0.196 −0.129 −0.145 −0.262 −0.026 −0.064 −0.225 0.307 −0.017 −0.271 −0.132 −0.115 −0.09 −0.08

0.307 −0.01 0.16 0.54 −0.161 0.128 0.307 0.091 0.215 0.34 0.297 0.116 −0.09 0.281 0.035 0.247 −0.084 0.343 0.23 0.385

using a constant bottom roughness friction coefficient. The simulation results were thoroughly validated against measured tides from 5 stations and measured currents at 4 locations in the central and southern parts. Water salinity and temperature were validated in space and time using observations spanning over 73 years from 1923 to 1996 for the AG, the Strait of Hormuz and the Gulf of Oman. 2.1. Climate change data The impacts of the anticipated climate change scenarios produced by different Atmospheric Ocean General Circulation Model (AOGCM) are included in the model by incorporating the projected anomalies of air temperature, precipitation and sea level rise in the initial setup input files to Delft3D-Flow prepared and used in the calibration phase. Air temperature is introduced in the heat flux file, precipitation in the evaporation file and sea level rise in the water level boundary files. The A2 and B1 climate change scenarios of the special report on emission scenarios (Nakićenović and Swart, 2000) were used representing the mid-high and low projected scenarios (see Fig. 1) used by the IPCC; the climate change fourth assessment report (IPCC, 2007). A2 and B1 climate change projections for the AG are obtained at 2020, 2050, and Table 4 AOGCM in order of their skill (precipitation).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Model

Total rank

GFDLCM20 MRI-232A CCCMA-31 CSIR0-30 MIROCMED UKHADCM3 INMCM-30 ECHO—G GFDLCM21 GISS—ER IPSL_CM4 MPIECH-5 UKHADGEM CCSM—30 FGOALS1G BCCRBCM2 CNRM-CM3 GISS—EH MIROC-HI NCARPCM1

7 7 6 5 5 5 4 3 3 3 2 2 2 1 1 0 0 0 0 0

CORR RMSE (ME)

CORR RMSE (G)

0.43 0.212 0.467 0.446 0.166 0.559 0.356 0.207 0.207 0.288 0.194 0.192 0.238 0.371 0.244 0.151 0.194 0.365 0.177 0.38

1.275 0.949 1.317 1.333 1.198 0.854 1.187 1.095 1.128 1.473 1.399 1.601 1.266 1.311 1.162 1.328 0.963 1.68 1.235 1.568

2080 with reference to 1990 from MAGICC/SCENGEN (version 5.3.v2) by using a multi-model approach. Air temperature is included in the heat flux file, precipitation is included into the evaporation file and the sea level rise is specified as a water level anomaly time series at the water level boundary. MAGICC/SCENGEN is a coupled gas-cycle/ climate model. (MAGICC is a model for the Assessment of Greenhousegas Induced Climate Change) that drives a spatial climate-change SCENarioGENerator (SCENGEN). MAGICC has been one of the primary models used by IPCC since 1990 to produce projections of future globalmean temperature and sea level rise. AOGCM data base in SCENGEN includes twenty CMIP3/AR4 models re-gridded to 2.5 × 2.5° latitude/ longitude and the global observed data bases (at 2.5 by 2.5 resolution) with a common 20-year reference period, 1980–1999 (MAGICC/ SCENGEN user manual, 2008). For impact studies it is useful to consider, not just a single model, or a set of single models, but the average over a number of models as some AOGCMs may simulate the observed climate better than others in particular regions and, importantly for impact studies, they may produce different climate. This is an idea first introduced by Santer et al. (1990). The justifications for use of a multimodel average are two-fold. First, multi-model averages are less spatially noisy. Second, by many measures of skill, multi-model averages are often better than any individual model at simulating present-day climate (Wigley, 2008). 2.1.1. Arabian Gulf projected climate change anomalies SCENGEN tools produces four statistics for model validation performance, calculated by comparing observed and present-day model control-run or 20th-century run data for temperature, precipitation and pressure. A semi-quantitative skill score is recommended to rank models using the “VALDIN.out” statistical analysis file. Each model is counted once if it is in the top seven (top third approximately) for any statistic over the globe or over the selected study region (Middle East in our case). The maximum skill score is therefore 8, which would mean that the model was in the top seven for all four statistics over both regions. 2.1.1.1. Air temperature. In Table 1, the regional (Middle East) and global statistical results of the four measures for temperature are listed, and in Table 2 the models are ranked in the order of their skill scores. The average temperature skill analysis has shown the superiority of the Japanese model MIROC-HI and the German model MPIECH-5 both scoring 8 meaning that the model was in the top seven for all four statistics over both regions. Next, a subjective choice must be made as to which

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

5

Table 5 Arabian Gulf projected Climate change anomalies. A2

1990 2020 2050 2080 2100

B1

Air temperature (°C)

Precipitation (%)

SLR (cm)

Air temperature (°C)

Precipitation (%)

SLR (cm)

0.00 +0.78 +2.10 +3.94 +5.18

0.00 −5.72 −14.23 −25.08 −32.87

0.00 +7.93 +22.45 +45.81 +63.63

0.00 +0.78 +1.63 +2.36 +2.60

0.00 −5.34 −11.59 −16.58 −17.95

0.00 +9.04 +23.82 +38.86 +49.06

Table 6 Arabian Gulf air temperature and relative humidity record analysis. Station

Kuwait Dhahran Bahrain Doha AbuDhabi Dubai B-Abbas B-Lingeh Bushehr Abu Musa Kish Island Sirri Island

Record (Yr)

Missing %

Ta (C)

Rh (fraction)

Mean

SD

Data trend line

R

Mean

SD

Data trend line

R2

40 65 40 40 30 40 24 24 9 9 9 9

2 8 0.4 1.3 0.01 1.8 1.3 4.2 6.8 13.9 7.5 14.5

26.37 26.75 26.94 27.84 27.98 27.82 26.85 27.5 25.13 27.67 27.4 28.05

9.49 7.67 6.74 7.0 6.46 6.22 6.29 6.07 7.19 5.15 5.61 5.06

y = 0.0001× + 23.543 y = 3E-05× + 26.326 y = 0.0001× + 25.016 y = 0.0001× + 25.391 y = 0.0002× + 24.507 y = 0.0004× + 21.4 y = 0.0002× + 22.792 y = 0.0004× + 19.299 y = 0.0026× − 34.023 y = 0.0003× + 21.158 y = 0.0002× + 23.046 y = 0.0003× + 21.445

0.0021 0.001 0.0033 0.0079 0.0088 0.0137 0.002 0.0002 0.0004 0.0025 0.001 0.0025

0.47 0.59 0.67 0.67 0.6 0.65 0.75 0.64 0.64 0.66 0.66 0.65

0.24 0.19 0.13 0.18 0.14 0.17 0.17 0.14 0.1 0.11 0.11 0.1

y = −3E−05× + 0.9035 y = −1E−05× + 4.5044 y = −2E−05× + 0.9239 y = −1E−04× + 5.777 y = −2E−05× + 0.8936 y = −3E−05× + 1.1104 y = −5E−05× + 1.7328 y = −1E−05× + 0.8705 y = −2E−05× + 1.1016 y = −4E−05× + 1.471 y = −8E−06× + 0.839 y = 0.0002× + 1.0187

0.1948 0.0371 0.2795 0.3059 0.127 0.4199 0.5368 0.0266 0.034 0.0841 0.0044 0.0186

2

Table 7 Arabian Gulf wind speed and cloud cover record analysis. Station

Kuwait Dhahran Bahrain Doha AbuDhabi Dubai B-Abbas B-Lingeh Bushehr Abu Musa Kish Island Sirri Island

Record (yr)

Missing %

U10 (m/s) Mean

SD

Data trend line

R2

Mean

Cloud cover (fraction) SD

Data trend line

R2

40 65 40 40 30 40 24 24 9 9 9 9

2 8 0.4 1.3 0.01 1.8 1.3 4.2 6.8 13.9 7.5 14.5

4.09 4.32 4.68 4.15 3.69 3.53 2.84 3.49 3.14 3.93 3.72 4.77

2.21 1.68 2.0 1.87 1.11 1.19 1.25 1.80 1.73 2.1 1.74 2.08

y = −4E−07× + 4.0981 y = −5E−06× + 0.6505 y = −5E−05× + 5.4358 y = −2E−05× + 1.062 y = 2E−06× + 3.6454 y = 1E−05× + 3.3093 y = 0.0002× − 1.8469 y = 4E−05× + 2.6543 y = −0.0002× + 6.7187 y = −3E−06× + 3.998 y = −1E−05× + 4.0099 y = −2E−05× + 0.9985

9E−7 0.0038 0.0094 0.0489 4E−5 0.0021 0.2427 0.0038 0.0072 2E−6 4E−5 0.0051

0.22 0.14 0.19 0.15 0.13 0.15 0.18 0.14 0.24 0.14 0.12 0.19

0.07 0.04 0.05 0.04 0.03 0.04 0.05 0.07 0.11 0.05 0.07 0.11

y = −0.0041× + 8.3748 y = −0.0016× + 3.3673 y = −0.0012× + 2.5857 y = −0.0017× + 3.5137 y = −0.0008× + 1.807 y = −0.0014× + 3.0232 y = −0.0037× + 7.5585 y = −0.008× + 16.145 y = −0.0024× + 5.1081 y = −0.0153× + 30.853 y = −0.0025× + 5.0485 y = 0.0042× − 8.2221

0.5002 0.2833 0.0942 0.2473 0.0512 0.2056 0.2279 0.715 0.0255 0.6481 0.0104 0.0101

models to retain for temperature multi-model averaging. Based on the results in Table 2, the models ranking 8 to 1 are chosen to develop the projected A2 and B1 future air temperature. ME: Middle East; G: global; RMSE: root mean square error; CORR: correlation. 2.1.1.2. Precipitation. The regional and global statistical results of the four measures for precipitation are listed and in Table 3 the models are ranked in the order of their skill scores of producing the observed

Table 9 BIA air temperature and Dhahran seawater temperature. Date

23/03/1986 13/07/1986 29/09/1986 04/05/1987 07/10/1987 12/12/1987 08/03/1989

Ta (BIA)

Tw (Dhahran)

(°C)

(°C)

20.9 34.4 32.6 30.3 32.3 18.6 22.5

20.4 30.7 33.2 26.5 31.5 22.2 18.2

Table 8 Seasonal evaporation trend line analysis at BIA.

Winter Spring Summer Autumn

Trend line equation

R2

0.037× − 71.54 0.1283× − 250.95 0.0867× − 166.47 0.0335× − 63.146

0.3115 0.6636 0.386 0.1987

Table 10 BIA data used to calculate C coefficient for the AG. E (mm/month)

Ta (°C)

es(mm Hg)

Tw (°C)

RH (fraction)

es (mm Hg)

U10 (km/h)

128.4

26.94

24.87

25.76

0.67

17.86

16.85

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

6

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

precipitation for the Middle East (ME). The average precipitation skill analysis has shown the superiority of the USA model GFDLCM20 and the Japanese model MRI-232A both scoring 7 (Table 4). No model has

Table 11 Coastal desalination plants (m3/day) in the AG (Source: GCC-Desalination, 2010; GWI, 2007). Country Plant

Technology Capacity Units (m3/day)

Commissioning

Kuwait

MSF MSF MSF MSF MSF MSF MSF MSF MSF MSF RO MSF MSF RO RO MSF +

163,636 190,909 501,818 88,636 523,636 454,545 223,000 280,000 137,727 947,727 90,909 22,727 113,636 75,000 45,455 409,091

6 7 16 3 16 4 10 8 6 20 15 2 5 10 8 10 +

1971–1975 1978 1983 1982 1988–2001 2006–2007 1982 2000 1982 1983 2001 1986 1975–1985 1984 1989 2000

MED MED VC MSF MSF MSF MSF MSF MED RO RO MSF MSF MSF

31,818 145 318,182 150,000 136,364 181,818 272,727 9091 909 909 68,182 86,364 109,091

4 4 2 14 5 3 4 4 2 − − 4 3 6

2004 1985 1977–1983-1994 1997–1998 2008 2004 2006–2008 1997 1982 1983 1977 1979 1980

MSF

104,545

4

1985

MSF MED

104,545 31,818

3 2

1987 2000

MSF MSF MSF MSF MSF MED MSF MSF MSF MSF MSF RO MSF MED-RO MSF MSF MSF MED RO RO MSF +

286,364 145,455 318,182 104,545 231,818 240,909 113,636 136,364 309,091 272,727 459,091 172,727 290,909 454,545 150,000 72,727 104,545 50,000 13,636 4545 34,545

5 4 6 3 4 14 − − − − 6 2 5 − 5 3 3 −

−

2002 1989 1995 2000 2001 2002 1981 1990 1994 1999 2004 2004 2004 2000 1981 1996 2001 2000 1995 1990 −

6818 13,636

− −

− −

KSA

Bahrain

Qatar

UAE

Iran

Ashwaiba-S Doha-E Doha-W Ashwaikh Azour-S Alsubia Khobar-2 Khobar-3 Al-Jubail-1 Al-Jubail-2 Al-Jubail-RO Khafji Sitra RasJarjoor Aldoor Alhidd Alba Howar RasFunats-A RasFunats-B Ras Funats-B2 RasLafan Qatar Taqa Dukhan Abu Samra Shamal AbuDhabi-steam UmmAlNar-E (A) UmmAlNar-W (1–6) UmmAlNar-W (7–8) UmmAlNar-E (B) UmmAlNar-W (9–10) UmmAlNar-W (B) Al-Tawila-A1 Al-Tawila-B1 Al-Tawila-B2 Al-Tawila-A2 Al-Tawila-A2 JabalAli-D JabalAli-E JabalAli-G JabalAli-K Shuwaihat Quidfa-RO Quidfa-MSF Quidfa-Hybrd Layyah Mirfa Mirfa New Ajman Ajman Zawra Bandar Abbas Kishm Bandar Lingeh

MED MSF MSF +

Kish Isl. YaLavan Isl.

MED MED + RO MSF +

5000 6364

− −

− −

Asaluyeh Bushehr

MED MED + RO MSF +

18,182 44,545

− −

− −

MED

scored the maximum of 8, this is expected due to the complexity of precipitation modeling. The first 15 models are retained for precipitation multi-model averaging. Comparing the list of models for both parameters shows that the top models in temperature show relatively low importance in precipitation this comparison is important to show that the prediction skill of a model is variable and is based on the tested parameter and the study area under consideration which emphasizes the need for adopting a multimodel approach. The multi-model statistical selection process described in the user manual was conducted as specified and resulted in selecting a set of fourteen models for temperature and another set of fifteen models for precipitation for reproducing the observed data of temperature and precipitation. 2.1.2. Input files preparation methodology The regional climatic trend changes obtained from long term data records and the future global projected A2 and B1 scenario anomalies obtained from the AOGCM multi-model approach and summarized in Table 5 are used here to build the input files of the future scenarios to be considered in the present Arabian Gulf hydrodynamic model with reference to the year 1990. The simple global projections of air temperature and precipitation are used to develop rational estimates of long term surface evaporation input files. Projected A2 and B1 sea level rise scenarios are simply added to the tidal Input files at the open boundary. The following parts discuss the preparation methodology and present the data used. 2.1.2.1. Heat flux and wind file future projections. The heat flux file includes 4 columns namely: time step, relative humidity, air temperature and cloud cover. Analysis of long term climate records from 12 climate stations around the AG (Tables 6 and 7) showed negligible changes in relative humidity and wind speed and marginal reduction in cloud coverage. Based on these findings the air temperature due to climate change is considered the most significant parameter affecting the future heat flux and was adjusted in the heat flux files considering the temperature anomalies of the A2 and B1 climate change scenarios. Other parameters were extended to 2080 without change. 2.1.2.2. Sea level rise future projection. Sea level rise anomalies from the 1990 baseline year (Table 5)are simply specified at the open boundary at 2020, 2050, and 2080 by creating a separate time series file with the target sea level rise value using the Delft3D-flow water level–time series option with “Filbc0” key word and the file name added in the additional parameters section of the Master Definition File. 2.1.2.3. Future evaporation projection. In certain simulations, similar to the AG arid region, evaporation plays an important role in the salt mass balance. Using the “Fileva” keyword, the time-series for evaporation and precipitation (and its temperature) can be specified as a separate ASCII file (Deltares, 2011). Daily evaporation measurements using “class A” evaporation pan for the period 1983–2009 provided to the study by Bahrain International Airport (BIA) is used in the study. Data analysis showed an increasing trend of evaporation at a rate of Table 12 Projected freshwater demand increase rates in GCC countries. Reference

Description

GCC-Desalination, 2010 World Bank-Fichtner, 2011

10 year increase rate (2000–2010) GCC demand increase (2000–2050) (maximum in Kuwait) Realistic development scenario UAE (2010–2030) Water demand increase (1995–2025) Optimistic development scenario UAE (2010–2030)

UAE-MOEW, 2010 Dawoud, 2007 UAE-MOEW, 2010

% 5.4 3.9 4.7 6.3 13.2

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

7

Table 13 Projected increase percentage by technology until 2080 for the realistic-BAU development scenarios.

MSF MED RO Sum

(2007–2012) increase %

Annual increase %

Market share % of technology

Annual increase of technology considering realistic development (5%)

Increase % in 2020 to 2012

Increase % in 2050 to 2012

Increase % in 2080 to 2012

27 113 477

5.4 22.6 95.4 123.4

4.4 18.3 77.3 100

0.2 0.9 3.9 5

2 7 31 40

8 35 147 190

15 62 263 340

Table 14 Brine production by desalination process. Environmental requirement or impact Volume of saline feed water per m3 of fresh water Volume of brine effluent per m3 of fresh water

MSF

MED

4 3

3 2

RO 2 to 2.5 1 to1.5

The regression relationship is used to translate the future air temperature of the AOGCM's to water temperature under the basic assumption that the air to water variation relation will hold in the future and that the location of the measurements at the central portion of the AG could safely be considered as an average representative value. Tw ¼ 3:6981 þ 0:8188Ta

0.002 mm per decade. Fig. 2 shows a plot of the measured evaporation after being adjusted by applying a 0.7 pan coefficient. The seasonal trend analysis for the historical data represented in Table 8 and Fig. 3 showed an increasing trend across all seasons with different rates. The highest increase was in spring (March–April–May) with 0.0064 mm/yr and the least increase in Autumn (Sept.–Oct.– Nov.) with 0.0017 mm/yr. The highest evaporation was always in summer seconded by spring. The projected evaporation files were estimated using Meyer formulae and temperature anomalies. Empirical formula (Eq. (1)) was developed by Meyer (1915) based on Dalton's Law (1802) (Harrold et al., 1986) and is widely used to estimate the evaporation from lakes or reservoirs and is used to estimate the monthly projected evaporation rates. E ¼ Cðes −ed Þð1 þ 0:1 U10 Þ

ð1Þ

where E is monthly evaporation rate, C is an empirical coefficient equal to 15 for small shallow pools and 11 for large deep reservoirs, es is saturation vapor pressure in mm of mercury corresponding to monthly mean air temperature observed at nearby stations for small bodies of shallow water or corresponding to water temperature instead of air temperature for large bodies of deep water, ed is actual vapor pressure in mm of mercury in air based on monthly mean air temperature and relative humidity at nearby stations (ed = RH . es) small bodies of shallow water or based on information obtained about 10 m above the water surface for large bodies of deep water, and U10 is monthly mean wind velocity (km/h) at 10 m above ground level. The projected water temperature is needed to complete the projected evaporation calculations by Meyer's formula. Since the AOGCM's do not produce water temperature projections, a regression relationship was produced between BIA air temperature and water temperature using measurements of a three year (March 1986–March 1989) oceanographic study of the western Arabian Gulf (Dhahran) some 30 km west of BIA conducted by King Fahd University of Petroleum and Minerals/Research Institute (KFUPM/RI, 1990). Table 9 lists the data used for developing the correlation relationship. A regression relationship (Eq. (2)) is produced using a 95% confidence level with an R square value of 80.4% using the data in Table 3.

ð2Þ

Where: Ta and Tw are the air temperature and water temperature (°C). Fig. 4 compares the resulting calculated Tw using the regression equation and the recorded field measurements. The saturated vapor pressure equation (Eq. (3)) from Ward and Trimble (2004) is used to estimate es and ed ink Pa unit which is then converted to mm Hg unit.   16:78T−116:9 es ¼ exp T þ 237:3

ð3Þ

Using the BIA annual average evaporation rate, Ta, RH, and U10 in Table 10 and Tw calculated from the developed correlation relationship (Eq. (3)), the values are substituted in the Meyer (1915) equation in order to verify the “C” coefficient for the AG water basin. The manipulation of data produced a representative value of C = 6.82 for the Arabian Gulf water basin. Future evaporation files for the A2 and B1 climate change scenarios were prepared in three steps. First, the projected monthly average evaporation for A2 and B1 climate change scenarios from 1990 to 2080 was calculated based on the projected air temperature (Ta) applied in Eq. (2) using the estimated C coefficient. The projected air temperature used in the hydrodynamic simulation is produced by adding the projected anomalies to the reference (1990) air temperature. As discussed earlier, based on the long term record analysis of 12 stations in the AG, the Ta was the only parameter that has shown significant change while the change of other parameters was insignificant. Reference values of monthly average wind speed and relative humidity were kept unchanged. Secondly, a regression relationship was produced between 1990 projected monthly evaporation in the first step above and the reference average monthly evaporation measured from the BIA. Finally, the produced regression relationship in the second step was used to scale the projected monthly variation to that at BIA. Fig. 5 represents a three year moving average of the used evaporation rate input file for A2 and B1 scenarios.

Table 15 Projected increase percentage by technology until 2080 for the optimistic development scenarios.

MSF MED RO Sum

(2007–2012) increase %

Annual increase %

Market share % of technology

Annual increase of technology considering optimistic development (13.2%)

Increase % in 2020 to 2012

Increase % in 2050 to 2012

Increase % in 2080 to 2012

27 113 477

5.4 22.6 95.4 123.4

4.4 18.3 77.3 100

0.6 2.4 10.2 13.2

5 19 82 105.6

22 92 388 501.6

39 164 694 897.6

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

8

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

Table 16 Increase in salinity and temperature at discharge by desalination technology (source: expert opinion SWCC). Technology

Recovery (%)

Salinity increase to ambient (ppt)

Temperature increase to ambient (°C)

MSF MED HYBRD RO

25 33 32.5 50

10 15 24 44.8

10 10 7.3 1

Table 18 Approximate withdrawals and consumptions, not accounting for ambient temperature or plant efficiency (rounded and adapted from EPRI 2002b; DOE, 2006). Plant and cooling system type water

Withdrawal (thousand liters/MWh)

Consumption (thousand liters/MWh)

Fossil fuel/biomass/waste | once-through cooling Fossil fuel/biomass/waste | closed-loop cooling Nuclear steam | once-through cooling Nuclear steam | closed-loop cooling

76–190 2–23 95–230 3–4

1 2 1.5 3

2.2. Coastal effluent data Effluents from desalination plants, electricity power stations, and oil refineries along the AG coastline are considered. The access to actual coastal effluent data is extremely limited. The GWI-IDA (2012) estimated that the AG holds a total seawater desalination production capacity of 20 Mm3/day, of which 80% is produced using thermal desalination and 20% by membrane desalination. The Arabian peninsula shoreline holds 97.5% of the total AG desalination capacity along its shallow shoreline characterized by low circulation and high evaporation. A rough estimate reveals that almost 70 Mm3/day is circulated by the desalination plants and reject brine is returned back to the AG with increased salinity and temperature. The World Bank data portal (http:// data.worldbank.org) provides annual statistics of the national electricity production. This data was used as the starting point to estimate the cooling water demands for power production following the estimated rates published by the U.S. Department of Energy “Report to Congress on the Interdependency of Energy and Water” (DOE, 2006). Similarly, the U.S. Energy Information Administration (http://www.eia.gov) provides annual statistics of the oil industry in which the refinery capacities are listed. To estimate the cooling water demands for refineries, we refer to (http://pubs.usgs.gov/wsp/1330g/report.pdf) “Water Requirements of the Petroleum Refining Industry” (OTTS, 1963). The following sections describe the data used for modeling the long term discharge from desalination, power, and refinery coastal facilities.

2.2.1. Desalination reject brine discharges and future projection The GCC water resource experts committee of the GCC board published its first report “Salt Water Desalination in the GCC, history, current and future” in 2010 using 2008 data and including for the first time a joint GCC governmental effort to disclose data sources of the desalination industry. This reflected the degree of cautiousness and concern of the GCC countries for the importance of data sharing for planning and research purposes as was clearly mentioned in the report preface. The report (GCC-Desalination, 2010) included a historical review of the desalination industry in the GCC. It also included a listing of all desalination plants along with the desalination technologies and production capacities and dates of commissioning as well as the projects under construction and a 5 year projection for desalination capacity growth expectance in the GCC. The report presented a very advanced platform for conducting this study with the detailed information it provided. The data of the report was analyzed thoroughly and was used to verify the GWI-IDA (2012) maps. Together the two sources were integrated to conclude the final distribution of the desalination plants and other Table 17 Annual increase rate (%) of electricity production in the AG (1990–2011). Country

Annual increase

United Arab Emirates Bahrain Iran Kuwait Qatar Saudi Arabia

23 14 15 10 26 12

details including the historical desalination status (1990), technology distribution, rate of capacity increase and latest production capacities. The largest plants in the GCC are as follows: (a) Al-Jubail in Saudi Arabia, at 2.01 Mm3/day; b) Jabal Ali on the coast of Dubai, at 1.17 Mm3/day; (c) Al-Taweelah, Um An Nar and Shuweihat on the coast of Abu Dhabi, at, respectively, 1.06, 0.86 and 0.45 Mm3/day; and (d) Doha-East and West in Kuwait at 0.69 Mm3/day. Table 11 lists the coastal desalination plants in the AG using 2008 data (GCCDesalination, 2010). Establishing the size and profile of Iran's desalination industry is a far from exact science (GWI, 2007). Iran desalination data in Table 11 was compiled from various sources but mainly from (http://www.globalwaterintel.com/archive/6/7/general/iransinstalled-desalination-profile.html). Lattemann et al. (2013) provided an overview of more recent information published in the GWI-IDA (2012) inventory report. The report indicated that the AG holds a total seawater desalination capacity of more than 20 million cubic meters per day (Mm3/d) representing 33% of the worldwide daily drinking water production from seawater. Accordingly this represented an increase of 66% over the total seawater desalination capacity of 12.1 Mm3/d in the AG in the year 2007, which was then representing 44% of the global capacity (Lattemann, 2010). They added that while multi-effect distillation (MED) and reverse osmosis (RO) capacities increased above average by 113% to 3.3 Mm3/d and by 477% to 4.2 Mm3/d, respectively, the multi-stage flash (MSF) distillation capacity increased by only 27% to 12.5 Mm3/d. However, MSF is still the predominant process in the Gulf region in absolute terms and

Table 19 Annual increase percentage of oil refinery production in the AG (1986–2010). Country

Annual increase

United Arab Emirates Bahrain Iran Kuwait Qatar Saudi Arabia

0.030 0.004 0.032 0.026 0.076 0.008

Table 20 Refineries in the AG with daily production (2010) (thousand barrels/day). Country Kuwait

Saudi Arabia Bahrain Qatar United Arab Emirates Iran

Refinery

Capacity

Alahmadi Shuaiba Abdulla RasTanura Jubail Jubail Jubail Bapco Um Said RasLafan Shaheen Abudhabi Ruwais Lavan Bandar Abbas

470 200 270 550 400 305 400 250 147 146 250 280 120 20 335

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

9

Table 21 Climate change and coastal effluent scenarios. Scenario #

Name

Description

Climate change A2

1 2 3 4 5 6 7

Ref. (RL) (BAU) (OP) (BAU) A2 B1 (A2) (OP) (BAU) (B1) (OP) (BAU)

Reference conditions from calibration phase at 1990 Individual realistic business as usual coastal effluents only Individual optimistic business as usual coastal effluents only Individual climate change scenario only Individual climate change scenario only Combined optimistic business as usual coastal effluents and climate change Combined optimistic business as usual coastal effluents and climate change

Future demand B1

RL

Cooling water OP

x x

x x

x x

x x

x x x x

Table 22 Summary of individual and combined scenarios on the long term salinity results. Category

Effluent change

Scenario

RL(BAU) OP(BAU)

Climate change

B1 A2

Combined

B1-OP(BAU) A2-OP(BAU)

Season

W S W S W S W S W S W S

Domain range

Salinity change (year 2080)

Scenario range

Discharge locations

Restricted flow locations

Shallow (b20 m)

Central (N20 m)

Increase (max. 1.9 ppt) Increase (max. 1.75 ppt) Increase (max. 2.7 ppt) Increase (max. 2.7 ppt) Decrease (max. −1.0) Decrease (max. −0.25 °C) Decrease (max. −0.75 °C) Decrease (max. −0.25 °C) Increase (max. 3.7 ppt) Increase (max. 3.7 ppt) Increase (max. 2.7 ppt) Increase (max. 3.5 ppt) 0.0 to 3.7 ppt

Increase (max. 1.2 ppt) Increase (max. 1.2 ppt) Increase (max. 2.5 ppt) Increase (max. 2.5 ppt) Decrease (max. −1.5 ppt) Decrease (max. −1.0 ppt) Decrease (max. −1.9 ppt) Decrease (max. −1.2 ppt) Increase (max. 1.5 ppt) Increase (max. 1.7 ppt) Increase (max. 2.5 ppt) Increase (max. 2.5 ppt) −1.9 to 2.5 ppt

Increase (max. 0.5 ppt) Increase (max. 0.5 ppt) Increase (max. 0.5 ppt) Increase (max. 0.5 ppt) Decrease (max. −0.8 ppt) Mixed (−0.5 to 0.2 ppt) Mixed (−0.5 to 0.4 ppt) Mixed (−0.5 to 0.2 ppt) Decrease (max. −1.1 ppt) Mixed (−0.1 to 0.1 ppt) Mixed (−1.5 to 0.5 ppt) Increase (max. 0.5 ppt) −1.5 to 0.5 ppt

Decrease (max. −0.6 ppt) No change Decrease (max. −0.5 ppt) No change Mixed (−1.0 to 0.6 ppt) Mixed (−0.7 to 1.0 ppt) Mixed (−0.2 to 0.7 ppt) Mixed (−0.7 to 1.0 ppt) Decrease (max. −1.2 ppt) Mixed (−0.5 to 1.2 ppt) Mixed (−1.0 to 1.0 ppt) Increase (max. 1.2 ppt) −1.2 to 1.2 ppt

accounts for 63% of the production whereas MED and RO account for only 16% and 21%, respectively. The approximate locations of MSF, MED and RO desalination capacities in the AG are shown in Figs. 6 to 8 with cumulative capacities for each location and hence a single data point may represent one or several desalination plants in the same city or location by showing their combined capacity. Where the cumulative capacity exceeds 100,000 m3 /d, the exact value is also shown. Future projections of desalination brine discharges (m3/day) is a very complex issue and is directly related to the past and future economic development trends, the freshwater demands in each country, and the choice of desalination technologies to be adopted. Several literature sources were investigated to estimate the projected increase rate of desalination capacities and the subsequent brine discharge to be employed for future projections and long term modeling. It was decided that an annual increase ratio of 5% representing a realistic rate was an acceptable rate which most governments move around while the optimistic annual development scenario used for the United Arab Emirates at 13.2% (UAE-MOEW, 2010) is used as the higher limit. The

−0.6 to 1.8 ppt −0.1 to 1.8 ppt −0.5 to 2.7 ppt −0.1 to 2.7 ppt −1.9 to 0.6 ppt −1.2 to 1.0 ppt −1.9 to 0.7 ppt −1.2 to 1.0 ppt −1.2 to 3.7 ppt −0.5 to 3.7 ppt −1.5 to 2.7 ppt 0.5 to 3.5 ppt

findings of the projected freshwater demand increase rates in GCC countries from the literature search are summarized in Table 12. As for the desalination technology future market shares, there is a clear trend of increase in MED and RO technologies over the MSF. This is due to the significant improvements and advantages that both emerging technologies have shown in terms of cost reduction, increased capacity of production, low energy consumption, and improved process reliability. The desalination technology market share growth rates from 2007 to 2012 (Lattemann et al., 2013) are considered a useful indicator of the degree of confidence of the desalination industry planners and will be used to formulate the projection scenarios of desalination capacity growth. Table 13 includes the increase percentage for each technology until 2080 with an estimated total increase to 340% in 2080. Recent data (GWI-IDA, 2012) is used as a datum to project the future increase. The volume of water processed by desalination plants to obtain fresh water is very huge with freshwater conversion efficiencies ranging from 25% for MSF technology to better than 50% for RO (UAE-MOEW, 2010). Table 14 states the ratios specified for estimating the brine discharge

Table 23 Summary of individual and combined scenarios on the long term temperature results. Category

Effluent

Scenario

RL (BAU) OP (BAU)

Climate change

B1 A2

Combined

B1-OP (BAU) A2-OP (BAU)

Domain range

Season

W S W S W S W S W S W S

Temperature change (year 2080)

Scenario range

Discharge locations

Restricted flow

Shallow (b20 m)

Central (N20 m)

Increase (max. 8.0 °C) Increase (max. 5.7 °C) Increase (max. 5.5 °C) Increase (max. 5.5 °C) Decrease (max. −0.5 °C) Increase (max. 2.7 °C) Increase (max. 1.0 °C) Increase (max. 3.5 °C) Increase (max. 5.2 °C) Increase (max. 6.0 °C) Increase (max. 6.0 °C) Increase (max. 7.0 °C) 0.0 to 8.0 °C

Increase (max. 0.5 °C) No change No change No change Decrease (max. −0.5 °C) Increase (max. 2.7 °C) Increase (max. 1.0 °C) Increase (max. 3.5 °C) No change Increase (max. 2.0 °C) Increase (max. 1.0 °C) Increase (max. 3.5 °C) 0.0 to 3.5 °C

Increase (max. 0.5 °C) No change No change No change No change Increase (max. 2.7 °C) Increase (max. 1.0 °C) Increase (max. 3.5 °C) No change Increase (max. 2.5 °C) Increase (max. 1.5 °C) Increase (max. 3.5 °C) −0.5 to 3.5 °C

Mixed (−0.5 to 1.0 °C) No change Increase (max. 0.5 °C) No change Mixed (−0.5 to 1.0 °C) Increase (max. 2.7 °C) Increase (max. 1.5 °C) Increase (max. 3.5 °C) Mixed (−0.5 to 1.0 °C) Increase (max. 2.5 °C) Increase (max. 2.5 °C) Increase (max. 3.5 °C) 0.0 to 3.5 °C

−0.5 to 8.0 °C 0.0 to 5.7 °C 0.0 to 5.5 °C 0.0 to 5.5 °C −0.5 to 1.0 °C 2.3 to 2.7 °C 0.5 to 1.5 °C 2.7 to 3.5 °C −0.5 to 5.2 °C 2.0 to 6.0 °C 0.0 to 6.0 °C 2.7 to 7.0 °C

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

10

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

Fig. 4. Measured and calculated water temperature (Tw) using the regression relationship.

used and the properties and degree of mixing with cooling water prior to discharge. Such information is site specific and may vary for the same site based on the season and the source water properties. For this study and upon consultation with desalination expert opinion from the Saline Water Conservation Corporation (SWCC), the increase in the brine salinity and temperature added by each desalination technology to the ambient intake salinity and temperature are shown in Table 16. These values are used to represent the increase during the future hydrodynamic model simulations.

Fig. 5. Projected evaporation for the A2 and B1 scenarios.

following international experience (World Bank, 2011). For the RO process, 50% recovery is used for estimating the brine discharge projection. Moreover, in order to include extreme case scenario in the study an optimistic development scenario was conducted using an annual increase rate of 13.2% according to the UAE 30-year water conservation strategy which represents a very high estimate given the global economic status and the local demand increase rates. Table 15 provides the increase percentages used for the optimistic scenario. The coastal desalination plants' reject brine properties (considering only salinity and temperature) are a function of the desalination process

2.2.2. Future projection of cooling water discharges from thermoelectric power and refinery coastal facilities Due to the large amount of water required to cool thermoelectric power plants, and in light of the predicted future increase in energy consumption for the coming years (UAE-MOEW, 2010; DOE, 2010; WEC, 2007; IPCC, 2000), cooling water discharges associated with power generation must be taken into consideration when the long term thermal impacts at desalination plant intakes are to be evaluated. The exact amount of cooling water required depends on the energy source used, cooling technology, plant efficiency, ambient temperature, and relative humidity, so it is difficult to obtain exact data without detailed records and the government capacity to process them (Kohli and Frenken, 2011). In the AG region thermal distillation, including multistage flash distillation (MSF) and multi effect distillation (MED), takes place almost predominantly in cogeneration plants where water is distilled from the

Fig. 6. MSF plants in the AG region (source:Lattemann et al., 2013).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

11

Fig. 7. MED desalination plants in the AG region (source:Lattemann et al., 2013).

Fig. 8. RO desalination plants in the AG region (source:Lattemann et al., 2013).

thermal energy used to produce electricity. Cogeneration setup (power/ desalination) has proven to be a cost effective operation with savings in thermal energy costs approaching 30% (Al-Mutaz, 2001). Moreover, the cooling water discharges from thermoelectric power plants are mixed with desalination brine discharge to lower the stream temperature and salinity prior to returning it back to the sea. Temperature rises of 10–15 °C might still be expected in the receiving water body (EPRI, 2002a). The increase in electrical power production in the AG region was dictated by development needs, population growth, urbanization, and improved life standards. As indicated earlier on several occasions, the data availability in the region is very limited and reliance had to be made on various international sources to bridge the gap of information. Fig. 9 represents the growth of electricity production in the AG region with the source data obtained from (http://data.worldbank.org/indicator/ EG.ELC.PROD.KH). It is worth noting that the shoreline of Iran on the AG is mostly undeveloped except for a limited number of industrial complexes such as Asalouyah, Mobin, Bandar Abbas, Kishm and Bushehr. Data collected from various sources indicated that the share of electricity production held by Iran on the AG was very small

(0.005%) compared to the actual electricity production of the country (http://enipedia.tudelft.nl/wiki/Iran/Powerplants). Saudi Arabia electricity production on the AG side is estimated at 64% of the total country production based on the thermal desalination plants' capacity share on the AG as revealed by Lattemann et al., (2013). For long term projection, the annual rate of electricity production increase (Table 17) was produced from the data in Fig. 9 for each country and was used to estimate the increase in the power plants' cooling water discharges. The discharge points of power plants cooling water in the hydrodynamic

Fig. 9. Electricity production capacity in the AG.

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

12

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

Fig. 10. Total oil refinery production capacity by country.

model are safely assumed to coincide with the desalination plants discharge locations based on the predominance of cogeneration in the AG region.

In order to estimate the cooling water demands, consumption, and discharges from thermoelectric power plants the U.S. Electric Power Research Institute report (EPRI, 2002a) and the U.S. Department of Energy report (DOE, 2006) were consulted and used as the best available estimates. Table 18 provides the information obtained on the ratios of withdrawal and consumption related to power plants used in estimating the cooling water needs as reported by Kohli and Frenken (2011). A review of the U.S. Energy information administration international data portal (http://www.eia.gov) for the oil refinery capacities in the AG revealed limited increase in the refinery capacities in the AG except for Iran (Fig. 10). Table 19 provides the average increase % rates from 1985 to 2011. For long term projection, the annual increase rate of oil refinery production (Table 19) is produced from the data in Fig. 10 for each country and is used to estimate the increase in the oil refinery cooling water discharges. Based on the National Iranian Oil Company information portal (http://en.nioc.ir), Iran holds two refineries along the AG namely Lavan refinery and Bandar Abbas refinery with a total capacity of 20,000 and 335,000 barrels/day respectively. This accounts for 21% of the countries refinery capacity.

Fig. 11. Seasonal surface salinity (top) and temperature (bottom) spatial distribution results for RL (BAU) scenario for winter (left) and for summer (right).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

Similarly, for Saudi Arabia two refineries are located along the AG namely Al-Jubail and RasTanura with a total capacity of 1,105,000 and 550,000 barrels/day respectively and accounting to 86% of the country's total refinery capacity. The list of refineries was defined using the information available at (http://en.wikipedia.org/wiki/List_of_oil_refineries) for the AG. Data of Fig. 10 and Table 19 were adjusted to reflect the actual share of Iran and Saudi Arabia oil refinery facilities discharged in the AG. Table 20 lists the main refineries in the AG with the corresponding capacity. The USGS report on “Water Requirements for Petroleum Refining Industry” (OTTS, 1963) was used to estimate the cooling and freshwater losses from the refineries. The report considers a ratio of 345 gal/barrel of crude as cooling water discharge and evaporation losses to 28 gal/ barrel of crude. 2.2.3. Future scenarios Projections of climate changes, coastal effluents and future demands were used to develop 6 scenarios (Table 21). Scenario 1 represents the baseline (year 1990) conditions developed at the initial calibration/validation phase of the hydrodynamic model. Scenarios 2 and 3 consider simulating the individual effect of coastal effluents (no climate change included). Based on the countries' expected water demands two future

13

annual increase rates are used, 5% for the realistic case (RL)(BAU) and 13.2% for the optimistic case (OP)(BAU).Estimated cooling water discharges from thermoelectric power and refinery coastal facilities are added to scenarios 2 and 3. Scenarios 4 and 5 consider simulating the individual effects of climate changes (scenarios A2 and B1) including projected air temperature, precipitation, sea level rise and the developed surface evaporation. Scenarios 6 and 7 consider simulating the combined effects of climate change scenarios A2 and B1 with the optimistic business-as-usual (OP)(BAU) scenario. The (OP)(BAU) scenario considers the annual increase of the desalination technology market share in the Arabian Gulf for the years 2007–2012as reported by Lattemann (2013) and assumes that it will hold in the future (Tables 13 and 15). This assumption is considered highly conservative due to exceptional growth witnessed during that period and was justified based on worst case scenario approach taken for the assessment of coastal discharges. From the resulting simulation files, spatial difference maps of salinity and seawater temperature were produced by subtracting the results of the tested scenarios from the reference scenario for year 1990 (baseline). Results of seasonal difference maps at 30 year intervals (2020, 2050, and 2080) were produced to monitor the changes. The future changes in response to the designed scenarios were used to explore the assimilative capacity of the AG water basin.

Fig. 12. Seasonal surface salinity (top) and temperature (bottom) spatial distribution results for OP (BAU) scenario for winter (left) and for summer (right).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

14

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

3. Results and conclusions: The hydrodynamic modeling results of surface salinity and water temperature changes in response to the attempted scenarios are presented below. The terminal simulation year 2080 is only presented for shortness as the review of the results revealed a relatively steady buildup of the resulting responses. 3.1. Results The presented surface map results of seasonal salinity and temperature for the year 2080 represents the terminal conditions of a 90 year simulation (1990–2080) obtained using the applied scenarios. Three scenario categories were addressed: the individual coastal effluents; the individual climate change; and the combined coastal effluents and climate change. The results showed variable responses based on the dynamic balance between the coastal effluents, the local conditions at the receiving

waters (assimilative capacity) and the climate change effects. The coastal effluent effects were mostly localized to the nearshore areas while the climate change showed a more general regional effect over the AG. The assimilative capacity is a function of circulation, flushing, location, depth, and degree of flow restriction. The amount of effluent discharged is a prime factor in the resulting effects. Generally, higher discharges are more persistent and cause larger changes in terms of value change and spatial extent. This is evident at the major discharge locations such as at Al-Jubail industrial complex in Saudi Arabia comprising a number of desalination, refinery, petrochemical, and massive power generation production plants. The circulation is generally weak along the shallow parts along the Arabian Peninsula shoreline which leads to less dissipation and allows for coastal effluent accumulation. Flushing taking place in the shallow areas along the shoreline results in the improvement of local water quality conditions. It takes place in the lower portion of the water column and is mainly of seasonal nature. Summer high evaporation conditions result in better mixing of the water column. The hyper-saline

Fig. 13. Seasonal surface salinity (top) and temperature (bottom) spatial distribution results for B1 climate change scenario for winter (left) and for summer (right).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

heavier water sinks down and moves with gravity to the deeper central portions and out of the Gulf through the Strait of Hormuz. With the residence time increasing northward away from the Strait of Hormuz the northern parts of the AG near Kuwait and Saudi Arabia are more exposed to discharge accumulation compared to the southern parts along the UAE coastline. Deeper bathymetry and unrestricted flow along the Iranian coast provide much better circulation and dissipation conditions. No accumulation was witnessed along the Iranian coast. Shallow and restricted flow around Bahrain, south of Qatar and inside the Kuwait bay played a major role in the accumulation of discharges and the increased salinity as well as temperature.The following sections include a graphical representation of the resulting seasonal surface salinity and temperature spatial distribution in the year 2080 for the tested scenarios. 3.2. Results summary and discussion. 3.2.1. Salinity The model was first introduced by the individual RL (BAU) and OP (BAU) coastal effluents at the designated discharge locations in the

15

AG. Results (Figs. 11 and 12) showed localized increasing salinity effects close to the discharge locations with the OP(BAU) showing more pronounced effects. The change in the deep water and closer to the Strait showed a tendency towards decreasing. The winter season shows slightly improved circulation in the central deeper and southern parts. Keeping in mind that the presented surface currents also represent the difference change of the simulated currents from the baseline current conditions, it is not clear whether the long term release of coastal effluents have contributed to this improved circulation near Qatar and UAE especially that no change was noticed in summer surface circulation. The results along the Iranian coast do not show any changes in response to the coastal effluents due to the better circulation and low discharge properties. The climate change scenarios (Figs. 13 and 14) showed dominant effects over the entire water basin. The changes differed with depth and circulation characteristics. Winter season is characterized by the IOSW influx moving along the deeper part of the AG and along the Iranian coastline. The overlaid difference circulation clearly indicates the presence of the IOSW winter influx. This influx is present in the difference circulation map probably due to the added climate change sea level

Fig. 14. Seasonal surface salinity (top) and temperature (bottom) spatial distribution results for A2 climate change scenario for winter(left) and for summer (right).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

16

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

rise component at the boundary. The lower (B1) and the higher (A2) IPCC climate projections were used.The salinity of both scenarios showed a consistent mild decreasing trend at the discharge locations (no discharges included) and restricted flow locations. Both scenarios had comparable effects with the A2 causing more decrease at the restricted flow locations. In the shallow and central parts the response was mixed but more towards decrease. The decrease is due to improved water column mixing most probably in response to higher evaporation resulting from higher air temperatures. A general observation is that the surface salinity decreases as depth decreases and shows mixed increase and decrease to slight increase with increasing depth. The combined effect of coastal effluents and climate change on the surface salinity was evaluated using the optimistic discharge properties with the B1 and A2 climate change projections (Figs. 15 and 16). The shown salinity increase due to the individual coastal effluents was balanced with the decreasing effects of the climate change. This dynamic balance caused lowering the salinity at the discharge locations where the discharge properties had the dominant effect. The B1-OP (BAU) scenario produced the highest salinity increasing effect by 3.7 ppt. The A2OP(BAU) scenario produced slightly lower maximum salinity due to the

improved water column mixing associated with higher evaporation. Table 22 summarizes the effect of each scenario and the limits of the obtained salinity results at the discharge locations, restricted flow locations, shallow and central deep parts. 3.2.2. Temperature Results of both individual effluent scenarios showed clear trend for local surface temperature increase at the discharge locations. Insignificant changes were seen elsewhere with less consistent responses which reflects the sensitivity of surface temperature to the dynamics of air water interface related to the heat flux and water circulation. Despite that the realistic discharge conditions have lesser discharge amounts they result in the highest change at the discharge locations with a maximum of 8.0 °C in winter and 5.7 °C in summer, while the optimistic conditions result in a typical change of 5.5 °C in both seasons. A possible interpretation of this is that the more intensive discharges at the relatively shallow discharge location manipulates the local circulation conditions more drastically resulting in better mixing and dilution which tends to reduce the surface water temperature change of the OP scenario.

Fig. 15. Seasonal surface salinity (top) and temperature (bottom) spatial distribution results of combined B1-OP scenario for winter (left) and for summer (right).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

The individual climate change scenarios showed clear temperature increasing trends. Winter related effects of improved circulation due to the IOSW influx are obvious in the deep parts while the summer produces more uniform conditions with the diminishing effects of the IOSW.With the B1 scenario, the shallower parts in the north and along the Arabian coastline showed marginal reduced surface temperature by 0.5 °C. The winter warm influx of the IOSW are obvious in the deep central parts and along the Iranian coast which causes a temperature increase up to 1.0 °C. Summer shows a general consistent increase response over the entire domain with 2.0–2.5 °C. The Hormuz vicinity showed slightly reduced temperature change of 1.0–1.5 °C affected by the exchange with the relatively cooler water entering from the Gulf of Oman. A2 represents a more extreme climate change scenario which produced larger spatial and higher temperature changes. Shallow locations are subject to faster heating and cooling and water column mixing. Winter temperature increase changes reaches slightly above

17

1.0 °C and up to 2.7 °C in summer along the shallow parts. The deeper parts are affected by the IOSW influx and reaches 1.5 °C in winter and as the effect of the IOSW diminishes the summer temperature change tends to be more uniform and reaches 3.5 °C. The combined scenarios results showed the predominant mild climate change impacts over the regional scale temperature while the coastal effluents effects were very drastic but spatially limited close to the discharge locations. The long term changes are actually the added effect of both the forcing factors. Table 23 summarizes the effect of each scenario and the limits of the obtained temperature results at the discharge locations, restricted flow locations, shallow and central deep parts. 3.3. Conclusions and recommendations: The obtained hydrodynamic evaluation results of surface salinity and seawater temperature for the tested scenarios represent an advanced

Fig. 16. Seasonal surface salinity (top) and temperature (bottom) spatial distribution results of combined A2-OP scenario for winter (left) and for summer (right).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

18

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx

and well substantiated long term and comprehensive evaluation of the projected impacts of climate change and coastal effluent factors. It revealed the clear effects of climate change and coastal effluents in increasing the salinity and temperature in the AG. With the projected limited precipitation, increase in temperature and limited exchange with the open ocean despite the projected sea level rise, the excessive evaporation acts to increase the salinity in the Gulf. The climate change impacts were variable based on the depth, circulation, degree of flow restriction and proximity to the Strait of Hormuz and the related exchange with open ocean water. Shallow areas along the Arabian coast experience better flushing and water column mixing with intensified seasonal cooling and heating cycles which result in marginal salinity reduction. The impacts of the climate change were found to be affirmative towards increasing the AG temperature and salinity. The impacts of coastal effluents on the water temperature and salinity were found to be variable and were based on the discharge properties as well as the receiving water properties. The impacts were most drastic at the shallow depths along the Arabian coastline. Increased temperature and salinity contours were seen approximately 20 km from Al-Jubail discharge locations and in Kuwait, Qatar, UAE, and the restricted flow area south of Bahrain as a result of massive discharges beside the shallow depths and low circulation water properties. The Iranian coastline by contrast showed very limited response to the applied scenarios due to the low discharges introduced and the much larger depth and better circulation properties. The combined scenarios affirmed the individual scenario results for the Arabian shoreline showing a resultant increase in the seawater temperature and salinity. With these obvious impacts on the nearshore water quality there is an apparent need to develop our understanding of the impacts of these changes on the local marine environment as well as on the long term operational conditions of the coastal facilities utilizing seawater in their processes. More strategically there is a need to evaluate the long term impacts on the coastal desalination plants and their ever increasing capacities representing the only viable solution for potable water supplies in the AG region. Highlights • Hydrodynamic modeling is used to evaluate long term changes of seawater temperature and salinity. • Individual and combined scenarios of projected climate change and coastal effluents were applied. • Climate change impact was regional and varied spatially based on circulation, depth, degree of flow restriction, and heat flux. • Coastal effluent impact is limited locally and depend on the dynamic balance of discharge and receiving water characteristics. • Combined scenarios affirmed net seawater temperature and salinity increase near major outfalls along Arabian coastline.

References Ahmad, F., Sultan, S.A.R., 1991. Annual mean surface heat fluxes in the Arabian Gulf and the net heat transport through the Strait of Hormuz. Atmosphere-Ocean 29, 54–61. Al Sarmi, S. and Washington, R. (2011): Recent observed climate change over the Arabian Peninsula, J. Geophys. Res.-Atmos., Vol. 116 Issue D11 Al-Dousari, A.E., 2009. Desalination leading to salinity variations in Kuwait marine waters. Am. J. Environ. Sci. 5 (3), 451–454. Alessi, C.A., Hunt, H.D., Bower, A.S., 1999. Hydrographic data from the U.S. Naval Oceanographic Office: Persian Gulf, Southern Red Sea, and Arabian Sea 1923–1996. Woods Hole Oceanog. Inst. Tech. Rep., WHOI-99-02. Al-Mutaz, I., 2001. Potential of nuclear desalination in the Arabian Gulf countries. Desalination 135, 187–194. Al-Osairi, Y., Imberger, J., Falconer, R., 2011. Mixing and flushing in the Persian Gulf (Arabian Gulf). J. Geophys. Res. 116, C03029. Areiqat, A., Mohamed, K.A., 2005. Optimization of the negative impact of power and desalination plants on the ecosystem. Desalination 185, 95–103.

Azam, M.H., Elshorbagy, W., Ichikawa, T., Terasawa, T., Taguchi, K., 2006. 3D application to study the residual flow in the Arabian Gulf. ASCE J. Waterw. Port Coastal Ocean Eng. 132 (5), 388–400. Darwish, M.A., Alotaibi, S., Alfahad, S., 2008. On the reduction of desalting energy and its cost in Kuwait. Desalination 220, 483–495. Dawoud, M., 2007. Water Scarcity in the GCC Countries: Challenges and Opportunities. Gulf Research Center, Dubai, United Arab Emirates. Deltares, 2011. Deft3D-Flow User Manual. Delft, Netherlands. Department of Energy-USA (DOE ), 2006. Energy demands on water resources: report to Congress on the interdependency of energy and water. Department of Energy-USA /Energy Information Administration (DOE), 2010. Annual Energy Outlook 2010 with Projections to 2035. Electric Power Research Institute (EPRI), 2002a. Comparison of Alternate Cooling Technologies for California Power Plants Economic. Environmental and Other Tradeoffs. Electric Power Research Institute (EPRI), 2002b. Water & Sustainability, U.S. water consumption for power production-the next half century. Electric Power Research Institute. Topical Report Vol. 3. Elhakeem, A., Elshorbagy, W., 2013. Evaluation of the long-term variability of seawater salinity and temperature in response to natural and anthropogenic stressors in the Arabian Gulf. Mar. Pollut. Bull. 76 (1–2), 355–359. Elhakeem, A., Elshorbagy, W., AlNaser, H., Dominguez, F., 2015a. J. Water Resour. Plan. Manag. http://dx.doi.org/10.1061/(ASCE)WR.1943–5452.0000507, 04015007. Elhakeem, A., Elshorbagy, W., Bleninger, T., 2015b. Long-term hydrodynamic modelling of the Arabian Gulf. Mar. Pollut. Bull. 94 (1–2), 19–36. Elshorbagy, W., Azam, M.H., Taguchi, K., 2006. Hydrodynamic characterization and modelling of the Arabian Gulf. J. Waterw. Harb. Coast. Eng. Div. ASCE 132, 47–56. Elshorbagy, W., Azam, M.H., Nakata, K., Terasawa, T., 2008. Temperature-salinity field of the shallow shelf of southern Arabian Gulf. Far East J. Ocean Res. Vol. 1 (Issues 2-3), 99–126. Elshorbagy, W., Azam, M.H., Elhakeem, A., 2013. Temperature-salinity modeling for Ruwais coastal area in United Arab Emirates. Mar. Pollut. Bull. 15;73 (1), 170–182. GCC-Desalination report, 2010. Saline Water Conversion History, Present and Future. 1st ed. Group of Experts, Desalination Committee on Water Resources In the countries of the Cooperation Council for the Arab Gulf States (in Arabic). Gill, A.E., 1982. Atmosphere-Ocean-dynamics International Geophysics Series, vol 30. AcademicPress Inc, New York p 662. GWI and IDA, 2007. DesalData.com by Global Water Intelligence and Water Desalination Report Incorporating the IDA Desalination Plants Inventory. Harrold, L.L., Schwab, G.O., Bondurant, B.L., 1986. Agricultural and forest hydrology. Agricultural Engineering Department, Ohio State University, Columbus (271 pp.). Hemming, D., Betts, R., Ryall, D., 2007. Environmental stresses from detailed climate model simulations for the Middle East and Gulf region. Report (ref. MOEN/04/04/ 02/02b) completed by the Met Office Hadley Centre for UK Ministry of Defence project ‘Defence and Security Implications of Climate Change commissioned by the UK Ministry of Defence (Contract ref.RT/COM/4/5015). Hughes, P., Hunter, J.R., 1979. A proposal for a physical oceanography program and numerical modeling of the KAP Region. UNESCO, Div. Mar. Sci., Paris, MARINF/27 (16 Oct. 1979. 102 pp.). Hunter, J.R., 1983. A review of the residual circulation and mixing processes in the kap region, with reference to applicable modeling techniques. In: El-Sabh, M.I. (Ed.), Oceanographic Modeling of the Kuwait Action Plan (KAP) Region. UNESCO Reports in Marine Science 28, pp. 37–45 (Paris). Intergovernmental Panel on Climate Change - IPCC, 2000. Special Report on Emissions Scenarios. Intergovernmental Panel on Climate Change — IPCC, 2007. Special Report on the Regional Impacts of Climate Change and Assessment of Vulnerability. Cambridge University Press, Cambridge, United Kingdom. Johns, W.E., Yao, F., Olson, D.B., Josey, S.A., Grist, J.P., Smeed, D.A., 2003. Observations of seasonal exchange through the Straits of Hormuz and the inferred freshwater budgets of the Persian Gulf. J. Geophys. Res. 108 (C12), 3391. Kämpf, J., Sadrinasab, M., 2006. The circulation of the Persian Gulf: a numerical study. Ocean Sci. 2, 27–41. KFUPM/RI, 1990. Aramco Sustaining Research Project, Environmental Studies, Volume VI, Oceanographic Investigations Coastal and Offshore Hydrography (Dhahran, Saudi Arabia). Kim, B.K., Jeong, Y.H., 2013. High cooling water temperature effects on design and operational safety of npps in the gulf region, nuclear engineering and technology, 45(7). Kohli, A., Frenken, K., 2011. Cooling water for energy generation and its impact on national-level water statistics. AQUASTAT Programme of FAO. Lane, A., 1989. The heat balance of the North Sea. Tech. Rep. 8. Proudman Oceanographic Laboratory 251, 261. Lattemann, S., 2010. Development of an environmental impact assessment and decision support system for seawater desalination plants (PhD thesis), CRC Press, Balkema (276 pp.). Lattemann, S., Höpner, T., 2008. Environmental impact and impact assessment of seawater desalination. Desalination 220, 1–15. Latteman, S., Morelissen, R., Gils, J., 2013. Desalination capacity of the Arabian Gulfmodelling. Monitoring and Managing Discharges. International Desalination Association World Congress on Desalination and Water Reuse. Tianjin, China. Lehr, W.J., 1984. A brief survey of oceanographic modelling and oil spill studies in the KAP region. In: El-Sabh, M.I. (Ed.), Oceanographic Modeling of the Kuwait Action Plan (KAP) Region. UNESCO Reports in Marine Science 28, pp. 4–11 (Paris). Linden, O., Larsson, U., Al-Alawi, Z.S., 1988. Effects of oil pollution in a shallow subtropical marine environment. Oil Chem. Pollut. 5, 65–79. Meshal, A.H., Hassan, H.M., 1986. Evaporation from the coastal waters of the central part of the Gulf. Arab J. Sci. Res. 4 (2), 649–655.

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032

A. Elhakeem, W. Elshorbagy / Marine Pollution Bulletin xxx (2015) xxx–xxx Meyer, A.F., 1915. Computing run-off from rainfall and other physical data. ASCE 79, 1056–1155. Nakićenović, N., Swart, R., 2000. Special Report on Emissions Scenarios. Cambridge University Press, Cambridge, UK (570 pp.). Nasrallah, H.A., Balling Jr., R.C., 1996. Analysis of recent climatic changes in the Arabian Peninsula region. Theor. Appl. Climatol. 53, 245–252. OTTS, L., 1963. Water Requirements of the Petroleum Refining Industry, Geological Survey Water-Supply Paper 1330-G. United States Government Printing Office, Washington. Perroud, M., Goyette, S., Martynov, A., Beniston, M., Anneville, O., 2009. Simulation of multiannual thermal profiles in deep Lake Geneva: a comparison of onedimensional lake models. Limnol. Oceanogr. 54, 1574–1594. Privett, D.W., 1959. Monthly charts of evaporation from the North Indian Ocean, including the Red Sea and the Persian Gulf. Q. J. R. Meteorol. Soc. 85, 424–428. Purkis, S.J., Riegl, B., 2005. Spatial and temporal dynamics of Arabian Gulf coral assemblages quantified from remote-sensing and in situ monitoring data. Mar. Ecol. Prog. Ser. 287, 99–113. Purnama, A., Al Barwani, H.H., Smith, R., 2005. Calculating the environmental cost of seawater desalination in the Arabian marginal seas. Desalination 185, 79–86. Reynolds, R.M., 1993. Physical oceanography of the Gulf, Strait of Hormuz, and the Gulf of Oman—results from Mt. Mitchell expedition. Mar. Pollut. Bull. 27, 35–59. Samuelsson, P., 2010. Using regional climate change models to quantify the impact of climate change on lakes. In: George, G. (Ed.), The Impact of Climate Change on European Lakes. Springer, New York, pp. 15–32. Santer, B.D., Wigley, T.M.L., Schlesinger, M.E., Mitchell, J.F.B., 1990. Developing climate scenarios from equilibrium GCM results. Max-Planck-Institut für Meteorologie Report No. 47, Hamburg, Germany 29 pp.

19

Sheppard, C., Loughland, R., 2002. Coral mortality and recovery in response to increasing temperature in the southern Arabian Gulf. Aquat. Ecosyst. Health Manag. 5, 395–402. Sheppard, C., Al-Husiani, M., Al-Jamali, F., Al-Yamani, F., Baldwin, R., Bishop, J., Benzoni, F., Dutrieux, E., Nicholas, K., Durvasula, S., Jones, D., Loughland, R., Medio, D., Nithyanandan, M., Pilling, G., Polikarpov, I., Price, A., Purkis, S., Riegl, B., Saburova, M., Namin, K., Taylor, O., Wilson, S., Zainal, K., 2010. The Gulf: a young sea in decline. Mar. Pollut. Bull. 60, 13–38. Smith, R., Purnama, A., Al-Barwani, H., 2007. Sensitivity of hypersaline Arabian Gulf to seawater desalination plants. Appl. Math. Model. 31 (2347–2354). Tolba, M.K., Saab, N.W., 2009. Impact of climate change on Arab countries, Report of the Arab Forum for Environment and Development (AFED). Technical Publications Beirut, Lebanon (181 pp.). UAE-MOEW (Ministry of Environment and Water), 2010. United Arab Emirates Water Conservation Strategy, UAE. Ward, A.D., Trimble, S.W., 2004. Environmental Hydrology. CRCPress, Boca Raton, Florida (481pp.). Wigley, T., 2008. MAGICC/SCENGEN 5.3 User Manual (version 2), National Center for Atmospheric Research, Boulder, Colorado. http://www.cgd.ucar.edu/cas/wigley/ magicc/. World Bank, 2011. Middle East and North Africa (MENA) regional water outlook-part II, desalination using renewable energy. Final Report, FICHTNER, 70191 StuttgartGermany (http://data.worldbank.org http://pubs.usgs.gov/wsp/1330g/report.pdf http://www.eia.gov http://www.globalwaterintel.com/archive/6/7/general/iransinstalled-desalination-profile.html). World Energy Council (WEC) (Ed.), 2007. Deciding the Future: Energy Policy Scenarios to 2050. London (UK).

Please cite this article as: Elhakeem, A., Elshorbagy, W., Hydrodynamic evaluation of long term impacts of climate change and coastal effluents in the Arabian Gulf, Marine Pollution Bulletin (2015), http://dx.doi.org/10.1016/j.marpolbul.2015.10.032