To predict climate change impacts on the CO2 sink, in a given waterbody, there would be a need for modelling and simulating the hydrodynamics of this waterbody. Such a simulation can provide insight into how the CO2 sink would be affected by climate change which in turn would give ideas about how much effort is to be given to mitigate and/or adapt to the change. Knowing the current and/or future sink capacities depends, to great extent, on understanding and then modelling the complex relationship between the waterbody’s temperature and salinity in one hand and the CO2 solubility in the other hand. Successful modelling of such complex relationship leads to better estimation of the waterbody potentials to dissolve CO2 from the atmosphere. 

Climate change and CO₂ sinks in waterbodies are closely interconnected due to the crucial role that waterbodies play in carbon sequestration as they are significant carbon sinks, absorbing roughly a quarter of the CO₂ emitted by human activities each year. This absorption helps moderate global temperatures by reducing the amount of CO₂ in the atmosphere, which in turn limits the greenhouse effect and slows climate change4.

Generally, the CO2 sink, as an ecosystem function of inland waterbodies, was repeatedly considered by a number of previous studies. For example, Raymond et al., (2013) provided an estimate for regional variations in global inland water surface area, dissolved CO2 and Carbon dioxide (CO2) transfer from inland waters to the atmosphere5. Other studies explained the biogeochemical and hydrological mechanisms driving CO2 concentrations in inland waters6,7. Usually, mega development projects and associated environmental and socioeconomic changes are expected to have significant impacts on ecosystem function as CO2 sink. In this respect, seasonal and annual fluxes of CO2 emissions from Chinese inland waterbodies was quantified and their changes during the period 1980s–2010s was evaluated8,9. Due to the essential role of GHGs emission of inland waterbodies in understanding their role as a CO2 sink, several studies attempted to estimate such emissions in different parts of the world10,11.

It was noted that different previous studies emphasized the significant impacts of climate change on the ecosystem functions of inland waterbodies as CO2 sink. Several literatures employed sampling, physiochemical analysis to lake water and/or sediment and statistical analysis to study the lakes’ CO2 sinks12,13,14,15. Conversely, integrating modelling CO2 solubility in water, hydrodynamic modelling and GIS analysis to study the CO2 sink capacity of a waterbody is still a little explored area of research. Moreover, limited number of previous studies, to our knowledge, provided estimates for the change in the waterbodies CO2 sink capacity in the future under different scenarios of climate change in general and specially for Egyptian lakes. Therefore, for the Middle East inland waters, this study can be considered as the first research work investigating this topic.

The aim of this work is to propose an integrated approach to quantitatively estimate the change in a waterbody CO2 sink capacity due to climate change by using hydrodynamic modelling and GIS analysis. Such an integrated approach can support climate change mitigation strategies as it provides an insight into the waterbody CO2 sink.

Study area

Wadi El-Rayan Lakes are two man-made lakes in the hyper-arid Western Desert of Egypt near Fayoum (Fig. 1)17,18,19. Their origin is traced back to 1973 when agricultural wastewater was diverted to Wadi El-Rayan depression forming the Upper Lake first and in 1980 the surplus water that exceeds the Upper Lake’s holding capacity fell through a waterfall to form the Lower Lake20,21,22. The two lakes are at different levels in relation to the mean sea level17,18,19,23,24: the Upper Lake is at − 10.00 m AMSL, while the Lower Lake is at − 32.00 m AMSL25,26. The Upper Lake obtains its water from a diversion of a main drain called “Wadi Drain”27 and gives water to nearby reclaimed lands and fish farms18,28. The Lower Lake is a semi-closed basin that receives its water from the Upper Lake’s surplus water and some minor drains from nearby agricultural lands and the main water loss is by evaporation29. The inlet and the intake of the waterbody (Fig. 1) can be pointed out as:

• The main discharge to the waterbody which comes from the Wadi Drain through a 9 km canal then an 8 km tunnel and dispenses its water in the north-east of the Upper Lake27.

• The pump station which is located in the west of the Upper Lake and it is responsible for the irrigation of the reclaimed lands in the west.

• The discharge to the fish farms that depends on the Upper Lake’s water30.

Wadi El-Rayan Lakes, geographic location, inlets and intakes and water depth.

Full size image

Wadi El-Rayan Lakes are located in a hyper-arid area that is characterized by hot and dry climate with high evaporation rates exceeding 4.65 mm/day on average31, and low precipitation17,23,32. The prevailing wind is Northern, varying from North-Western to North-Eastern33 resulting in the formation of extensive sand dunes34,35.

The two lakes have different salinity levels ranging between 2‰ in the Upper Lake and 13‰ in the Lower Lake29,36,37,38,39. This remarkable difference in salinities may imply that they have two different ecosystems.

Data and methodology

To estimate the impacts of climate change on the CO2 sink capacity under different scenarios, a methodology of four main steps was applied including simulating the CO2 solubility (mg/L) as a function of temperature and salinity, developing a hydrodynamic model for Wadi El-Rayan Lakes, predicting key physical and chemical parameters of Wadi El-Rayan Lakes under climate change scenarios, and finally estimating the change in Wadi El-Rayan CO2 sink capacity (Fig. 2).

Methodology flowchart.

Full size image

Miscellaneous data are used in this procedure. Some of them are generated from tabulated data, interpolated from given points or contour maps (Table 2).

Modelling the CO2 solubility at different temperatures and salinities

At constant CO2 partial pressure above the waterbody, the CO2 solubility depends on both temperature and salinity. The lakes’ water temperature and salinity are changeable through different times of the year. These changing parameters of temperature and salinity control the amount of soluble CO2 in water. As both water temperature and salinity increase, the solubility of CO2 decreases16. Data given in Table 1 is used for mathematical analysis to model the CO2 solubility (mg/L) in water at different temperatures and salinities assuming two conditions:

1. (1)

   The waterbody is exposed to moist air containing 0.04% CO2.

2. (2)

   The total air pressure is 760 mmHg (1 atm).

A non-linear mathematical analysis is used to related CO2 solubility (mg/L) to both temperature (°C) and salinity (‰) using OriginLab software (OriginPro 2016, b9.3.226)43 which is a high-performance statistical analysis software. The fit equation is used later to quantitatively estimate the potential soluble amount of CO2 in a known volume of water (lake’s water).

Developing a hydrodynamic model for Wadi El-Rayan Lakes

To simulate hydrodynamic flow of Wadi El-Rayan Lakes, Delft3D-FLOW model was selected because it is an open-source, relatively easy to use, well-documented and technically supported modelling software. Besides it is used extensively and successfully to simulate lake hydrodynamics in different parts of the world44,45,46,47,48,49,50,51,52. A representative hydrodynamic model goes through three steps: model set-up, calibration and validation. Model set-up requires descriptive data about Wadi El-Rayan Lakes in addition to meteorological data. Descriptive data includes the geometrical boundaries of Wadi El-Rayan Lakes, the bathymetry, water level, temperature and salinity, bottom roughness and water viscosity besides the quantities of water inflow and intake. Meteorological data includes air temperature, humidity, cloud coverage, solar radiation, wind speed and direction. Model calibration includes the comparison of the model outputs with ground truth data such as water level, for example, and the adjustment of the model’s different parameters such as roughness and viscosity, for example, in order to yield as close values to ground truth as possible. In some rare cases, adjustments to water flow can be done to compensate for a source of water gain or loss that is not included in the model set-up process. Model validation is a performance test made for the calibrated model using a different set of data rather than those used in the model set-up and calibration and compare the model output to ground truth data using error matrices. If the error in the validation process is acceptable, the model is then representative to the waterbody under investigation.

Model set-up

The main steps to set-up the model are as follows:

• Define model dimensionality according to the purpose of the study and the data availability.

• Delineate the waterbody physical domain and construct the suitable grid representing it.

• Specify the time frame for the simulation process.

• Determine driving forces working on the waterbody and provide time series data covering the selected time frame.

• Identify the discharges to and from the waterbody and provide time series data for their quantities.

• Indicate the initial conditions at the beginning of the simulation.

For model set-up, two-dimensional (2D) model is used to represent Wadi El-Rayan Lakes. 2D model is simple enough to attain the goal of this work and complies with available data. Keeping the model as simple as needed saves the time and effort of the processing and requires less data storage capabilities. A mesh of 75 × 107 is created to represent the waterbody domain. This resolution is a compromise to ensure the accuracy of results and a suitable run time. The depth values are assigned to the grid cells by using the formerly prepared depth raster layer (Fig. 1) generated from a contour map given by38. An interpolation is made to fill-up any empty cell on the grid. To reduce the time and effort of calculations, dry cells which represent little islands and/or dry areas that exist into the waterbody’s domain are blocked.

Simulation of a whole year starts on the 1st of February 2013 (01 02 2013 00:00) and ends on the 31st of January 2014 (31 01 2014 23:00). The time frame is determined based on available data36,37,38. Forces working on the waterbody are defined to be salinity, temperature, and wind. The salinity of the two basins (the Upper and the Lower Lakes) are completely different. The Upper Lake has an average salinity value of 1.6‰ while that of the Lower Lake is 15.6‰ according to the collected field data from monitoring stations provided by the Egyptian Lakes’ Monitoring Program36. This detail is well represented by defining the salinity values in the form of spatially varied map instead of a solid value in the initial conditions in the beginning of the simulation. Water temperature and water level are also being given in the form of spatially varied maps as there are differences between the two basins of the waterbody. The mean water temperature of the waterbody is about 15.4 °C. The water level is given a value in reference to the mean sea level to reflect the difference in the two basins’ levels (the Upper Lake is at 10 m below the mean sea level while the Lower Lake is at 32 m below the mean sea level).

The roughness Chezy value is assigned to 40 and the eddy viscosity and diffusivity to 0.5 m2/s and all can further be adjusted in the calibration step if needed. The heat flux model uses relative humidity (%), air temperature (°C) and total radiation (J/m2/s), which is equivalent to (W/s), as the mean forces to determine the heat transfer and calculate the volume loss due to evaporation. Data used in the heat flux model is of high resolution (hourly-based data). Wind force is represented by high resolution data of wind speed (m/s) and direction (degrees). The prevailing wind is coming from the north and the average wind speed is about 4.4 m/s (Table 3).