The Impact of Tides and Monsoons on Tritium Migration and Diffusion in Coastal Harbours: A Simulation Study in Lianyungang Haizhou Bay, China

Abstract

Many nuclear power plants have been built along China’s coasts, and the migration and diffusion of radioactive nuclides in coastal harbours is very concerning. In this study, considering the decay and free diffusion of radioactive nuclides, a local hydrodynamic model based on the FVCOM was built to investigate the migration and diffusion of the radioactive nuclide tritium in Haizhou Bay, China. This model was calibrated according to the observed tidal level and flow velocity and direction, which provide an accurate background. This study aimed to evaluate the impact of tides and monsoons on the migration path and concentration variations in tritium over time. The results demonstrated that the simulated flow field can reflect real-life receiving waters. The distribution of the tritium concentration is affected by the flow field, which is related to the tides. Moreover, more severe radioactive contamination was exhibited in winter than in summer because monsoons may have hindered the migration and diffusion of tritium within the harbour. Given the poor hydrodynamic conditions and slow water exchange in the open ocean in Haizhou Bay, the diffusion rate of radioactive nuclides outside the bay area was higher than that within it.

1. Introduction

Around the world, nuclear power plants are mainly built in coastal areas. Nuclear accidents not only pose a direct threat to marine ecosystems, but also to humans through the food chain, potentially risking our survival and health [1,2,3]. Since the Fukushima Daiichi nuclear power plant accident in March 2011, large amounts of artificial radionuclides have been released into the atmospheric and marine environments [4,5]. The total amount of released radionuclides is estimated to be about 520 PBq, including a wide range of radioisotopes, such as ¹³¹I, ¹³⁷Cs, and radioactive tritium [6]. In addition, small-scale releases of artificial radionuclides continue to occur [7,8,9,10]. The Fukushima nuclear power plant accident has triggered widespread concern about radioactive contamination in nearshore waters, especially in nearshore harbours [11,12]. Therefore, we must grasp the migration and diffusion patterns of nuclides in coastal harbours and bays.

Tritium (³H) is often used as a marine tracer due to its suitable properties [13]. Currently, many scholars have addressed the distribution and transport of radionuclide tritium in the North Pacific Ocean after the Fukushima nuclear accident [14,15]. Although the general principles of tritium’s distribution and transfer processes within and between the various hydrological compartments are known, the variations in timescale and aspects of altitude dependence are still under debate [16]. This is especially the case in nearshore harbour waters with complex environmental conditions. The distribution of tritium in harbours of different seas is different, and some scholars have studied the transport of tritium in Daya Bay in China [17,18,19,20]. In other studies [21,22], the characteristics of radionuclide transport in China’s nearshore waters were explored using an ocean numerical model combined with random particle travelling. In a different study [23], a hydrodynamic model of the Skagit River estuary was developed based on the Finite-Volume Coastal Ocean Model (FVCOM). In another study [24], a high-resolution regional model (Nucleus for European Modelling of the Ocean) simulated the distribution of tritium in the Mediterranean Sea. In a different study [25], a combined biophysical and water cycle model simulated the dispersion of Atlantic salmon in a semi-enclosed bay. In another study [26], the combination of a hydrodynamic model and evaporation–atmosphere transport model was constructed, which assessed the average tritium release from the coast into the English Channel.

The majority of these scholars’ studies have focused more on modelling the transport and diffusion of radionuclide tritium in harbour waters and less on the factors affecting its dispersal within harbours. In another study [27], the seasonal transport patterns of pollutants from coastal farms were assessed, and the effects of positive- and oblique-pressure coastal dynamics and different farming practices on the transport of dissolved pollutants were evaluated. In a different study [28], a well-calibrated FVCOM model was used to simulate the effects of tropical cyclones on marginal seas. In another study [29], the role of tides on the ³H dispersion of liquid discharges from the Daya Bay Nuclear Power Plant was analysed. Haizhou Bay in Lianyungang, China, has a typical monsoon climate and many nuclear power plants, but no one has studied tritium dispersion modelling or the factors affecting dispersion in this harbour.

In this study, by analysing the characteristics of the hydrodynamic environment in Haizhou Bay, Lianyungang, a radionuclide transport and diffusion model was constructed using the FVCOM. Tritium was selected as a typical nuclide to explore its migration and diffusion mechanism in the coastal harbours. The hydrodynamic flow field in the region was simulated and analysed. We simulate the migration trajectory and concentration evolution of radionuclides from the outer sea to the harbour under a hypothetical accident scenario and analyse the effects of tidal action and monsoon changes on nuclide propagation and its concentration distribution. We aim to provide a scientific basis for the environmental impact assessment of radioactive liquid discharges from nuclear power plants and accumulate data for the rapid evaluation of emergency situations.

2. Modelling and Numerical Experimental Design

2.1. Model Description

A radionuclide transport diffusion model based on the Finite-Volume Coastal Ocean Model (FVCOM) [30,31] was developed to simulate tritium activity in the ocean. This model is adept at meticulously delineating complex coastal boundaries, while preserving computational expediency, thereby asserting its pre-eminence in coastal area simulation endeavours. This model’s representation of irregular bathymetric contours is accomplished through $\sigma$ coordinate vertical transformation. In terms of turbulence modelling, the horizontal and vertical dimensions are addressed using the Smagrinsky turbulence closure and Mellor–Yamada level 2.5 turbulence closure models, respectively. This dual application ensures that the model’s formulations are both mathematically and physically robust and comprehensive. Temporal discretization within the FVCOM is managed via time-splitting techniques, which differentiate between the temporal resolutions of the two-dimensional external and the three-dimensional internal modes. This stratagem significantly shortens the overall computation duration. Additionally, this model incorporates wet–dry determination to handle the tidal dynamics at the boundary, thereby enhancing the conservation properties of the calculations in regions subject to alternating wet and dry conditions.

The model uses $\sigma$ vertically oriented coordinates with the following coordinate transformation [32]:

$$\sigma =\frac{z-\xi }{H+\xi }=\frac{z-\xi }{D}$$

(1)

where $H$ represents the static water depth; $\xi$ represents the height of the free surface; and $D$ represents the total water depth. For the $\sigma$ coordinate system, the three-dimensional internal mode control equations are

$$\frac{\mathit{\partial }\zeta }{\mathit{\partial }t}+\frac{\mathit{\partial }Du}{\mathit{\partial }x}+\frac{\mathit{\partial }Dv}{\mathit{\partial }y}+\frac{\mathit{\partial }Dw}{\mathit{\partial }\sigma }=0$$

(2)

$$\begin{array}{c}\frac{\mathit{\partial }uD}{\mathit{\partial }t}+\frac{\mathit{\partial }{u}^{2}D}{\mathit{\partial }x}+\frac{\mathit{\partial }uvD}{\mathit{\partial }y}+\frac{\mathit{\partial }uw}{\mathit{\partial }\sigma }-fvD=-gD\frac{\mathit{\partial }\zeta }{\mathit{\partial }x}-\frac{gD}{{\rho }_0}\left[\frac{\mathit{\partial }}{\mathit{\partial }x}\left(D{\displaystyle {\int }_{\sigma }^0\rho d{\sigma }^{\prime }}\right)+\sigma \rho \frac{\mathit{\partial }D}{\mathit{\partial }x}\right]+\\ \frac{1}{D}\frac{\mathit{\partial }}{\mathit{\partial }\sigma }\left(K_m\frac{\mathit{\partial }u}{\mathit{\partial }\sigma }\right)+D{F}_x\end{array}$$

(3)

$$\begin{array}{c}\frac{\mathit{\partial }vD}{\mathit{\partial }t}+\frac{\mathit{\partial }uvD}{\mathit{\partial }x}+\frac{\mathit{\partial }{v}^{2}D}{\mathit{\partial }y}+\frac{\mathit{\partial }vw}{\mathit{\partial }\sigma }-fuD=-gD\frac{\mathit{\partial }\zeta }{\mathit{\partial }y}-\frac{gD}{{\rho }_0}\left[\frac{\mathit{\partial }}{\mathit{\partial }y}\left(D{\displaystyle {\int }_{\sigma }^0\rho d{\sigma }^{\prime }}\right)+\sigma \rho \frac{\mathit{\partial }D}{\mathit{\partial }y}\right]+\\ \frac{1}{D}\frac{\mathit{\partial }}{\mathit{\partial }\sigma }\left(K_m\frac{\mathit{\partial }v}{\mathit{\partial }\sigma }\right)+D{F}_y\end{array}$$

(4)

$$\frac{\mathit{\partial }TD}{\mathit{\partial }t}+\frac{\mathit{\partial }TuD}{\mathit{\partial }x}+\frac{\mathit{\partial }TvD}{\mathit{\partial }y}+\frac{\mathit{\partial }Tw}{\mathit{\partial }\sigma }=\frac{1}{D}\frac{\mathit{\partial }}{\mathit{\partial }\sigma }\left(K_h\frac{\mathit{\partial }T}{\mathit{\partial }\sigma }\right)+D\widehat{H}+D{F}_T$$

(5)

$$\frac{\mathit{\partial }SD}{\mathit{\partial }t}+\frac{\mathit{\partial }SuD}{\mathit{\partial }x}+\frac{\mathit{\partial }SvD}{\mathit{\partial }y}+\frac{\mathit{\partial }Sw}{\mathit{\partial }\sigma }=\frac{1}{D}\frac{\mathit{\partial }}{\mathit{\partial }\sigma }\left(K_h\frac{\mathit{\partial }S}{\mathit{\partial }\sigma }\right)+D{F}_S$$

(6)

$$\rho =\rho (T,S)$$

(7)

where u, v, and w represent the velocity components in the u, v, and w directions, respectively; T, S, and $\rho$ represent the temperature, salinity, and density, respectively; g represents the gravitational acceleration; f represents the Koch force parameter; $K_m$ and $K_h$ represent the vertical vortex viscous and vertical vortex viscous thermal diffusion coefficients; $F_x$ and $F_y$ represent the horizontal momentum terms; $F_T$ represents the horizontal temperature diffusion term; $F_S$ represents the horizontal salinity diffusion term; and $\widehat{H}$ represents the vertical gradient of short-wave radiation.

The boundary conditions are as follows:

The free surface boundary conditions in a free sea, $\sigma =0$:

$$\left(\frac{\mathit{\partial }u}{\mathit{\partial }\sigma },\frac{\mathit{\partial }v}{\mathit{\partial }\sigma }\right)=\frac{D}{{\rho }_0{K}_m}\left({\tau }_{sx},{\tau }_{sy}\right),\omega =\frac{\widehat{E}-\widehat{P}}{\rho }$$

(8)

$$\frac{\mathit{\partial }T}{\mathit{\partial }\sigma }=\frac{D}{\rho c_p{K}_h}\left[Q_n(x,y,t)-SW(x,y,0,t)\right]$$

(9)

$$\frac{\mathit{\partial }S}{\mathit{\partial }\sigma }=\frac{S(\widehat{P}-\widehat{E})D}{K_h\rho },q^{2}l=0,q^2=B_1^{2/3u_{\tau s}^2}$$

(10)

The seafloor boundary conditions on the seafloor boundary, $\sigma =-1$:

$$\left(\frac{\mathit{\partial }u}{\mathit{\partial }\sigma },\frac{\mathit{\partial }v}{\mathit{\partial }\sigma }\right)=\frac{D}{{\rho }_0{K}_m}\left({\tau }_{bx},{\tau }_{by}\right),\omega =\frac{{\Omega }_b}{\Omega }$$

(11)

$$\frac{\mathit{\partial }T}{\mathit{\partial }\sigma }=\frac{A_{H}D\tan \alpha }{K_h-A_H{\tan }^2\alpha }\frac{\mathit{\partial }T}{\mathit{\partial }n}$$

(12)

$$\frac{\mathit{\partial }S}{\mathit{\partial }\sigma }=\frac{A_{H}D\tan \alpha }{K_h-A_H{\tan }^2\alpha }\frac{\mathit{\partial }S}{\mathit{\partial }n},q^{2}l=0,q^2=B_1^{\frac{2}{3u_{\tau b}^2}}$$

(13)

In this formula, $A_H$ represents the horizontal diffusion coefficient; ${\tau }_{sx}$ and ${\tau }_{sy}$ represent the sea surface shear stresses in the x and y directions, respectively; ${\tau }_{bx}$ and ${\tau }_{by}$ represent submarine friction forces in the x and y directions, respectively; $u_{\tau s}$ and $u_{\tau b}$ represent the friction of the sea surface and floor on the boundary layer, respectively; $Q_n(x,y,t)$ represents the net heat flux at the sea surface; $SW(x,y,0,t)$ represents the short-wave radiative flux at the sea surface; $c_p$ represents the specific heat coefficient of seawater; and $q$ and $l$ represent the turbulent kinetic energy and turbulence scale, respectively.

Closed boundary conditions:

The closed boundary represents the land boundary, and the normal velocity and flux are 0.

$$v_n=0,\frac{\mathit{\partial }T}{\mathit{\partial }n}=0,\frac{\mathit{\partial }S}{\mathit{\partial }n}=0$$

(14)

Advection and diffusion are two major processes that determine the transport and dispersion of tritium in an ocean. Self-decay is responsible for removing the tritium dissolved in seawater. The self-decay process depends solely on the half-life of tritium, Comprehensively, the governing equation can be presented as

$$\frac{\mathit{\partial }C}{\mathit{\partial }t}=C_o(x,y,\sigma ,t)+F_c+\frac{1}{D^2}\frac{\mathit{\partial }}{\mathit{\partial }\sigma }\left(K_h\frac{\mathit{\partial }C}{\mathit{\partial }\sigma }\right)-\left(\frac{\mathit{\partial }uC}{\mathit{\partial }x}+\frac{\mathit{\partial }vC}{\mathit{\partial }y}+\frac{1}{D}\frac{\mathit{\partial }\omega C}{\mathit{\partial }\sigma }\right)-\lambda C$$

(15)

where $C$ represents the concentration of the tracer; $D$ represents the total water depth; u, v, and w represent the x, y, and $\sigma$ components of the water velocity; $K_h$ represents the vertical diffusion coefficient that is calculated using the selected turbulent closure scheme in the FVCOM; and the horizontal diffusivity in $F_c$ is calculated using the Smagorinsky eddy parameterization method in the FVCOM. $\lambda$ represents the radioactive decay constant, and $C_o$ represents the concentration injected from the source point.

Tritium, which is usually released in large quantities during nuclear accidents, is an isotope of hydrogen and exists in the offshore environment mainly in the form of water molecules, with a half-life of about 12.32 years; it has no particulate activity and is little affected by biogeochemical processes. Tritium is highly radioactive and is capable of penetrating deep into human tissues and releasing radiation. Long-term exposure to high concentrations of tritium increases the risk of leukaemia and other cancers.

2.2. Numerical Experiment Design

Lianyungang Haizhou Bay, a model harbour, is not only one of the most important bays in Jiangsu Province, it is also a pivotal port and the shipping heartland of China. The bay area harbours rich marine and fishery treasures. The computational domain in this study is defined as 119.18°–120.27° E and 34.4°–35.78° N; bathymetric information is derived from detailed nautical charts. This computational process uses the mean sea level and a spherical coordinate system, and a fine grid refinement is implemented for the coastline and islands around Haizhou Bay. The entire computational domain consists of 46,024 grid cells and 23,489 nodes, with 103 nodes at the open boundary. Vertically, the model is divided into four layers with the σ coordinate system. A dry and wet flooded grid discrimination mechanism is introduced during the model startup, with cold start processing. The turbulent eddy–viscosity coefficients in the horizontal direction are determined using the Smagorinsky parametric turbulence closure model, while the Mellor–Yamada level 2.5 turbulence closure model is used for calculation in the vertical direction.

In a different study [33], the TPXO8, TPXO_Yellow Sea 2010, and TPXO_China&Ind models exhibit a higher prediction accuracy compared to the MIKE Global Tide and TPXO7.2 models along our coast. In this study, the subtidal tidal harmonic constants generated by Chinatide are used as open boundary conditions. The selected simulation area is shown in Figure 1, while the calculated bathymetry distribution and grid division in the area are shown in Figure 2.

2.3. Initial Condition Setting

The hydrodynamic environment is simulated to provide a basic flow field for the study of radioactive material migration and diffusion. In order to simulate the diffusion of pollutants, it is assumed that the radioactive substances carried by the Fukushima nuclear sewage flow into Haizhou Bay in Lianyungang in China’s Yellow Sea from the outer sea; the radioactive substances begin to appear in the eastern sea area of the outer sea of Haizhou Bay, and the diffusion of the nuclides is mainly driven by the diffusion effect of the tides and monsoons to simulate the migration and diffusion of radionuclides. Assuming that the initial pollutant distribution field is zero, the attenuation effect of pollutants is considered in the calculation. The long-half-life nuclide tritium is the main object of study, so the pollutants that are predicted in this study are conservation-type pollutants.

To thoroughly examine the dispersion of radionuclides across Haizhou Bay’s subregions, our study delineates the bay into three primary areas: Lanshan, Lianyun, and Ganyu. A comprehensive delineation of these regions is depicted in Figure 3.

Initial condition: the initial radionuclide concentration field is 0.

Boundary conditions: on a closed boundary, the normal gradient of nuclides is 0. On an open boundary the nuclide concentration on the boundary as the water flows into and out of the calculation area is 1 Bq/L.

Drawing upon historical data, this study scrutinizes the dispersal patterns of nuclides throughout the summer and winter periods, delving into the complex interplay of various elements. To see an exhaustive array of the computational parameters, please consult

Table 1. Table of various combinations of risk conditions in Haizhou Bay.

Case	Time	Considerations	Continuous Pollutant Release (Activity Concentration, Bq/L)
1	summer	tides	1
2	winter	tides	1
3	summer	tides	1
		summer monsoon	1
4	winter	tides	1
		winter monsoon	1

.

Table 2. Settings of main parameters in simulation.

Hydrological/Radiological Parameters	Value
AH	1 m2s−1
Km	2.825 m2s−1
Kh	1.5 × 10−4 m2s−1
w	0 ms−1
$\lambda$	1.8 × 10−9 s−1

enumerates the parameters adopted during the computational process of the model. These parameters represent the synthesis of a literature review, expert assessments, and various model calibration outcomes, as referenced in other studies [34,35].

3. Hydrodynamic Simulation Results and Discussion

In this study, we simulated the hydrodynamic conditions in Haizhou Bay in Lianyungang from 15:00 on 17 October 2019 to 15:00 on 16 November 2019. The model’s fidelity was subsequently corroborated through comparison with the empirical data collected during the same time frame.

3.1. Model Description

In order to validate the hydrodynamic model, four stations (Lanshangang Station, Lianyungang Station, and Stations 1 and 2) within the sea area of Haizhou Bay were selected for calculations and comparison with the observed data. Dalian Ocean University conducted the on-site monitoring of the water level and flow-field changes in the near-sea area of Tianwan Nuclear Power Plant in October 2019 and provided monitoring data [36]. Based on the real-time meteorological data from 15:00 on 17 October to 15:00 on 16 November 2019, the water level changes for this complete tidal cycle were calculated, and the calculated water levels were compared with the measured water levels from Dalian Ocean University. The change patterns of the calculated and measured water levels are shown in Figure 4 and Figure 5.

Figure 4 and Figure 5 reveal that the variation in water levels within the modelled region over a month is minimal, exhibiting periodic fluctuations. Haizhou Bay experiences a semi-diurnal tidal pattern, characterized by two high and two low tides daily. The discrepancy between the predicted and observed water levels remains below 10 cm, accurately reflecting the tidal movements in the vicinity of the nuclear power plant. Furthermore, the maximum deviation is within 10%, underscoring the hydrodynamic model’s validity.

To further assess the reliability of the model simulation results, the number of tidal levels was analysed using the root-mean-square error (RMS).

$$RMS={\left\{\frac{1}{N}{\displaystyle \sum _{i=1}^N{(g_{\mathrm{mod}}-g_{obs})}^2}\right\}}^{\frac{1}{2}}$$

(16)

where $g_{\mathrm{mod}}$ represents the simulated value and $g_{obs}$ represents the observed value. The results are shown in

Table 3. Comparison between measured tidal level data and modelling results.

Tide Level	Maximum Error (m)	Minimum Error (m)	Average Error (m)	RMS
Lanshangang Station	0.35	0.01	0.13	0.02
Lianyungang Station	0.41	0.01	0.13	0.02

.

Table 3 clearly demonstrates that the root-mean-square errors for tide levels in Lanshangang Harbour and Lianyungang Harbour are uniformly 0.02, signifying a high degree of concordance between the model’s simulated values and actual observations, thereby affirming the model’s superior capability in replicating the tidal level dynamics.

3.2. Trend Validation

The tidal validation data were derived from in situ current measurements at two distinct stations within Jiangsu Province: Lianyungang Stations 1 (Figure 6A,B) and 2 (Figure 6C,D). The analysis entailed a comparative assessment between the vertically averaged empirical data and mid-layer computational outcomes from the model, with the results being depicted in Figure 6. A juxtaposition of the observed tidal data with the simulated findings is presented in

Table 4. Comparison between measured tidal current data and modelling results.

Currents	Maximum Error	Minimum Error	Average Error	RMS
Station 1	0.38 m/s	0.03 m/s	0.04 m/s	0.02 m/s
Station 1	30°	0.05°	3.4°	2.68°
Station 2	0.31 m/s	0.13 m/s	0.02 m/s	0.02 m/s
Station 2	13°	0.07°	8.8°	8.88°

, illustrating the degree of concordance between the empirical measurements and model predictions.

Table 4 reveals that the mean absolute discrepancy between the simulated and observed flow velocities is a mere 0.02 m/s, while the mean absolute divergences in flow direction are 2.68° and 8.88°, respectively. Overall, the computational outcomes for current velocity demonstrate consistency with the observed data, and the calculated flow directions broadly capture the authentic dynamics of tidal fluctuations. The primary source of deviation likely stems from the concentration of current observation points in the coastal zone, where the model accuracy may be compromised by less-precise bathymetric data and a coarser seabed topography. Consequently, the model’s results are deemed to be substantially reliable, offering a solid foundation for hydrodynamic conditions essential for downstream modelling.

4. Tritium Migration and Diffusion under Different Operating Conditions

Figure 7, Figure 8, Figure 9 and Figure 10 illustrate the temporal distribution of tritium activity concentrations over a 50-day simulation period, each under varying operational scenarios. Within these visual representations, the red zone delineates the migration and diffusion scope of tritium, whereas the blue zone demarcates the regions of the sea that remain uncontaminated.

4.1. Tidal Effects on Radionuclide Transport Diffusion

4.1.1. Radionuclide Tritium Diffusion Path Analysis

The simulated assessments reveal that upon scrutiny of cases 1 and 2, subsequent to its emanation from the source of leakage, radioisotope tritium initially disseminates westward along the harbour’s shoreline, propelled by the prevailing marine currents. Upon reaching the coast, it advances south-westward towards Lanshan District, where it engages in extended diffusion. In due course, it permeates Ganyu District and subsequently broadens its reach to Lianyun District. This dispersion route diverges from the anticipated trajectory commonly envisaged by the populace.

4.1.2. Analysis of the Transmission Range of Radionuclide Tritium

Based on the analysis of radionuclide propagation, the radioisotope tritium spread rapidly in a south-westerly direction during the initial three days after release under the action of tidal forces and the monsoon winds in scenarios 1 and 2. From the fourth to the tenth day, its propagation speed was significantly reduced, and by about the tenth day, it had spread throughout Arashiyama, affecting the harbour. From the tenth day to the fiftieth day, the spread of these radionuclides in the harbour was extremely slow, and the entire harbour area was fully contaminated by the fiftieth day. In other words, even after 40 days of continuous dispersion, the contaminant did not completely cover all the corners of the harbour.

The inadequate hydrodynamic conditions within the bay contribute to the sluggish renewal of the water column, leading to the swift dispersion of radionuclides in the outer regions, which markedly diminishes in the inner recesses. Consequently, the water flow at the bay’s entrance outpaces that of its interior region, fostering more vigorous activity at the mouth, while the inner flow slowly fades, impeding the effective dilution of pollutants. Hence, the radioactive contamination levels within the bay surpass those in the peripheral zones, amplifying the potential hazards to humans during fishing.

Upon meticulous comparison of the simulation outcomes for cases 1 and 2, in-depth analysis of the radionuclide dispersion dynamics and concentration trajectories within Hazhou Bay’s maritime domain across the summer and winter periods was conducted, juxtaposing Figure 7a–c against Figure 8a–c. It is apparent that during the summer months, the initial westward proliferation of contaminants occurs at a markedly accelerated pace relative to that during the winter season. Scrutinizing Figure 7d–f alongside Figure 8d–f indicates that within ten days the pollutants arrive at the Arashiyama region sooner during the winter than they do in the summertime. Further examination of Figure 7g,h in relation to Figure 8g,h unequivocally demonstrates that after a fifty-day period post release, the pollution levels in the Ganyu and Lianyungang areas are higher in winter when contrasted with their summer counterparts.

In the comprehensive analysis of seasonal radionuclide dispersion patterns, the overall trajectory during the winter is congruent with that in the summer. However, the initial retardation of contaminant propagation is discernible in early winter, followed by the acceleration of diffusion that culminates in a widespread distribution throughout Haizhou Bay within around 50 days. Notably, comparative assessments reveal that radioactive materials infiltrate the bay earlier during the winter, leading to a more pronounced contamination when juxtaposed with the summer dispersion events.

4.1.3. Characterization of the Distribution of Highly Concentrated Nuclides and Analysis of Diffusion Rates

Upon monitoring the dispersion of tritium, it was observed that substantial levels were present not only at the source but also in the southwestern maritime region within seven hours post-release. After three days, elevated concentrations of tritium emerged adjacent to the coastline west of the point of origin. By the tenth day, extensive areas of Hazhou Bay’s outer waters exhibited high levels of tritium. Despite radionuclide penetration into the port and bay, the zones with the highest concentration from the initial release to day fifty had not yet reached the inner harbours of Lanshan, Ganyu, or Lianyungang.

The examination of diffusion rates reveals that the radionuclide levels persist at elevated intensities for ten days subsequent to the initial discharge, with diffusion predominating over dilution. Subsequent to this interval, as the nuclides migrate towards the harbour and near the coastline, a marked decline in concentrations ensues. From the thirty-first to the fiftieth day, the pace at which the concentrations disperse decelerates considerably, with peak concentrations consistently appearing in the harbour’s outer waters.

4.2. Monsoon Effects on Radionuclide Transport and Dispersion

Upon examining the impact of the monsoon by juxtaposing and scrutinizing Figure 7a–c with Figure 9a–c, it is evident that the radioisotope tritium is initially profoundly influenced by monsoonal dynamics, facilitating its swift westward dispersion. Observing Figure 7d–f alongside Figure 9d–f, the diffusion pattern of radioisotopes appears to be predominantly governed by tidal forces over time, particularly upon their encroachment into the bay region. A comparison of Figure 7g,h with Figure 9g,h discloses the summer monsoon’s suppressive role on the migration and diffusion of radioisotopes. Similarly, an analysis of Figure 8g,h and Figure 10g,h indicates that the winter monsoon likewise impedes the spread of these radioisotopes.

The enduring presence of tritium in the harbour is alarming, as its prolonged retention impedes the natural dispersion by external currents. This stagnation markedly increases the absorption of radioactive materials by marine life, leading to pronounced bioconcentration, with profound implications for the ecosystem. Consequently, the vigilant monitoring of aquatic environments and organisms proximate to national coastal harbours must be a top priority.

All four scenarios showed agreement that the tritium-contaminated water would reach the coast of Haizhou Bay in Lianyungang within 50 days (see Figure 11).

5. Conclusions

Taking Haizhou Bay in Lianyungang as the research object, a radionuclide migration diffusion model was established by considering the decay and free diffusion of radionuclides. The model results are highly consistent with the observed flow within Haizhou Bay. Building upon this foundation, we conducted simulations to trace the trajectories and spatiotemporal variation in tritium under various operational scenarios. Moreover, we discussed the effect of tidal dynamics and monsoonal forces on the dispersal characteristics of tritium. This research culminated in the following insights:

• Under the influence of tides and monsoon winds, the radionuclides continue to diffuse into the harbour over time. The hydrodynamic effect in Haizhou Bay is weak, the water exchange rate is slow, and the diffusion of radionuclides in the bay is slow, while that outside the harbour is fast. The water body at the mouth of the bay is more active, and the flow rate inside the bay is weak, which is not conducive to the dilution of pollutants.
• The concentration distribution of the radionuclide tritium is affected by the flow field, which is mainly influenced by the tides. Due to the tides, radionuclide contamination is more severe in winter than it is in summer.
• The simulation of the transport diffusion of nuclides shows that, in general, monsoons have a weak effect on the diffusion of radionuclides in Haizhou Bay. Monsoons promote the diffusion of radionuclides outside the harbour and inhibit their diffusion inside the harbour.