R. Peridfiez et al. 
Environmental Modelling and Software 122 (2019) 104523 
Release 
in air 
Alr . 
Contamination ” 
Immersion in 
Plume 
ap 
Dissolved 
| phase 
Uptake be 
biota 
Faod 
ingestion 
Release 
n ocean 
—' 
2 
Inhalatian 
A F 
| Dose to 
=> humans 
Particulated 
phase 
Uptake by 
bottom 
‚ediments 
t External 
exposure 
Fig. 1. Pathways for human exposure from releases of radioactivity into the ocean 
>nvironment (based on Hunt, 2004). The dotted rectangle identifies processes under 
:onsideration in this review. 
nuclear facilities, such as La Hague nuclear fuel reprocessing plant in 
the English Channel (Breton and Salomon, 1995) or from nuclear waste 
dumped in the sea (Harms, 1997; Cetina et al,, 2000). Marine transport 
models attracted more attention after the accident in Fukushima Daiichi 
Nuclear Power Plant (FDNPP) resulting from the March 2011 earth- 
Juake and tsunami (e.g, Kawamura et al., 2011; Behrens et al,, 2012; 
Tsumune et al., 2012; Dvorzhak et al., 2012; Masumoto et al., 2012; 
5CJ, 2014, etc.). However, it should be commented that last reviews 
>f marine radionuclide transport models were published more than a 
decade ago (Harms et al., 2003; Periänez, 2005a), 
The International Atomic Energy Agency (IAEA) has organized ac- 
Ävities on marine radioactivity transport model testing since the VAMP 
(Validation of Model Predietions) program in 1988 (see IAEA, 2000, 
for the aquatic group work). The most recent programmes are the 
MODARIA? (Modelling and Data for Radiological Impact Assessments) 
project, launched in 2012, and MODARIA-IL,} launched in 2016. The 
narine working groups in these programs were motivated by the recent 
developments in marine science and marine modelling, as well as the 
‘adioactive pollution due to the FDNPP disaster. In general, TAEA model 
intercomparisons are also motivated by the need to link modelling and 
data with radiological impact assessments, to enhance the capabilities 
of member states to simulate radionuclide transport and also to un- 
lerstand their effects in the environment. State-of-the-art models were 
applied by different teams to several radionuclide transport problems 
and results were compared , with special emphasis on the sources of 
discrepancies between models. In addition to what was said above, it 
may be useful to have a summary of the basic Principles of such models, 
which include all processes within the dashed square in Fig. 1 and are 
‘he most advanced radionuclide transport models at the present time. 
The purpose of this paper, in view of our previous comments, 
zonsists of addressing three main Points. These points correspond to 
he main sections of the paper: 
1. To provide an updated review of marine radionuclide transport 
modelling techniques, including a brief description of model 
(ypes (box models, Eulerian and Lagrangian models). Different 
model structures, techniques presently used to obtain water 
circulation and to describe other processes like water/sediment 
interactions and biological uptake are described in Section 3. 
? http://www-ns.iaea.org/projects/modaria/default.asp?l=1 16. 
; http://www-ns.iaea.org/projects/modaria/modaria2.asp?s=88&1=120, 
First, the main radionuclide transport processes in the sea are 
discussed in Section 2. 
To briefly review modelling works which were carried out to 
simulate FDNPP releases in the Pacific Ocean since, as men- 
tioned above, the most recent modelling efforts were devoted 
to this task. This review in presented in Section 4. 
To discuss the main difficulties and problems in marine ra- 
dionuclide transport models found during the MODARIA and 
MODARIA-IT programs, with special emphasis on sources of 
model uncertainties. Challenges in developing models for emer- 
gency situations (as FDNPP accident for instance) are also dis- 
cussed. All this is presented in Section 5. 
2 
3 
2. Radionuclide transport processes in the marine environment 
Radionuclides are considered to be conservative or NOon-conservative 
according to their geochemical behaviour. Some radionuclides remain 
dissolved in the water column since adsorption by the solid phases 
(suspended and bed sediment particles) is negligible. These are the 
3o-called conservative radionuclides. Radionuclides which are signifi- 
cantly adsorbed by sediment particles (suspended in the water column 
and on the seabed) are denoted as non-conservative. 
In general, the sea is a stratified environment where the pycnocline 
usually acts as a barrier for vertical mixing, Radionuclides may be 
deposited on the sea surface due to atmospheric fallout (dry and/or 
wet deposition) or directly introduced into the sea from a point source 
due to releases from a given industrial facility (chronic or accidental), 
an accident in a nuclear vessel etc, Other sources of radionuclides in 
the sea are river runoff, distributed influx due to the coastal wash- 
off and undersea groundwater discharges (Sanial et al., 2017). Once 
in the water column, radionuclides are transported by the currents 
(advection) and also mixed by turbulent diffusion. Diffusion produces 
a transport from high to low concentration regions, which makes 
peak concentrations decrease as the radionuclide patch size increases, 
Radionuclides may be fixed to suspended matter particles present in 
he water column, which sink by gravity (settling) if their densities 
are higher than water density and reach the deep layer. Advection 
and diffusion processes also occur in the deep layer. Suspended par- 
ticles transporting radionuclides may be deposited on the seabed. Also, 
direct adsorption/desorption reactions between the seabed and water 
may occur. Radionuclides introduced into the seabed sediment are 
eventually buried by the further deposition of suspended particles or re- 
suspended by near bed turbulence., Eventually, there will be a transfer 
of radionuclides to biota, Other Processes, as sea spray transfer from 
sea to land and evaporation of gaseous radionuclides may occur as well 
(Vives i Batlle et al., 2018). 
Plutonium and thorium isotopes are examples of non-conservative 
radionuclides, Radionuclides which are considered to be conservative 
are, for instance, “Sr, 125Sb, 9”Tc and 12971, It is quite usual to find mod- 
elling studies in which !?7Cs is treated as a conservative radionuclide 
(Prandle, 1984; Estournel et al., 2012; Behrens et al., 2012; Tsumune 
at al., 2012, among many others). Actually!97Cs is measured in seabed 
sediments, thus this is just an approximation in which adsorption- 
desorption processes are neglected, These processes were considered 
for !97Cs in Abril and Garcta-Leön (1993), Margvelashvily et al. (1997), 
Aldridge (1998), Aldridge et al. (2003) and Goshawk et al. (2003) 
among others. 
All these processes, which are represented in Fig. 2, can be described 
by means of appropriate differential equations, A computer code can 
then be written to numerically solve the equations, which constitutes 
the transport model. The following ingredients are required to sim- 
ulate the transport of non-conservative radionuclides in the marine 
environment: