Chapter 10 – The biogeochemical component of NEMO TOP interface for tracers and biogeochemistry

10.1 Scope

TOP (Tracers in the Ocean Paradigm) is the NEMO interface that provides the physical constraints and boundaries conditions for oceanic passive tracers transport. It also handles a seamless, hardwired coupling with biogeochemical (BGC) models.

In this chapter the NEMO-TOP evolution is addressed by considering the need i) to improve the orthogonality between physical processes and oceanic tracers dynamics, ii) to consolidate the modularity of the interface sub-components and iii) to foster the code readiness to handle future evolution of biogeochemical models (e.g., interactive benthic and sea-ice biogeochemical models, increased number of pelagic tracers).

The description of overlapping issues between TOP and other NEMO components is detailed within the specific chapters. In a nutshell, these will focus on the efficient implementation of computational algorithms (Chapter 3), a flexible method to adapt grid resolution over the oceanic domain (Chapter 6), and the interface with the forthcoming sea-ice model (Chapter 8).

10.2 Main issues

10.2.1 Tracers vertical sinking

Despite the large variety of advection and diffusion schemes made available through the OPA engine, the current NEMO code is still lacking a general scheme to deal with the vertical sinking of tracers in the pelagic environment. This is a key physical process that directly affects the spatio-temporal distribution of particulate suspended matter and the dynamics of phytoplanktonic groups, like e.g. diatoms. Among the large ensemble of approaches available from literature (Yool et al., 2013; Aumont et al, 2015; Vichi et al., 2015; Buthenschön et al., 2016) it is here necessary to identify the most suitable and up-to- date scheme to be included in TOP that will ensure an adequate balance between accuracy and computational efficiency. Furthermore, current schemes are mainly derived from typical advection schemes that were developed to handle transport by fluids. The potentially very high sinking speeds of particles (which can reach hundreds of meters per day) puts very high constraints on the timestep to respect the CFL criterion, especially where the vertical resolution is very high. To circumvent this constraint, time-splitting methods have been adopted by some biogeochemical models embedded in NEMO (Aumont et al., 2015). Yet, this can become every CPU intensive. Alternative solutions have been proposed in other modeling systems such as CROCO. These solutions will be investigated with care and if satisfying, the most appropriate one will be implemented in TOP.

10.2.2 Vertical light penetration scheme

Vertical penetration of visible light depends on the optical properties of seawater. It is mainly controlled by the content in pigments (i.e., chlorophyll), colored dissolved substances and particles. Computation of the vertical attenuation of light is required for the upper heat budget of the ocean and by various biogeochemical processes, the most well known being photosynthesis by phytoplankton (but all photochemical processes depend on the amount of light). Currently, in NEMO, modules performing this computation are developed separately for each biogeochemical model as well as in the dynamical core of NEMO (Lengaigne et al., 2007; Hernandez et al., 2017). The development of a standalone module to compute the vertical distribution of shortwave radiation (including visible light) would help rationalizing the treatment of light and would avoid the existence of multiple, sometimes redundant, modules. Furthermore, it will be useful for scientists interested in modeling photochemical processes in new biogeochemical schemes.

10.2.3 Carbonate System module (CSM)

The pelagic carbonate system has become a pillar of ocean sciences spreading from the biogeochemistry up to the climate studies through Earth System Models. In the last decade a strong effort was spent in the consolidation of scientific and modelling knowledges (Zeebe and Wolf-Gladrow, 2001; Dickson 2007; Orr et al., 2015) and, nowadays, a generalized and common approach for carbonate system simulation is well established (Orr et al., 2016). The inclusion of a state-of- the-art standalone module for the pelagic carbonate system in NEMO-TOP represents a community-based evolution of the interface and it will offer an easy-to- use tool for ocean scientists who want to deal with the CO2 problem.

10.2.4 Air-sea gas exchange module

At present, the different modules of TOP interface dealing with both inert chemistry components (e.g., CFCs, SF6) and biogeochemistry (mainly oxygen and CO2) use internally defined routines to solve the air-sea gas exchange. Given the wide variety of inert tracers used to perform the tracking of water masses and deep ocean ventilation studies, the development of a common air-sea gas transfer scheme (see e.g., Orr et al., 2016) will generalise the workflow of TOP existing modules, by ensuring also an easier maintenance of the code. In addition, this will simplify the inclusion of new inert chemistry tracers and the coupling with biogeochemical models.

10.2.5 Implementation of new BGC components

The overall structure of TOP is focused on the marine pelagic component (arrays for state variables, time integration, etc.) and only few elements are available to enable the treatment of additional dynamical components. Given the emerging interest in modelling both benthic (Capet et al. 2013) and sea-ice (Tedesco and Vichi, 2014) systems, the implementation of a new set of generalized routines to facilitate the coupling with complex biogeochemical models is a strategic issue. This will be achieved by following the modular layout available for the pelagic system, by including for these components the definition of 2D/3D state variables, time integration schemes, and I/O handling.

10.2.6 Implementation within PISCES biogeochemical model

Since PISCES is the marine biogeochemical model embedded in NEMO, this model will serve as a basis to test and validate the developments described in this chapter as well as in others chapters (see section 10.1). This task will require to modify the architecture of PISCES to redefine its perimeter : Currently, the modules that are planned to become part of NEMO-TOP are embedded within PISCES. Furthermore, Task 10.2.5 plans to implement a set of generalized routines to couple new BGC components with NEMO. This includes benthic models. Such a benthic model is already existing within PISCES and will be used as a basis to develop and validate the generalized routines. As for the oceanic part of PISCES, this activity will require to redefine the architecture of the sediment module of PISCES and its interfaces within NEMO.