Gas flux investigations have attempted to explain the transfer of mass across the air-sea interface. In most cases, the gas flux has been parameterised in terms of the product of a thermodynamic constraint (i.e., the difference in concentration of a gas across the air-sea interface) and a kinetic constraint (i.e., the piston velocity, which acts as a type rate constant in a first order decay process wherein the gas achieves equilibrium across the air-sea interface), which depends on wind speed and temperature. This relationship has been calibrated using tracer-release experiments on the ocean, wave tank studies, and micrometeorological techniques. The ocean-going tracer-release experiments and the wave tank experiments deduce the gas flux term as the only unknown in a budget equation where the increase or decrease of the tracer in the aqueous and/or gas phase is otherwise well known. Piston velocity estimates have been obtained from seagoing investigations of 222Rn, CO2, 3He, and SF6, and from wave tank studies using a number of other gases whose source and sink terms have been monitored in a controlled environment.

A number of investigations have been made of small photochemical species to resolve measured upper ocean concentrations with measured sources, sinks, and mixing rates. The simplest budgets place the known sources and sinks in a homogeneously-mixed box with a volume determined by the depth of the upper oceanic boundary layer. Parameterisations for photochemical and biological source and sink terms, mixing, and surface outgassing are incorporated, and the model is stepped forward in time to produce estimates of upper ocean concentration. Oceanographic box models are only successful when the chemical lifetime of the tracer is longer than about one day; i.e., in cases where the dynamics of diurnal mixing in the upper boundary layer may be neglected. More sophisticated models for short-lived tracers have employed one dimensional mixing models to simulate the downward transport of mass during the course of the diurnal mixing cycle. These types of models have been successfully used to simulate the diurnal cycle of a number of photochemically-produced tracers including H2O2, CO, COS, CS2, and a series of alkenes. In addition to providing insight into the chemical/biological processes of the upper ocean, these types of investigations have also provided a means to evaluate the skill of the mixing models that have almost entirely been developed to explain the distribution of temperature and salinity without consideration of another tracer. More advanced models have started simpler vertical mixing schemes and incorporated ecosystem models and photochemical parameterisations to simulate the temporal and depth variations of a tracer such as dimethyl sulfide (DMS) whose budget is mainly controlled by biological factors.

Inverse models are a class of simulation that has mainly not been not been exploited in marine chemistry. Given an adequate parameterisation of the terms in a tracer budget, an inverse model can be used to find optimised values of the rate constants and resolve the modelled and measured concentrations. The calibration of the early gas flux parameterisations are examples of inverse models in that experiments were set up and systems observed in which all the free variables governing tracer concentration were known except for gas exchange. The calculation of piston velocity was then established as a linear problem as the flux value that matched the measured changing tracer concentrations. The use of this kind of inverse model for deducing air-sea gas exchange is only successful for tracers that are relatively inert (i.e., with chemical lifetimes exceeding one month). For more reactive tracers, air-sea gas exchange may be ignored, and an inverse model may be used to deduce the best values of chemical source/sink terms to resolve observed and modelled concentrations.