Atmosphere-ocean balances and fluxes of gases are fundamental components of the global biogeochemical cycles. The specific balances and fluxes of greenhouse gases and aerosols are fundamental for Earth-System models forecasting climatic change and its impact on the biosphere. The Global Ocean acts as a net sink of CO2 from the atmosphere. However, the Coastal Ocean dynamics is much intricate and variable.  Coastal zones can either uptake CO2 form the atmosphere or outgas it, depending on processes as continental loads, upwelling, plancton productivity, the metabolism of coastal benthic ecosystems and changes in solubility. Contrasting with the CO2 dynamics, the Coastal Ocean acts as source of CH4, N2O and DMS to the atmosphere. Besides the balances/unbalances between atmospheric and oceanic concentrations, the rates at which gases cross the air-water interface are also extraordinarily variable. Although they are basically mediated by turbulence at the sea-surface, the drivers of this turbulence and its intensity are highly heterogenic. Wind drag is the dominant source of turbulence at the Global Ocean's surface, where other factors are neglected by current Earth System modelling. However, other factors have been demonstrated to also be essential for accurate estimates of gas transfer velocities across the Coastal Oceans' surface, namely the atmospheric stability, sea-surface roughness, fetch, rain, currents and the presence of surfactants.

The scientific community has produced many formulations for the estimation of the solubilities and transfer velocities of gases. Most of the formulations for transfer velocity are either rough generalizations or specific to particular conditions. However, both Earth-System and regional modelling need to overcome this limitation and ideally use an algorithm capable of providing accurate estimates in any type of environment. It was with this intent that the FuGas framework was started in 2010 and endures permanent upgrades from the state-of-the-art. In this framework the user can choose among tens of formulations from the bulk literature and compare their output. Different degrees of complexity can be used and the software automatically makes use of simpler solutions whenever/wherever the lack of data does not allow for the application of the user's predefined assemble. The software also allows to compare among alternative situations and quantify how much each input variable was responsible for the change in the output. This is achieved using a Taylor expansion of the model, that is adapted to the user's assemble estimating the partial derivatives from a multivariate version of Newton's Finite Difference formula. Its application was demonstrated in the works by Vieira et al. (2013, 2015a,b, 2016), as well as at the EGU 2016 General Assembly, ESA's 2016 Living Planet Symposium and OMICS 5thInternational Conference on Earth Science and Climate Change.