Abstract
Abstract. Given current threats to ocean ecosystem health, there is a growing demand for accurate biogeochemical hindcasts, nowcasts, and predictions. Provision of such products requires data assimilation, i.e., a comprehensive strategy for incorporating observations into biogeochemical models, but current data streams of biogeochemical observations are generally considered insufficient for the operational provision of such products. This study investigates to what degree the assimilation of satellite observations in combination with a priori model calibration by sparse BGC-Argo profiles can improve subsurface biogeochemical properties. The multivariate deterministic ensemble Kalman filter (DEnKF) has been implemented to assimilate physical and biological observations into a three-dimensional coupled physical–biogeochemical model, the biogeochemical component of which has been calibrated by BGC-Argo float data for the Gulf of Mexico. Specifically, observations of sea surface height, sea surface temperature, and surface chlorophyll were assimilated, and profiles of both physical and biological variables were updated based on the surface information. We assessed whether this leads to improved subsurface distributions, especially of biological properties, using observations from five BGC-Argo floats that were not assimilated. An alternative light parameterization that was tuned a priori using BGC-Argo observations was also applied to test the sensitivity of data assimilation impact on subsurface biological properties. Results show that assimilation of the satellite data improves model representation of major circulation features, which translate into improved three-dimensional distributions of temperature and salinity. The multivariate assimilation also improves the agreement of subsurface nitrate through its tight correlation with temperature, but the improvements in subsurface chlorophyll were modest initially due to suboptimal choices of the model's optical module. Repeating the assimilation run by using the alternative light parameterization greatly improved the subsurface distribution of chlorophyll. Therefore, even sparse BGC-Argo observations can provide substantial benefits for biogeochemical prediction by enabling a priori model tuning. Given that, so far, the abundance of BGC-Argo profiles in the Gulf of Mexico and elsewhere has been insufficient for sequential assimilation, updating 3D biological properties in a model that has been well calibrated is an intermediate step toward full assimilation of the new data types.
Highlights
Given the multiple and increasing pressures of ocean warming, acidification, deoxygenation, and changes in primary productivity on ocean ecosystem health, accurate model simulations are urgently needed to assess past and current states of marine ecosystems, forecast future trends, and predict the ocean’s response to different scenarios of climate change and management policies
By comparing forecast results from the assimilative model with independent observations from five BGC-Argo floats that are not assimilated but used in a priori tuning of the biogeochemical model, we rigorously evaluate whether the main biological observation stream in combination with physical observations can inform the 3D ocean distributions in high spatial and temporal resolution
A coupled data assimilation scheme for both physical and biological satellite observations was implemented to investigate whether these observations can inform subsurface distributions
Summary
Given the multiple and increasing pressures of ocean warming, acidification, deoxygenation, and changes in primary productivity on ocean ecosystem health, accurate model simulations are urgently needed to assess past and current states of marine ecosystems, forecast future trends, and predict the ocean’s response to different scenarios of climate change and management policies. B. Wang et al.: Can assimilation of satellite observations improve subsurface biological properties?. Cal approach used to compensate for these model deficiencies. It is a statistical method to interpolate and extrapolate sparse observations into the regular model space in a dynamically consistent way. While any practice to constrain a model by observations can be referred to as data assimilation, in this paper we refer to state estimation, i.e., sequential updates of the model state
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
