[A-OS12] Ocean Science Revealed by Global Observation System

[AOS12-P01] Generating analysis-ready BGC-Argo datasets using Jupyter Notebook

Haruto Fujishima^1,2,3, *Hakase Hayashida¹ (1.Japan Agency for Marine-Earth Science and Technology, 2.Tokyo University of Marine Science and Technology, 3.Tohoku University)

Keywords:BGC-Argo, ocean biogeochemistry, ocean ecosystem health, profiling floats, ocean observing system, Python

Biogeochemical Argo (BGC-Argo) is a global array of autonomous profiling floats that monitor marine ecosystem health in the upper 2,000 meters of the water column on a weekly basis. There are six key variables measured by BGC-Argo floats: chlorophyll-a, pH, oxygen, nitrate, irradiance, and suspended particles. Profiles of these variables are made publicly available as netCDF files on a near real-time basis. While these profiles may seem easy to use, they often contain erroneous and suspicious values due to poor sensor calibration and high sensitivity to noises and artifacts. Furthermore, the profiles may require bias corrections and derivations of desired variables using empirical equations. As such, the original data require post-processing before they can be used for scientific analysis. Here we develop such a post-processing tool that generates “analysis-ready” time series of biogeochemical profiles from a selected float. This tool is written in Python and uses Jupyter Notebook. Users can produce, visualize, and analyze the time series of their interest based on the time, geographic region, and biogeochemical variables, without having to download the original data. The output can be saved as a netCDF file for potential application such as model evaluation and data assimilation. By construction, our tool is generic, designed to mitigate known issues that are most common across the float profiles in the original data. We discuss the pros and cons of our approach and further development ideas for prospective end-users.