NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
The Ocean Carbon States Database: A Proof-of-Concept Application of Cluster Analysis in the Ocean Carbon CycleIn this paper, we present a database of the basic regimes of the carbon cycle in the ocean, the 'ocean carbon states', as obtained using a data mining/pattern recognition technique in observation-based as well as model data. The goal of this study is to establish a new data analysis methodology, test it and assess its utility in providing more insights into the regional and temporal variability of the marine carbon cycle. This is important as advanced data mining techniques are becoming widely used in climate and Earth sciences and in particular in studies of the global carbon cycle, where the interaction of physical and biogeochemical drivers confounds our ability to accurately describe, understand, and predict CO2 concentrations and their changes in the major planetary carbon reservoirs. In this proof-of-concept study, we focus on using well-understood data that are based on observations, as well as model results from the NASA Goddard Institute for Space Studies (GISS) climate model. Our analysis shows that ocean carbon states are associated with the subtropical-subpolar gyre during the colder months of the year and the tropics during the warmer season in the North Atlantic basin. Conversely, in the Southern Ocean, the ocean carbon states can be associated with the subtropical and Antarctic convergence zones in the warmer season and the coastal Antarctic divergence zone in the colder season. With respect to model evaluation, we find that the GISS model reproduces the cold and warm season regimes more skillfully in the North Atlantic than in the Southern Ocean and matches the observed seasonality better than the spatial distribution of the regimes. Finally, the ocean carbon states provide useful information in the model error attribution. Model air-sea CO2 flux biases in the North Atlantic stem from wind speed and salinity biases in the subpolar region and nutrient and wind speed biases in the subtropics and tropics. Nutrient biases are shown to be most important in the Southern Ocean flux bias.
Document ID
20180002175
Document Type
Reprint (Version printed in journal)
Authors
Latto, Rebecca (Columbia Univ. New York, NY, United States)
Romanou, Anastasia (Columbia Univ. New York, NY, United States)
Date Acquired
April 4, 2018
Publication Date
March 27, 2018
Publication Information
Publication: Earth System Science Data
Volume: 10
Issue: 1
ISSN: 1866-3508
Subject Category
Meteorology and Climatology
Oceanography
Report/Patent Number
GSFC-E-DAA-TN54528
Funding Number(s)
CONTRACT_GRANT: NNX14AB99A
CONTRACT_GRANT: NNX15AJ05A
Distribution Limits
Public
Copyright
Other
Keywords
carbon cycle
Oceans
data mining