PMEL CO2: Carbon Synthesis

Return to the Carbon Cycle Studies Page

The statistical detection of climate change signals in the atmosphere, and the attribution of those signals to independent forcing mechanisms, is well established in the climate modeling community. Detection and attribution studies using oceanic observations are less common and tend to focus on physical variables involving the heat budget. The application of these techniques to ocean biogeochemistry is even less common, due both to the relative scarcity of oceanic observations, and to the fact that the complex fully-coupled models required to undertake this task have just recently become available. As we move into the future, the continued accumulation of data and expected improvements in earth system modeling will act together to improve our ability to undertake detection and attribution studies. We are developing a new program to facilitate the development of this field of study and begin to work on the detection of ocean biogeochemical responses to climate change. This will give us the opportunity to not only to do pioneering work in this area, but also to provide guidance for future studies and observational programs. Our particular goals are to:

characterize the impact of global warming on the oceanic uptake of anthropogenic CO₂ and therefore future atmosphere-ocean partitioning of CO₂;
develop the tools and synthesize the data required to detect climate change, i.e., to separate the impact of global warming on the ocean carbon cycle from natural variability;
develop strategies for future observations, analyses, and improvement of models in order to improve and optimize future detection and attribution efforts.

In order to undertake a detection study, three main elements are required: (i) An observational data base, (ii) a firm estimate of natural processes in order to establish the patterns and magnitude of background variability, and (iii) an estimate of the signatures associated with climate change. We are initially focusing on the first element of this process.
We divide our data analysis activities into two categories depending on the frequency of the variability that can be examined. High resolution ocean time-series and underway pCO2 measurements can yield a reasonable description of seasonal to interannual surface ocean variability. These data will be used primarily for assessing the “natural” climate variability. Data from oceanographic sections that have been reoccupied will be used to examine lower frequency variations that are a very important component of the detection studies.

Analysis of seasonal to interannual variability
To characterize high frequency (seasonal to inter-annual) natural carbon system variability we are focusing on the air-sea difference in CO₂ partial pressure (DpCO₂). The available database has extensive temporal and spatial coverage and provides a direct assessment of oceanic CO₂ uptake. To date, little work has been done to assess the variability of the observations. We are examining the DpCO₂ variations on seasonal and annual time scales. The required data density will be achieved by minimal binning and de-trending using the methodology developed by T. Takahashi (LDEO). We are also working with various groups attempting to combine satellite data with carbon observations to try and develop a longer record of CO₂ variations with better spatial coverage. Time series data will be used to assess the sample size and signal amplitude required to define seasonal to interannual changes. Interannual variability will be more difficult to determine, but will be addressed with a combination of time series stations (e.g. HOT, BATS), ship-of-opportunity observations (e.g. Drake Passage), satellite proxies, moorings (e.g. TAO/TRITON), and hydrographic cruise data (e.g. WOCE, JGOFS).

Analysis of interannual to decadal variability
By looking at data in the main thermocline, weaker and longer term secular trends and cycles can be analyzed. This approach takes advantage of the fact that the residence times of waters and hence biogeochemical properties increase with depth and that flow in the ocean interior is primarily along constant potential density surfaces. Biogeochemical variations along isopycnal surfaces may therefore reflect past variations. This approach implicitly focus’ on inter-annual to decadal variability because the higher frequency “noise” has been smoothed out.
We are working with the Repeat Hydrography CO₂ Science Team to synthesize the quality controlled data sets, including data from the CLIVAR/CO₂ Repeat Hydrography Program cruises, to be used for examining changes along isopycnal surfaces. We are investigating changes in oxygen, nutrients, dissolved inorganic carbon, total alkalinity and anthropogenic CO₂ over time and space. We will examine oxygen utilization rates and corresponding organic carbon remineralization rates using appropriate stoichiometric C:O₂ ratios and corrected CFC age data.

This work is seen as the first step in a larger collaborative effort to conduct formal detection and attribution of biogeochemical changes in the global oceans.

Return to the Carbon Cycle Studies Page