CO2 is released from the P/T layer of the deep ocean at 4500 meters called the Calcite or Carbonate Compensation Depth. Below this depth carbonates (i.e. CaCO3) of all sorts are 100% soluble and CO2 is sent into solution. Carbonate cannot persist below this depth.
When ocean temperatures rise, then the CO2 concentration then exceeds the saturation P & T and must be released.
This phenomenon has been known for 150 years or more. Historical CO2 TRAILS temperature! No need for any other debate on "historical" CO2. Having said that, I am in NO way implying that present CO2 levels are a result of this mechanism. I am just trying to point out the irrelevance of the relationship as somehow having "been seen before" as associated with warming trends.
The role of CO2 as a greenhouse gas (which it is) and the degree of warming (that it does cause) are much more complicated than NAS, NASA, IPCC and others would like you to believe. I have linked a person's calculations on the thermodynamics of CO2 just to illustrate the complexity. I do not know if the guy is right or wrong or is a hack or a genius. He does however attempt to address some of the very major problems that Greenhouse theory must address!