The continued increase in atmospheric carbon dioxide (CO₂) due to anthropogenic emissions (mostly by fossil burning and land use and land cover change) is believed to have caused significant changes in global climate and biology. Because terrestrial ecosystems play an important but ambiguous role in the global carbon budget, a prime focus of much research in the global change studies is on the carbon accounting of the terrestrial ecosystems. Despite carbon sequestration by terrestrial ecosystems has been thought to occur predominantly in the Northern Hemisphere, the magnitude, longitudinal distribution and dynamics of carbon sinks remain uncertain. To fully balance carbon budget, uncertainties in estimating carbon sinks / sources must be quantified that include soil respiration variations with temperature, terrestrial responses to climate change, major disturbances in forest ecosystems, and total amount of sequestrations by the global oceans and terrestrial ecosystems. Most of these uncertainties mainly result from the complexity of terrestrial ecosystems. To reduce uncertainties in carbon budget estimation, many methods have been developed in the past several decades. However, these methods may not necessarily result in uncertainties reduction, instead, they often constrain the generality of carbon estimates. The purpose of this study is to provide comprehensive review of current methods, specifically focusing on terrestrial ecosystems.
Eight methods commonly used in estimating terrestrial ecosystem carbon budget were reviewed with respects to their merits and limitations that include 1) the whole-plant or soil chamber method, 2) forest inventory, 3) the eddy covariance techniques, 4) plant-soil C isotopes techniques, 5) land use and land cover change, 6) process-based modeling, 7) atmospheric tracer-transport modeling, and 8) remote sensing techniques. These methods were divided into two groups: bottom-up (or land-based) and top-down (or atmosphere-based). Generally, the bottom-up approaches are more appropriate in understanding and predicting changes in carbon balance of terrestrial ecosystems. However, they are difficult to be scaled up to represent the regional or global terrestrial ecosystems. The top-down approaches, similarly, have the advantage of estimating carbon budget over large spatial areas, but provide little information about the processes. As each approach has its own limitations, results from each study often vary significantly and difficult to compare. It is suggested that a study be made to compare and analyze the results from different methods using a single dataset, which will provide some quantitative assessment of uncertainties from each method.