Background
The potential of forests in mitigation of climate change has created a need to develop and improve methods to estimate the carbon budgets of forests. Soil monitoring methods applicable on the large scale carbon inventories are lacking. National and international soil surveys are carried out in European countries, but none of them has been designed for reliable carbon inventory on forest land. Due to high costs of the regionally representative soil surveys and soil carbon monitoring, sampling efforts need to be effectively allocated within a monitoring programme.
In order to design an effective and a reliable method to monitor changes in soil carbon, two kinds of information are necessary. Information is needed, first, on the rate of the change in soil carbon, and, second, on its spatial variation. Recent advances in simulation models of soil carbon make it possible to estimate the rate of the change in soil carbon in forests with different stand, site and climatic conditions. When these estimates are combined with information on the spatial variation of soil carbon, it is possible to calculate the accuracy of estimates for soil carbon changes obtained using different sampling designs and sample sizes. Since the reliability of this analysis depends on the validity of the model simulations and the representatives of the estimates for the spatial variation, both of these need to be tested with empirical data.
