Construction of land-use change matrix using National Forest Inventory data and a moving average method

Emission factor and activity data are necessary to calculate the removal and emission of greenhouse gases following the scheme of land use, land-use change, and forestry (LULUCF). The land-use change matrix comprises basic data for calculating greenhouse gas statistics in the LULUCF sector. In this study, the applicability of the point-sampling method suggested in the IPCC guidelines was analyzed using fixed points of the National Forest Inventory (NFI). A moving average method was applied to cluster plots to develop a land-use change matrix, using data from the 5th (2006 ~ 2010) and 7th (2016 ~ 2020) NFIs to analyze land-use change. When analyzing the matrix based on the number of cluster plots used in Finland, which assigns high weight to the latest year, we found significant decrease in the number of cluster plots and areas corresponding to forest and cropland. The moving average method showed low uncertainty in forest land, cropland, and settlements and high uncertainty in grassland, wetlands, and land use change area. The relative standard error for land-use change among land-use categories ranged widely from 3.7-100%. Comparing forest land from 2010 to 2020 with various statistics through the calculated areas, the area of forest was about 1.0% lower. Applying the area calculation method of this study to recent data could enable detailed, annual predictions of future land-use change. As in the United States and Japan, it could also help with the preparation of the NIR and other statistical reports.