In this communication, we intend to present the sampling design of the French National Forest Inventory (NFI), a sixty years old statistical survey.

Although NFIs have a long experience in providing quantitative and reliable information on forest resources, their methods need a deep mutation in order to address new challenges and be more responsive.

In many European countries, forests cover a significant part of the land area, reaching 30 % in France. The diversity of climatic conditions over the French territory have shaped different type of forests, which are overall more diverse and heterogeneous than their European counterparts. These forests also resulted from long term interactions with human populations. In particular, French forests went through an important transition since the mid nineteenth century, when the industry shifted from wood to non-renewable resources (coal and oil). This resulted in a continuous expansion of forest particularly on abandoned agricultural lands. This expansion is still ongoing up to now and forests continue to gain approximately 100,000 ha per year.

Nevertheless, the pressure on forests is increasing very quickly. Because of climate change, forest stands have to face new threats like water shortage or pathogens, among others. In the meantime they are expected to mitigate climate and provide new renewable resources for a low-carbon economy. The expected changes will be quick or even abrupt and NFIs have to evolve to follow and measure accurately such non-stable systems.

The sampling of the French NFI was designed to cope with these new challenges. Numerous efforts have been made to build a flexible and responsive tool. An important part of the system’s properties lies in its spatio-temporal organization: unlike other NFIs it operates on a flexible, annually adaptable sampling covering the whole country. This sampling design is continuous in space and time and is organized in interpenetrating annual grids thanks to a square grid of 1 x1 km whose properties will be detailed in this communication.

Another corner stone for the future of inventories lies in their ability to use relevant external information in order to enhance the precision of estimates. The availability of new and dense spatial information at country and European level allows NFIs to provide unprecedented results on small scales, thank to statistical approaches like post-stratification or imputation through kNN techniques, among others. These approaches are not fully operational and require further research to be implemented as new statistical techniques for the NFI.