An assessment of forest biomass maps in Europe using harmonized national statistics and inventory plots

Abstract

Maps of aboveground forest biomass based on different input data and modelling approaches have been recently produced for Europe, opening up the possibility for several applications and products not obtainable by summary statistics. However, the accuracy assessment of the existing maps is limited by the lack of reference data consistent over the study region and representative of the maps cells. Here, we used harmonized forest biomass data for 26 European countries derived by National Forest Inventories using a common biomass definition and estimator to assess four biomass maps. The assessment was performed at regional, national and sub-national scales using harmonized statistics derived from almost half million ground plot measurements, and at pixel level using a subset of 22,166 plots covering most European forest types. The field plots were temporally aligned with the maps using growth rates and further screened using an innovative approach based on tree cover variability to remove the plots not representative of the map cells. The harmonized reference data showed that all maps tended to overestimate at low biomass (<100 Mg ha^-1) and underestimate at medium - high biomass (>100 Mg ha^-1), resulting in an overall negative bias (23-43 Mg ha^-1 at national level) relative to the harmonized estimates. The maps relative errors ranged from 29% to 40% at national level and increased at higher resolutions, reaching 58-67% at pixel level. We also assessed the effect of the harmonization of the national statistics and report that the harmonized biomass values present significant differences compared to the national estimates for 14 countries, and provide a slightly higher stock (+3.8%) at European scale. We show that harmonized and representative reference data are essential to properly assess the accuracy of biomass maps, and we further identify the factors affecting the maps performance and provide indications for their improvements.

Keywords: Aboveground biomass; Carbon cycle; Forest inventory; Forest plot; NFI; Remote sensing.

Fig. 1

Spatial distribution of the harmonized EFDAC biomass plots for the 26 countries included in the EFDAC dataset.

Fig. 2

The four biomass maps for Europe: Thurner, Barredo, Kindermann and Gallaun (clockwise, from upper left).

Fig. 3

Comparison of the biomass stocks estimated by the biomass maps with the EFDAC statistics at national (left) and sub-national (right) levels.

Fig. 4

Comparison of the mean biomass densities estimated by the biomass maps and the EFDAC statistics at national (left) and sub-national (right) levels.

Fig. 5

Comparison of the biomass maps with the selected EFDAC plot dataset. The red lines represent the linear regressions between the two datasets and the black lines represent the 1:1 line. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)