Arumäe, T.* and Lang, M. 2016. A Validation of Coarse Scale Global Vegetation Height Map for Biomass Estimation in Hemiboreal Forests in Estonia. Baltic Forestry 22(2): 275-282.

   A public release of global vegetation height map, based on data from spaceborne lidar GLAS, was validated using forest management inventory (FI) data and airborne laser (ALS) data from two 15 × 15 km test sites in Estonia: the first one in Aegviidu and the second one in Laeva. For each global vegetation height (GVH) map pixel located in the test sites we calculated forest height based on the FI data and on ALS data. Linear regression analysis was then used to evaluate the relationships between GVH map values (HGVH), FI forest height (HFI) and ALS-based Lorey’s forest height (HALS). In the second test HGVH and HALS were evaluated for estimating forest biomass using regression analysis.

   The biomass was calculated for each GVH pixel using FI data and allometric regression models. The correlation between HGVH and HFI or HALS in both test sites was weak – in Aegviidu r < 0.25 and in Laeva r < 0.15; and, the relationship was not statistically significant in Laeva. The airborne lidar based HALS had a strong positive correlation with forest biomass and the determination coefficient of linear regression was R2 > 0.6 (p < 0.01) in both test sites. The relationship between HGVH and biomass was scattered and determination coefficient for linear model was small (R2 < 0.15, p < 0.01). Although in this study only weak correlation between measured forest heights (HFI and HALS) and spaceborne lidar based HGVH was found, the GVH type estimates are essential for the areas, where forest inventory data or airborne lidar data is not available. The obtained results show that forest height estimates from ALS or spaceborne lidar could be used directly for estimating biomass in managed hemiboreal forests at coarse spatial resolution.

Keywords: hemiboreal forests, airborne laser data, forest inventory data, global vegetation height map, forest biomass.