Dataset to article entitled "Large-scale validation of forest attribute maps across different spatial resolutions"

The dataset consists of input to reproduce the results of the article, "Large-scale validation of forest attribute maps across different spatial resolutions." We applied the regression models to predict biomass, volume, basal area, and Lorey’s height at the different spatial resolutions for the pixels covering the validation stands. We then estimated forest attributes for the validation stands by calculating the mean of the predicted forest attributes at stand-level. Pixel predictions were weighted according to the proportion by which they covered the plots to account for the fact that not all pixels fully fall within them. The observed and predicted values on plot-level were then averaged on stand level. This process resulted in a dataset with (synthetic) estimates and observed values of forest attributes based on different pixel sizes which was used in the uncertainity assessment. This repository consist the code (Analysis_clean.R) used to produce the results in the article and the input dataset (Observed_vs_predicted_clean_input.csv) Variables: FID: Stand ID Obs_vol: observed volume (m3ha-1) Obs_lh: observed Lorey’s height (m) Obs_ba: Observed basal area (m2ha-1) Obs_bm: Observed biomass (Mg ha-1) Pred_volmb: Predicted volume (m3ha-1) Pred_lh: Predicted Lorey’s height (m) Pred_ba: Predicted basal area (m2ha-1) Pred_biom:Predicted biomass (Mg ha-1) Res_class: Resolution class (1,5,10,16,30 m) Area_ha: area of stand [JB1] in hectare (Ha) Dom_sp: Dominant tree species in Norwegian either spruce (GRAN), pine (FURU) or boardleaf (BJERK) The field observation dataset has been collected between 01.01.2018-31.12.2022.