Alton, PB (2020). Representativeness of global climate and vegetation by carbon-monitoring networks; implications for estimates of gross and net primary productivity at biome and global levels. AGRICULTURAL AND FOREST METEOROLOGY, 290, 108017.

One of the major uncertainties in estimating global Net Primary Productivity (NPP) and Gross Primary Productivity (GPP) is the ability of carbon-monitoring sites to represent the climate and canopy-density of global vegetation ("representativeness"). These sites are used for empirical upscaling and calibration of global landsurface models. The current study determines the representativeness of two important carbon-monitoring networks - FLUXNET2015 and the Ecosystem Model-Data Intercomparison (EMDI) - by calculating the euclidian distance in climate-canopy space between each global 0.5 degrees cell and all carbon-monitoring sites of the same biome or Plant Functional Type (PFT). Reliance on the single (most similar) site has been adopted in the past. A straightforward weighted upscaling, using inverse euclidian distance, identifies which PFTs contribute most to global primary productivity in the context of how well they are represented in carbon-monitoring networks. Some vegetation types, which are numerically well-represented within the network, are sampled at the 'wrong' latitude and in more temperate climes than their global distribution. This includes non-mediterranean needleleaf forest which is one of the main vegetation types contributing to global GPP and NPP. (Semi-)arid regions (mean annual precipitation < 400 mm yr(-1)) are undersampled as well as the sparse vegetation that tends to characterise them. These regions include the tundra and the northern half of the boreal forest where growth is disproportionately affected by climate change. We find a large spread in NPP and GPP recorded at sites of the same PFT (standard deviation is 56% mean). Consequently, our bootstrap error analysis indicates that a minimum of 50 climate-representative sites per PFT is required to quantify adequately (2% precision) the primary productivity of each global vegetation type. Selecting unchartered climate-canopy space for new sites appears to be more important than a simple increase in site numbers.