Impacto do uso do filtro “velocidade de fricção” em estimativas anuais de carbono sobre ecossistemas de pastagem natural

Annual carbon estimation of the most diverse ecosystems is a recurring theme in meetings that address climate change mitigation, as it is essential to have a realistic inventory of carbon stock in the biosphere and the ability to assimilate atmospheric carbon. Measurements of CO2 flux over ecosystems after being taken undergo rigorous post-processing to remove spurious and unrealistic data. In addition, a correction for low turbulence situations, where the eddy-covariance technique may be underestimated, is to take the friction velocity (u*) as a threshold to establish valid measurements, especially at night. This method, although widely used by the scientific community, is not unanimous. Especially since u* is itself a flow and consequently its value correlates with the time scale used for the analysis. This paper presents the annual carbon estimate of a natural pasture ecosystem, Pampa biome, in an experimental site established in Santa Maria - RS. We evaluated three distinct situations in the annual carbon estimate (NEP): i) without a u* filter; ii) with a fixed filter u* for all evaluated years and; iii) with the filter u* varying seasonally. The methodology used to estimate u* is the same as that used by Papale et al. (2006). The results show a total annual carbon sequestration variability of up to 10% depending on the methodology employed. The ecosystem in question, regardless of the method used, proved to be a carbon sink. However, the use of one methodology or another in ecosystem situations that are close to carbon assimaltion neutral should be closely scrutinized for an accurate annual balance.