Assessment of aboveground carbon mass in a Mediterranean downy oak ecosystem using airborne lidar and NASA GEDI measurements

Abstract. The forest system is the main carbon sink after the oceans. However, due to climate change, an alarming number of tree species of the Northern Hemisphere are at risk of migrating northwards or becoming extinct. This is the case of the downy oak (Quercus pubescens), one of the main species making up the forests close to the Mediterranean Sea in France. Our aim is to retrieve aboveground carbon (AGC) and underground root carbon (UGC) stocks of the downy oak forest at Observatoire de Haute Provence (OHP), located about 80 km north of Marseille, in order to provide a baseline against which to assess the effect of climate change on this model species. The study presented here is based on airborne lidar observations gathered in May 2012 and field measurements from 2012, 2018 and 2023 in the OHP forest. The OHP forest consists of ~75 % downy oak, which is highly sensitive to global warming. Field measurements indicate minimal changes in tree growth and density between 2012 and 2023, and that its carbon storage efficiency remains stationary. As retrieved by lidar measurements, tree top heights (TTHs) are mostly between 5 and 12 m, with an uncertainty of around 1 m. The slow evolution of trees at the OHP site makes it appropriate to use lidar data recorded in 2012 to assess the carbon stock trapped in current forest biomass. By coupling allometric laws established from field measurements with lidar observations, we show that the quantities of carbon trapped in aboveground biomass are double those trapped in the root system. Over an area of ~24 ha, mean values of 15±14 tC ha^-1 are assessed for the aerial biomass, against 8–10±3–7 tC ha^-1 for the roots of diameter larger than 1 cm for low and high assessments, respectively. These values depend heavily on the height of the sampled trees themselves, as well as on their location on the OHP plateau (smaller trees, 5–6 m) or on the slope (tallest trees, 10–12 m). Using a Monte Carlo approach, the relative uncertainties in AGC have been calculated to be of the order of 17 % and 11 % for trees 5–6 m and 10–12 m tall, respectively. For UGC, the relative uncertainties were calculated as 8 and 5 % for the same tree heights, but the assumptions on the allometric model are associated with biases that can easily reach 100 %. Although the surface footprints are different, we show that there is a reasonable agreement between our airborne lidar measurements and the level 2B (TTH) and (aboveground biomass) operational products of the Global Ecosystem Dynamics Investigation mission on the International Space Station for data acquired between 2019 and 2022.