Improving non-photochemical quenching, fluorescence emission and gross primary production of boreal evergreen needleleaf forests in a land surface model

Climate change is caused by the ever-increasing anthropogenic CO₂ emissions and the resulting accumulation of carbon dioxide in the atmosphere, whose effects are so far mitigated by the oceanic and continental CO₂ uptakes. The terrestrial sink is the most uncertain component of the global carbon cycle mainly because the gross primary production (GPP), which is the quantity of atmospheric carbon absorbed by plants through photosynthesis, is highly variable in time and space, and because the involved processes are complex. As photosynthesis is not directly measurable at a scale larger than the leaf, land surface modelers have been looking for large-scale proxies of GPP. Solar-induced chlorophyll fluorescence (SIF) estimates from satellite instruments have emerged as a promising resource to inform on the space-time distribution of GPP, and have been increasingly used over the last decade. However, numerous challenges remain to be addressed to acutely understand the SIF signal, and its relationship with plant photosynthesis, to be able to correctly exploit space-borne SIF retrievals to constrain GPP simulated by land surface models (LSMs). Notably, we still lack knowledge on the non-photochemical quenching (NPQ), which represents the third deactivation pathway of the light energy absorbed by chlorophyll pigments, alongside photochemistry and fluorescence. In this study, we focused on boreal evergreen needleleaf forests (BorENF), which cover 29% of the world's total forest area, and whose GPP budget is still debated. We took advantage of both passive and active fluorescence measurements to improve the representation of NPQ, SIF and GPP in the ORCHIDEE LSM. We first used active measurements taken at the Hyytiälä BorENF site on Pinus sylvestris trees, to separately model the sustained and reversible NPQ components. Indeed, it was previously documented that during winter such evergreen trees suppress photosynthesis and sustain molecular modifications of the photosynthetic chain allowing the dissipation of the excess energy absorbed as heat (sustained NPQ). The reversible NPQ occurs during growth season in response to environmental stress (e.g., excessive light or droughts). In a second step, we optimised several ORCHIDEE parameters related to NPQ, SIF and GPP representations, using data assimilation techniques. We performed a multi-variables and multi-sites approach, simultaneously assimilating in situ GPP estimates at nine BorENF FLUXNET sites and collocated SIF estimates from the TROPOMI satellite instrument. The improvements brought to SIF and GPP were evaluated at those sites over independent years (i.e. not used in assimilation) with positive results, and at the regional scale against the FLUXCOM and FLUXSAT GPP products, as well as against TROPOMI SIF data.