Modelling sun-induced chlorophyll fluorescence (SIF) in evergreen conifer forests with a terrestrial biosphere model

Solar-induced chlorophyll fluorescence (SIF) is a small light signal emitted during the initial steps of photosynthesis and can be observed across scales (from photosystem level to satellites). To be able to model SIF, we need to understand the mechanistic processes (including both physical and biological) leading to the observed SIF signal. In this work, we implemented a representation of SIF emission and transmission processes into the terrestrial biosphere model QUINCY ('QUantifying Interactions between terrestrial Nutrient CYcles and the climate system'). We tested the model across three different boreal coniferous forests located in North America and Europe that have eddy covariance derived CO2 fluxes and tower-based SIF observations. We find that alternative SIF radiative transfer approaches (one based on mSCOPE, one on two-stream radiative transfer model L2SM, and one empirically based) overestimate the SIF signal, but show no large differences in the timing of their seasonal and diurnal predictions. The two-stream radiative transfer model approach, L2SM, provided stable performance while being comparatively computationally efficient. We find that our parameterization for sustained non-photochemical quenching is important for successfully simulating the timing of the SIF seasonal cycle. However, our parameterization did not work equally well across all three sites, likely because of different temperature regimes at the sites. We further evaluated the potential of remote sensing -based SIF from TROPOMI (the TROPOspheric Monitoring Instrument) to provide accurate information on SIF and found that it can potentially be used in model development. This study illustrates the usefulness of observations at different spatial scales and the linkages between SIF and GPP and their seasonal development at three different evergreen forest sites.