Monitoring fine-scale gross primary productivity along the Canadian treeline through multi-source remote sensing

Vegetation productivity along the northern treeline is undergoing rapid change in response to climate warming. Due to sparse in situ flux data, remotely sensed data are essential in monitoring the carbon cycle in these environments. However, challenges persist in boreal-tundra landscapes, where complex land cover patterns limit the utility of current coarse-resolution modelling efforts. Here, we apply the CAN-TG model to produce detailed (30 m) estimates of gross primary productivity (GPP) across the Canadian boreal-tundra transition zone, using satellite observations of vegetation greenness, land surface temperature, soil moisture, and freeze/thaw states. These variables represent key indicators of boreal productivity, capturing information on vegetation physiology, hydrological constraints, and seasonal phenology at varying spatial scales. GPP estimates were evaluated using independent GPP and solar-induced fluorescence (SIF), yielding strong agreement at coarse spatial scales (R² ≈ 0.91 and 0.88, respectively). Along treeline transects, CAN-TG estimates demonstrated notable correlations with distance above and below the treeline. Shifts in annual GPP of 0.12-0.36 g C m² year^-1 per 1 km shift in treeline position were observed, which may yield annual GPP increases of 0.77-2.32 Tg C year^-1 across the study domain over the next 100 years. Given the tight coupling between GPP and respiration in northern ecosystems, these increases represent a critical contribution to the maintenance of high-latitude terrestrial carbon sinks. This work offers a robust, scalable framework for monitoring productivity at higher spatial resolutions than previously possible, and provides insight into future changes under warming climate and shifting forest structure at the treeline.