Scale variance in the carbon dynamics of fragmented, mixed-use landscapes estimated using model–data fusion

Abstract. Many terrestrial landscapes are heterogeneous. Mixed land cover and use generate a complex mosaic of fragmented ecosystems at fine spatial resolutions with contrasting ecosystem stocks, traits, processes, each differently sensitive to environmental human factors. Representing complexity within models is key challenge for understanding regional carbon dynamics, their sensitivity gradients, resilience in the face climate change. Heterogeneity underpins this due trade-off between fidelity representation modelling frameworks computational capacity required fine-scale model calibration simulation. We directly address by quantifying simulated fluxes mixed-use landscape UK resolution analysis. test two different approaches combining Earth observation (EO) data into CARDAMOM model–data fusion (MDF) framework, assimilating time series satellite-based EO-derived estimates leaf area biomass stocks constrain parameters uncertainty an intermediate C cycle. In first approach, calibrated pixel level, representing “community average” encompassed management. This represents our baseline approach. second, we stratify based on (e.g. coniferous forest, arable/pasture) calibrate independently using EO specific stratum. scale dependence these grid spanning 1 0.05∘ over mixed-land-use region UK. Our analyses indicate that matters MDF. Under community average approach biological (gross primary productivity, Reco) relatively insensitive resolution. However, disturbance exhibit variance increases greater fragmentation coarser domains. contrast, stratification assimilated fine-resolution distributions resolved dependence, leading were 40 %–100 % higher than experiments. The differences result balance stratified experiment suggest weaker sink compared experiment. also find stratifying domain leads retrieved reflect variations function neighbouring areas use. emergent strata give rise divergent responses future Accounting structure heterogeneous stratification) therefore vital ensuring ecological large-scale MDF frameworks. need arises because places strong controls distribution plant functional traits processes controlling through landscapes, particularly those related management disturbance. Given importance global fluxes, together widespread increase forest results carry broader significance application national scales.