Gap-filling carbon dioxide, water, energy, and methane fluxes in challenging ecosystems: Comparing between methods, drivers, and gap-lengths

Summary

This paper extends existing gap-filling research by evaluating random forest regression for methane flux data and comparing multiple gap-filling methodologies across challenging ecosystems including European managed pastures, Southeast Asian converted peatlands, and North American drylands. The authors demonstrate that RFR performs competently as an alternative to standard algorithms, with MDS recommended for short gaps in CO₂ data but RFR superior for longer gaps and other flux types. Notably, RFR reliably filled cumulative fluxes over gaps exceeding three months whilst preserving key environment-flux relationships in the gap-filled data.

UK applicability

The findings are applicable to UK managed pasture systems, which are commonly studied using eddy covariance techniques and regularly encounter data gaps. The recommendation to use RFR for longer gaps and for trace gases such as methane will be relevant to UK research on grassland carbon and methane budgets, though the study's inclusion of drylands and converted peatlands extends beyond typical UK agricultural conditions.

Key measures

Gap-filling accuracy for CO₂, water, energy, and methane fluxes; performance across gap lengths (< 12 days, > 30 days, > 3 months); reliability of cumulative flux estimates; preservation of environment-flux relationships

Outcomes reported

The study evaluated random forest regression (RFR) and marginal distribution sampling (MDS) methods for filling missing data in eddy covariance measurements of carbon dioxide, water, energy, and methane fluxes across managed pastures, converted peatlands, and drylands. Performance was assessed across different gap lengths and flux types to identify optimal approaches for challenging ecosystems.

Topic tags