Digging deeper: Assessing the predictive power of common greenhouse gas accounting tools for soil carbon sequestration under organic amendment

Abstract

Effective project implementation and quantification of emissions reduction in climate-smart agriculture initiatives face challenges in measurement, monitoring, and verification. To address these challenges, predictive models are regularly used to estimate the emissions reduction potential of land management changes and to prioritize funding for projects. Despite their growing utility, few studies have evaluated the performance of publicly available model tools using site specific data. This study evaluated the performance and utility of four common model tools that represent the three Intergovernmental Panel on Climate Change model tiers to predict soil organic carbon storage and estimate greenhouse gas emissions on working lands under organic matter amendment. Field data from two long-term, compost application experiments in Washington State formed the basis for model simulations using DayCent, COMET-Farm, Cool Farm, and the Washington State Climate Smart Estimator (WaCSE). Soil carbon sequestration and emissions estimates varied among the evaluated models, which was expected given their differential data requirements and input capabilities. COMET-Farm, although easier to use, exhibited a higher level of bias compared to DayCent, which was expected as a mixed tier model. The DayCent model, the model engine for the COMET-Farm tool, demonstrated the ability to explain ∼50% more of the variation in the observed values compared to COMET-Farm when initiated using the same parameters. Cool Farm was unsuitable for estimating SOC sequestration benefits from compost application primarily because it did not add carbon to the soil pool following amendment. The differences in emissions estimates derived from WaCSE compared with other tools could be attributed solely to its highly constrained input parameters and basis in tier 1 emissions factors. We conclude that online tools can provide rapid estimates of greenhouse gas emissions reduction potential over larger areas or groups of farms but should be used with caution for site-specific estimates. Hence, it is crucial to clarify the intended purpose of an assessment and the designed function of model tools when evaluating their suitability for prioritizing funding for climate-smart agriculture initiatives at the individual farm level.