A Monte Carlo approach to estimate the uncertainty in soil CO2 emissions caused by spatial and sample size variability

Abstract

The soil CO2 emission is recognized as one of the largest fluxes in the global carbon cycle. Small errors in its estimation can result in large uncertainties and have important consequences for climate model predictions. Monte Carlo approach is efficient for estimating and reducing spatial scale sampling errors. However, that has not been used in soil CO2 emission studies. Here, soil respiration data from 51 PVC collars were measured within farmland cultivated by maize covering 25 km2 during the growing season. Based on Monte Carlo approach, optimal sample sizes of soil temperature, soil moisture, and soil CO2 emission were determined. And models of soil respiration can be effectively assessed: Soil temperature model is the most effective model to increasing accuracy among three models. The study demonstrated that Monte Carlo approach may improve soil respiration accuracy with limited sample size. That will be valuable for reducing uncertainties of global carbon cycle.