Estimation of uncertainties due to data scarcity in model upscaling: a case study of methane emissions from rice paddies in China

Abstract

Abstract. Data scarcity is a major cause of substantial uncertainties in regional estimations conducted with model upscaling. To evaluate the impact of data scarcity on model upscaling, we introduce an approach for aggregating uncertainties in model estimations. A data sharing matrix was developed to aggregate the modeled uncertainties in divisions of a subject region. In a case study, the uncertainty in methane emissions from rice paddies on mainland China was calculated with a local-scale model CH4MOD. The data scarcities in five of the most sensitive model variables were included in the analysis. The national total methane emissions were 6.44–7.32 Tg, depending on the spatial resolution used for modeling, with a 95% confidence interval of 4.5–8.7 Tg. Based on the data sharing matrix, two numeral indices, IR and Ids, were also introduced to suggest the proper spatial resolution in model upscaling.