PNNL

Abstract

Temperature sensitivity of soil respiration (Q10) serves as a critical role in benchmarking the potential intensity of regional and global terrestrial soil carbon fluxes-climate feedbacks. Although the field observations have demonstrated the strong spatial heterogeneity of Q10, knowledge gap still exists in the environmental factors influencing the spatial and temporal patterns of Q10 at the regional scale. Therefore, a gridded Q10 dataset from 1994 to 2016 with a spatial resolution of 1 km was developed from 515 field observations at 5 cm soil depth using a machine learning approach linked to climate, soil and vegetation variables across China. Results showed that annual predicted Q10 varied from 1.54 to 4.17 across China, with an area-weighted average of 2.52. Although there was no significant temporal pattern for Q10 (p = 0.32) across China, annual vegetation productivity (indicated by normalized difference vegetation index, NDVI), rather than temperature or precipitation, was positively correlated to Q10 (p