Search

Soil heterotrophic respiration (RH) is one of the largest and most uncertain components of the terrestrial carbon cycle, directly reflecting carbon loss from soils to the atmosphere. However, high variations and uncertainties of RH existing in global carbon cycling models require RH estimates from different angles, e.g., a data-driven angle. To fill this knowledge gap, this study applied a Random Forest (RF) algorithm (a machine learning approach) to (1) develop a globally gridded RH dataset and (2) investigate its spatial and temporal patterns from 1980 to 2016 at the global scale by linking field observations from the Global Soil Respiration Database and global environmental drivers (temperature, precipitation, soil water content, etc.). Finally, a globally gridded RH dataset was developed covering from 1980 to 2016 with a spatial resolution of half a degree and a temporal resolution of 1 year. Globally, the average annual RH was 57.2±0.6 Pg C a -1 from 1980 to 2016, with a significantly increasing trend of 0.036±0.007 Pg C a -2. However, the temporal trend of the carbon loss from RH varied in climate zones, and RH showed a significant and increasing trend in boreal and temperate areas. In contrast, such a trend was absent in tropical regions. Temperature-driven RH dominated 39 % of global land and was primarily distributed at high-latitude areas. The areas dominated by precipitation and soil water content were mainly semiarid and tropical areas, accounting for 36 % and 25 % of global land area, respectively, suggesting variations in the dominance of environmental controls on the spatial patterns of RH. The developed globally gridded RH dataset will further aid in the understanding of the mechanisms of global soil carbon dynamics, serving as a benchmark to constrain terrestrial biogeochemical models. The dataset is publicly available at 10.6084/m9.figshare.8882567 (Tang et al., 2019a). [ABSTRACT FROM AUTHOR]

Belowground autotrophic respiration (RA) is one of the largest but most highly uncertain carbon flux components in terrestrial ecosystems. However, RA has not been explored globally before and still acts as a "black box" in global carbon cycling currently. Such progress and uncertainty motivate the development of a global RA dataset and understanding its spatial and temporal patterns, causes, and responses to future climate change. We applied the random forest (RF) algorithm to upscale an updated dataset from the Global Soil Respiration Database (v4) – covering all major ecosystem types and climate zones with 449 field observations, using globally gridded temperature, precipitation, soil and other environmental variables. We used a 10-fold cross validation to evaluate the performance of RF in predicting the spatial and temporal pattern of RA. Finally, a globally gridded RA dataset from 1980 to 2012 was produced with a spatial resolution of 0.5 ∘ × 0.5 ∘ (longitude × latitude) and a temporal resolution of 1 year (expressed in g C m -2 yr -1 ; grams of carbon per square meter per year). Globally, mean RA was 43.8±0.4 Pg C yr -1 , with a temporally increasing trend of 0.025±0.006 Pg C yr -2 from 1980 to 2012. Such an incremental trend was widespread, representing 58 % of global land. For each 1 ∘ C increase in annual mean temperature, global RA increased by 0.85±0.13 Pg C yr -2 , and it was 0.17±0.03 Pg C yr -2 for a 10 mm increase in annual mean precipitation, indicating positive feedback of RA to future climate change. Precipitation was the main dominant climatic driver controlling RA, accounting for 56 % of global land, and was the most widely spread globally, particularly in dry or semi-arid areas, followed by shortwave radiation (25 %) and temperature (19 %). Different temporal patterns for varying climate zones and biomes indicated uneven responses of RA to future climate change, challenging the perspective that the parameters of global carbon stimulation are independent of climate zones and biomes. The developed RA dataset, the missing carbon flux component that is not constrained and validated in terrestrial ecosystem models and Earth system models, will provide insights into understanding mechanisms underlying the spatial and temporal variability in belowground vegetation carbon dynamics. The developed RA dataset also has great potential to serve as a benchmark for future data–model comparisons. The developed RA dataset in a common NetCDF format is freely available at 10.6084/m9.figshare.7636193 (Tang et al., 2019). [ABSTRACT FROM AUTHOR]