Terrestrial ecosystems provide multiple services interacting in complex ways. However, most ecosystem services (ESs) models (e.g., InVEST and ARIES) ignored the relationships among ESs. Process‐based models can overcome this limitation, and the integration of ecological models with remote sensing data could greatly facilitate the investigation of the complex ecological processes. Therefore, based on the Carbon and Exchange between Vegetation, Soil, and Atmosphere (CEVSA) models, we developed a process‐based ES model (CEVSA‐ES) integrating remotely sensed leaf area index to evaluate four important ESs (i.e., productivity provision, carbon sequestration, water retention, and soil retention) at annual timescale in China. Compared to the traditional terrestrial biosphere models, the main innovation of CEVSA‐ES model was the consideration of soil erosion processes and its impact on carbon cycling. The new version also improved the carbon‐water cycle algorithms. Then, the Sobol and DEMC methods that integrated the CEVSA‐ES model with nine flux sites comprising 39 site‐years were used to identify and optimize parameters. Finally, the model using the optimized parameters was validated at 26 field sites comprising 135 site‐years. Simulation results showed good fits with ecosystem processes, explaining 95%, 92%, 76%, and 65% interannual variabilities of gross primary productivity, ecosystem respiration, net ecosystem productivity, and evapotranspiration, respectively. The CEVSA‐ES model performed well for productivity provision and carbon sequestration, which explained 96% and 81% of the spatial‐temporal variations of the observed annual productivity provision and carbon sequestration, respectively. The model also captured the interannual trends of water retention and soil erosion for most sites or basins.

Chinese forests cover most of the representative forest types in the Northern Hemisphere and function as a large carbon (C) sink in the global C cycle. The availability of long-term C dynamics observations is key to evaluating and understanding C sequestration of these forests. The Chinese Ecosystem Research Network has conducted normalized and systematic monitoring of the soil-biology-atmosphere-water cycle in Chinese forests since 2000. For the first time, a reference dataset of the decadal C cycle dynamics was produced for 10 typical Chinese forests after strict quality control, including biomass, leaf area index, litterfall, soil organic C, and the corresponding meteorological data. Based on these basic but time-discrete C-cycle elements, an assimilated dataset of key C cycle parameters and time-continuous C sequestration functions was generated via model-data fusion, including C allocation, turnover, and soil, vegetation, and ecosystem C storage. These reference data could be used as a benchmark for model development, evaluation and C cycle research under global climate change for typical forests in the Northern Hemisphere.

While a number of machine learning (ML) models have been used to estimate RE, systematic evaluation and comparison of these models are still limited. In this study, we developed three traditional ML models and a deep learning (DL) model, stacked autoencoders (SAE), to estimate RE in northern China’s grasslands. The four models were trained with two strategies: training for all of northern China’s grasslands and separate training for the alpine and temperate grasslands. Our results showed that all four ML models estimated RE in northern China’s grasslands fairly well, while the SAE model performed best (R² = 0.858, RMSE = 0.472 gC m⁻² d⁻¹, MAE = 0.304 gC m⁻² d⁻¹). Models trained with the two strategies had almost identical performances. The enhanced vegetation index and soil organic carbon density (SOCD) were the two most important environmental variables for estimating RE in the grasslands of northern China. Air temperature (Ta) was more important than the growing season land surface water index (LSWI) in the alpine grasslands, while the LSWI was more important than Ta in the temperate grasslands. These findings may promote the application of DL models and the inclusion of SOCD for RE estimates with increased accuracy.

It is important to accurately evaluate ecosystem respiration (RE) in the alpine grasslands of the Tibetan Plateau and the temperate grasslands of the Inner Mongolian Plateau, as it serves as a sensitivity indicator of regional and global carbon cycles. Here, we combined flux measurements taken between 2003 and 2013 from 16 grassland sites across northern China and the corresponding MODIS land surface temperature (LST), enhanced vegetation index (EVI), and land surface water index (LSWI) to build a satellite-based model to estimate RE at a regional scale. First, the dependencies of both spatial and temporal variations of RE on these biotic and climatic factors were examined explicitly. We found that plant productivity and moisture, but not temperature, can best explain the spatial pattern of RE in northern China’s grasslands; while temperature plays a major role in regulating the temporal variability of RE in the alpine grasslands, and moisture is equally as important as temperature in the temperate grasslands. However, the moisture effect on RE and the explicit representation of spatial variation process are often lacking in most of the existing satellite-based RE models. On this basis, we developed a model by comprehensively considering moisture, temperature, and productivity effects on both temporal and spatial processes of RE, and then, we evaluated the model performance. Our results showed that the model well explained the observed RE in both the alpine (R² = 0.79, RMSE = 0.77 g C m⁻² day⁻¹) and temperate grasslands (R² = 0.75, RMSE = 0.60 g C m⁻² day⁻¹). The inclusion of the LSWI as the water-limiting factor substantially improved the model performance in arid and semi-arid ecosystems, and the spatialized basal respiration rate as an indicator for spatial variation largely determined the regional pattern of RE. Finally, the model accurately reproduced the seasonal and inter-annual variations and spatial variability of RE, and it avoided overestimating RE in water-limited regions compared to the popular process-based model. These findings provide a better understanding of the biotic and climatic controls over spatiotemporal patterns of RE for two typical grasslands and a new alternative up-scaling method for large-scale RE evaluation in grassland ecosystems.