Hui Zheng

Traditional land–atmosphere coupling models assume that land productivity decreases monotonically as meteorological aridity increases. However, a strong resilience paradox exists across the Loess Plateau in China: moderate drying in 2001 led to the collapse of canopy carbon, while record compound dry–heat extremes in 2006 resulted in functional biophysical buffering. In this work, we extend the ecological amplitude modulation (EAM) model from humid to water-limited drylands and solve this paradox by modeling it as a state-dependent dynamic bifurcation, mediated by cross-scale vegetative-hydrological interactions. EAM identifies an empirical boundary at approximately 40 days that separates high-frequency synoptic atmospheric fluctuations (periods 25 days). In 2001, early-spring structural overshoot depleted deep vadose-zone moisture reserves (100∼289 cm), driving the scale separation ratio toward unity (κ→1) and locking daily canopy assimilation into destructive, in-phase synchronization with sub-seasonal drought (r=0.324, p<0.001). Conversely, retained deep subsoil reserves in 2006 maintained dynamical scale orthogonality (κ≫1), enabling active anti-phase buffering (r=-0.468, p<0.001) under extreme vapor pressure deficit (VPD). Micro-scale synthesis reveals that subsoil moisture exhaustion activated dried soil layers, triggering a 5.07-fold non-linear amplification in stomatal sensitivity to rising VPD (dΔGPP/dΔVPD collapsing from -0.0162 to -0.0819) as an early-onset conservative defense. These findings demonstrate that unmanaged vegetative overshoot traps restored drylands in chronic subterranean drought, confirming that EAM diagnoses cross-scale hydraulic failure and indicating that moderate stand thinning (30% to 50%) is vital for climate resilience.

Terrestrial ecosystems frequently exhibit a "resilience paradox" under compound drought–heatwave extremes, where moderate meteorological anomalies trigger catastrophic carbon collapse while harsher compound events induce minimal functional loss with resilience. Here, we evaluate this paradox across South Chinese monsoon forests during the contrasting 2003–2004 (collapse) and 2013–2014 (buffered) compound extremes. We demonstrate that linear time-domain causal discovery (PCMCI) fails to distinguish these opposite states due to scale-mixing artifacts that convolve fast atmospheric pulses with slow hydrological memory. To resolve cross-scale non-linear interactions, we advance an Ecological Amplitude Modulation approach integrating spatiotemporal Fast-MEEMD with Normalized Hilbert spectral analysis. We identify an intrinsic quasi-40-day spectral gap separating stationary high-frequency physiological carriers (26 days). In 2D Holo-Hilbert phase space, the 2003 collapse (ΔGPP=-1.90σ) was driven by Nonlinear Multiplicative Amplification, where spring structural overshoot depleted root-zone water, constraining physiological activity and precipitating hydraulic breakdown. In contrast, the 2013 resilience (ΔGPP=+0.61σ) arose from Adaptive Decoupling, where intact antecedent moisture preserved the 40-day buffer against atmospheric drying. This multi-scale diagnostic approach provides a physics-based biomarker for anticipating non-linear ecosystem tipping points.

Assimilating the Gravity Recovery and Climate Experiment (GRACE) terrestrial water storage (TWS) has been proven a promising approach to improving terrestrial hydrological simulation, but its impact under different climatic regimes and on hydrological responses to important climatic oscillations (i.e., El Niño‐Southern Oscillation (ENSO)) remains unclear. This study constructed a land surface data assimilation (DA) system based on a multi‐physics ensemble constructed by the Noah‐Multiparameterization land surface model (Noah‐MP) and the Data Assimilation Research Testbed over the southern contiguous United States (CONUS). The system features a set of high‐skill and low‐interdependence parameterization configurations as the assimilation ensemble, based on which the daily‐scale GRACE TWS was assimilated using the Ensemble Adjustment Kalman Filter. Results show that DA enhances the performance of TWS anomaly in the southeastern/south‐central CONUS, the Rocky Mountains, the Sierra Nevada, and the Pacific Northwest. For soil moisture, DA offers more skill gains in the deep layer than in the surface layer. For snow water equivalent, DA improves the performance in the northeastern/central CONUS and the western mountains, but offers limited gains in the southern CONUS, where the snow cover is shallow and seasonal. DA improves the runoff simulations in arid and semi‐arid regions where Noah‐MP performs poorly, but slightly degrades the performance in humid regions where Noah‐MP performs well. DA also improves the fidelity of runoff–ENSO linkages in semi‐arid and arid regions by eliminating spurious correlations. Our findings confirm the effectiveness of GRACE DA in reducing hydrological simulation uncertainties and demonstrate its potential for improving water management planning under climatic oscillations.

Surface solar radiation (SSR) is a fundamental component of the Earth’s energy budget, driving climate dynamics, ecosystem stability, and solar energy potential. By integrating high-quality in situ observations from 87 ground stations with multi-source satellite products (CERES) and reanalysis data, this study systematically quantifies the spatiotemporal evolution and driving mechanisms of SSR across China from 1994 to 2022. Nationally, SSR exhibits a significant brightening trend of 72 MJ m⁻² per decade, despite pronounced regional and seasonal divergence. Annual SSR increases significantly in North, Central, Southwest, and South China—with North China experiencing the fastest growth (160 MJ m⁻² per decade). Strikingly, the Tibetan Plateau exhibits a persistent and significant “dimming” trend (−86 MJ m⁻² per decade). Furthermore, spatial analysis reveals that summer SSR trends show a unique meridional “+ − +” dipole structure that closely resembles the dominant mode of summer precipitation over eastern China. Quantitative attribution analysis indicates that declining aerosol optical depth (AOD) is the primary driver of brightening across most of China, whereas the dimming on the Tibetan Plateau is specifically triggered by increased atmospheric water vapor, cloud cover, and precipitation. These findings underscore that in sensitive high-altitude regions, regional climate dynamics and hydrological feedback can override the radiative effects of anthropogenic aerosol reductions, highlighting the necessity for localized solar resource assessments.

Evaluating kilometer‐scale atmospheric models in data‐sparse mountains is challenging because in situ meteorological observations are scarce and remote‐sensing products are uncertain. Using hydrological models to link atmospheric model‐simulated precipitation to streamflow is equally problematic, because those models carry substantial structural and parameter uncertainty in mountain terrain. We therefore propose a simple, low‐uncertainty alternative: route atmospheric model runoff through a network routing model calibrated against observed streamflow. We apply this approach to assess thirteen 3‐km Weather Research and Forecasting (WRF) experiments over the Yarlung Zangbo River basin on the Tibetan Plateau, with systematically varied parameterizations for radiation, microphysics, planetary boundary layer, and orographic drag. In mountainous basins, routing can be simplified using the Muskingum model. This parsimonious approach proves effective and robust: hourly streamflow simulations achieve a median Pearson correlation of 0.75 across all six examined gauging stations, and calibrated wave celerity shows little sensitivity to the driving WRF experiment. Statistical analysis confirms that routing model uncertainty is small enough to distinguish performance differences among the WRF configurations. Compared with precipitation‐based evaluation, the routing‐based approach provides complementary and more hydrologically relevant measures of atmospheric model performance, especially when basin‐mean precipitation is already well captured. The method offers a discriminative tool that leverages the superior spatial representativeness of streamflow observations to evaluate atmospheric models in data‐sparse mountainous regions.

Regional air‐sea coupling critically influences East Asian summer climate, yet its representation in current modeling systems remains challenging. This study enhances mechanistic understanding of this coupling by comparing two long‐term simulations (1991–2014) using the new Regional Integrated Earth Model System (RIEMS) with air‐sea coupling and its standalone atmospheric model (WRF). Results show both RIEMS and WRF demonstrate advantages and added values in reproducing summer air temperature (T2mean and T2max) and precipitation compared to the ensemble mean of global climate models (ENS). Meanwhile, RIEMS better reproduces T2mean and T2max compared to WRF, particularly over East China. RIEMS reduces the T2max warm bias in WRF by 0.7°C on average over land. Additionally, RIEMS offers more accurate simulations of surface sensible and latent heat fluxes while WRF overestimates sensible heat flux by nearly 7.0 W m ⁻² over land. Mechanistic analyses indicate that WRF reproduces less cloud cover and increased incoming shortwave radiation, resulting in higher sensible heat flux and warmer air temperature. The climatological decrease in cloud cover in WRF can be attributed to its warmer air temperature. Meanwhile, WRF simulated less outgoing shortwave radiation and trapped more longwave radiation, leading to warmer surface air temperature. Furthermore, without air‐sea coupling, WRF strengthened water and energy transportation from ocean to land through a stronger southwesterly airflow, which is more conducive to simulating higher surface air temperatures in WRF over a large area. These findings provide mechanistic evidence that regional air‐sea coupling is essential for credible climate simulations and projections over East Asia.

Understanding how compound extremes affect terrestrial ecosystems is a major challenge in Earth system science. Although the combined effects of stressors are recognized, the manner in which the prestress state determines the basic response mechanism remains unclear. In this study, we used the “natural experiment” methodology to compare two major extreme events within a monsoon‐influenced, low‐latitude highlands setting to explain mechanistic changes in land‐atmosphere interactions. By analyzing an extensive set of remote sensing and reanalysis data with nonlinear structural equation modeling, we show that the ecosystem response shifted from a traditional water‐limited paradigm during the 2010 drought to an energy‐governed paradigm during the 2019 heatwave. Our results suggest that this transition is governed by the antecedent root zone soil moisture status, which acts as a tipping point fundamentally shifting the impact of atmospheric factors on canopy evapotranspiration screens, such as temperature and vapor pressure deficit. This study highlights a possible threshold‐type state dependence non‐linearity, which is lacking in major Earth System Models. Incorporating this “hydrological memory” is crucial for minimizing uncertainties in climate projections and for correctly assessing the vulnerability of ecosystems in a warming world.

Accurate representation of drainage density in a river network delineation is crucial for large-domain hydrodynamic simulations, but existing digital elevation model (DEM)-based methods fail to preserve observed drainage density across the entire network. This study introduces a fundamentally improved DEM-based river network delineation method that preserves observed drainage density by incorporating a novel concept of upstream accumulation length, which integrates flow direction and drainage density. A test case demonstrated the method’s compatibility with widely used critical drainage area-based methods; they produced identical results when using the same drainage density. The method was then applied to the Chinese Mainland using the MERIT-Hydro flow direction dataset and a drainage density dataset from the Third National Land Resources Survey of China. The resulting dataset shows superior performance in capturing drainage density variations, particularly in regions with complex topography, compared to existing datasets. The method is promising for delineating more accurate river networks by combining satellite-derived or surveyed drainage density with DEMs, thereby laying a valuable foundation for large-domain hydrodynamic simulations in vast regions of the globe, where there is a lack of manually maintained river geometry datasets.