Abstract Wetland water quality plays a vital role in controlling emissions of methane, a strong greenhouse gas, and it is regulated by the sequestration of nutrients and toxic metals on iron‐rich soils. However, the stability of these functions under shifting precipitation and temperature regimes remains unknown, because of complex and unconstrained mechanisms through which climate and hydrology drive coupled biogeochemical processes. In particular, although increasing molecular‐scale evidence suggests that “cryptic” sulfur cycling can influence carbon and iron cycles in freshwater wetland porewaters, how these interactions scale up under dynamic hydrological conditions to influence atmospheric feedbacks and water quality have been unexplored. Barriers include uncertainties in reaction rates and networks involving intermediate‐valence sulfur forms, especially under variable hydrological conditions. Here we present a predictive, interdisciplinary framework that quantifies coupled iron‐sulfur‐carbon cycling under dynamic hyporheic flows by integrating reactive transport modeling, metagenomic sequencing, and multi‐scale geochemical analyses at a riparian wetland. Simulations show that even at low sulfate concentrations, cryptic sulfur cycling enhances dissolved iron concentrations by threefold, promoting iron flocs that improve water quality, and it suppresses methane production by 26%, tempering methane emissions. Moreover, incorporating weather‐driven water flux reversals on daily to seasonal scales further amplifies these effects, promoting an additional 25% in iron(II) accumulation and 20% in methane attenuation in porewater. By mechanistically linking microbial metabolism and hyporheic flow fluctuations with redox dynamics and methane production, our work demonstrates the need to consider hydrologically driven sulfur interactions, when predicting atmospheric feedbacks and contaminant dynamics in wetlands under changing climate.
Abstract Reservoirs provide water supply, clean energy, and numerous functions for human well‐being. However, reservoir construction often leads eutrophication due to altered hydrological conditions, presenting a major global environmental challenge. Hydrodynamic regulation is an effective measure for controlling reservoir eutrophication, yet current knowledge is largely based on the hydrodynamics‐water quality‐phytoplankton (HWP) pathway. This pathway also forms the basis for widely used hydrodynamic models simulating eutrophication processes, such as CE‐QUAL‐W2 and Delft3D. Despite its widespread application, the mechanisms and pathways beyond the classical HWP pathway that influence phytoplankton remain poorly explored. To address this gap, empirical dynamic modeling and Bayesian network were employed to construct causal networks illustrating how water level fluctuations influence phytoplankton dynamics, using 10‐year continuous daily data from the Three Gorges Reservoir. Our findings reveal that water level fluctuations impact water temperature, turbidity, and nitrogen concentrations, thereby influencing phytoplankton biomass through the classical HWP pathway. More importantly, the results indicate the presence of a possible non‐HWP pathway through top‐down control, most likely driven by zooplankton predation. This pathway exerts a stronger influence on phytoplankton biomass than the classical HWP pathway, suggesting an important mechanism through which water level fluctuations affect reservoir eutrophication has previously been overlooked. These findings expand our understanding of hydrodynamic impacts on phytoplankton and underscore the importance of non‐classical HWP pathways in determining eutrophication, offering new insights for reservoir management and the development of more robust hydrodynamic models.
Abstract Kinematic wave models are widely used to represent hillslope runoff in distributed hydrological models. However, their simplified descriptions often prevent the effective use of observed soil hydraulic parameters during calibration. To address this limitation, we derived a new analytical storage–runoff relationship that explicitly incorporates percolation under wet‐soil conditions and rapid drainage through subsurface pipes. This relationship was formulated using an approximate quasi‐steady‐state solution of the vertical Darcy equation, enabling computationally efficient evaluation without vertical discretization. Integrated into a kinematic wave framework, the new formulation was implemented in a distributed rainfall‐runoff model and tested using flood events in a steep mountainous dam basin in Japan. Constraining the calibration with locally observed soil hydraulic parameters used as reference values allowed the model parameters to remain within realistic and observationally consistent ranges. This approach accurately reproduced flood hydrographs while reducing parameter and predictive uncertainty. Leave‐one‐out cross‐validation further demonstrated improved performance in simulating unobserved flood events. By improving the process representation of hillslope runoff, the proposed model allows calibration to converge toward soil parameters within observationally reasonable ranges, thereby reducing equifinality without compromising simulation performance.
Abstract Traditional derivations of soil moisture flow equations are based on conservation laws and physical principles, but the resulting equations still depend on how individual processes and constitutive relationships are represented. Data‐driven equation discovery offers a complementary way to identify governing terms from observations, yet previous applications to soil moisture flow have relied mainly on numerical experiments. Here, we develop Spline‐SR, a spline‐regularized sparse regression framework for discovering nonlinear soil moisture flow equations from sparse fixed‐depth observations. Spline‐SR reconstructs continuous water‐content dynamics, evaluates candidate derivatives through automatic differentiation, and selects governing terms with state‐dependent spline coefficients. Numerical experiments show that the method reliably recovers the known three‐term structure of the water‐content form of the Richards equation under sparse and noisy sampling. We then apply Spline‐SR to measurements from sand and loamy‐sand columns under wetting, drying, and alternating conditions. For both soils, the experimental analysis identifies a parsimonious structure containing the three terms expected from the water‐content form of the Richards equation. The inferred coefficient functions vary across forcing conditions, while the resulting equations reconstruct the main features of soil moisture dynamics. These results show that sparse experimental observations can contain sufficient information to identify the governing structure of soil moisture flow equations, while coefficient relationships remain more sensitive to observational coverage and flow conditions.
Abstract Remote‐sensing spectral observations are widely used to monitor crop growth and productivity, yet their potential for agrohydrological forecasting remains limited by a lack of mechanistic links to crop and soil processes. Existing models often prescribe canopy states, simplify physiological responses, or omit explicit spectral simulations, limiting their ability to use spectral signals to constrain key crop and soil variables. To address this, we developed the STEMMUS‐SCOPE‐PG model, integrating a process‐based plant growth (PG) module into the STEMMUS‐SCOPE framework. In this system, simulated carbon assimilation and soil moisture jointly drive dynamic changes in leaf area index (LAI) and biomass, which in turn feed back into canopy radiative transfer, enabling bidirectional coupling between spectral signals (e.g., reflectance and solar‐induced chlorophyll fluorescence (SIF)) and hydrological fluxes. The model was evaluated using three years of field data from maize (C4) and soybean (C3) AmeriFlux sites (US‐Bo1 and US‐Ne3), together with MODIS/TROPOMI satellite observations. Results show that STEMMUS‐SCOPE‐PG reproduced LAI and biomass across both sites, and captured seasonal SIF dynamics consistent with TROPOMI observations. Relative to the original SCOPE and STEMMUS‐SCOPE configurations, the model reduced mean RMSE by 28.8% and 28.3% for gross primary productivity, and by 15.4% and 14.3% for evapotranspiration, respectively. Moreover, we revealed a linear relationship between daily LAI change and SIF during vegetative stages, and a nonlinear relationship during reproductive stages, highlighting phenological effects on crop‐SIF dynamics. These findings demonstrate the value of linking crop physiology with satellite observations, supporting applications in precision agriculture and land‐surface modeling.
