Search

Hydrological information is essential for planning water resource projects and for simulating hydrological models to calibrate and generate streamflow data. This study evaluates the performance of distributed and semi-distributed hydrological models in the Tamakoshi River Basin (TRB), Nepal. The models assessed include the Spatial Process in Hydrology (SPHY), Hydrologiska Byråns Vattenbalansavdelning (HBV), and Hydrologic Engineering Centre Hydrologic Modeling System (HEC-HMS). The models were evaluated using metrics such as Nash-Sutcliffe Efficiency (NSE) ranging from 0.62 to 0.77, Coefficient of Determination (R2) from 0.76 to 0.77, and Percentage Bias (P Bias) from 26.16 % to −3.54 % during daily calibration. All three models were successfully calibrated and validated on a daily and monthly basis using observed discharge data at the Busti station in the TRB. The SPHY model performed better during the calibration period compared to the validation period, while the HBV and HEC-HMS models showed better performance during validation than calibration. Among the statistical indicators, the coefficient of determination closely matched, but NSE and P Bias showed some deviations. This comparative analysis of calibrated and validated models is valuable for filling gaps in hydrological data records and for selecting appropriate hydrological models for planning, managing, and developing water resource projects in the TRB.

Numerical models have risen to prominence as indispensable tools for the in-depth study of the water cycle and these extreme hydrological phenomena， gaining widespread application across the globe.To delve into the spatiotemporal evolution patterns of global terrestrial water circulation against the backdrop of climate change and to decipher the intricate feedback mechanisms among atmospheric， land， and hydrological systems， the exploration of coupled atmospheric-land-hydrological models has emerged as a pivotal area of focus in the international research landscape dedicated to atmospheric and hydrological studies.This paper embarks on its journey by meticulously reviewing and delineating the evolution of coupled models， shedding light on the distinct advantages of the Weather Research and Forecasting Model Hydrological （WRF-Hydro） modeling system.It methodically dissects the primary sensitivity parameters of the WRF-Hydro model， while extensively covering its applications in analyzing surface runoff， soil moisture， the energy-water cycle， and the intertwined atmospheric and hydrological processes.The discourse culminates in a forward-looking exploration of the future directions in the development of the WRF-Hydro coupled model.Emphasizing strategic advancements， the paper advocates for a concerted effort towards the creation of robust scale conversion schemes， the refinement of parameterization methods， and the execution of high-resolution simulations.These simulations are crucial for accurately mapping the spatial and temporal dynamics of atmospheric and hydrological variables within basins， thereby significantly enhancing the model's capacity to intricately depict the interactions among atmospheric conditions， land surface phenomena， and hydrological processes.This comprehensive approach underlines the imperative to deepen our understanding and improve our modeling capabilities， aiming at a more effective prediction and management of the impacts arising from climate change and extreme hydrometeorological events.

Soil temperature is an important physical quantity in earth science research.In the study of Land-atmosphere interaction， soil temperature not only affects the physical， biological and chemical processes of the underlying surfaces， but also plays an important role in the energy and material exchange between land and atmosphere.With the availability of more and more relevant data， machine learning methods have been introduced into soil temperature prediction by more and more researchers， and have surpassed the performance of statistical models and physical models in many tasks.This paper compares three common methods of soil temperature calculation： statistical model， physical model and machine learning method， and briefly introduces the principles and characteristics of various machine learning models applied to soil temperature research.Based on domestic and foreign literatures， the research progress of traditional machine learning and deep learning in three aspects of soil temperature is summarized.In the study of the spatial distribution of soil temperature， traditional machine learning methods can learn spatial characteristics through the spatial heterogeneity of influencing factors， and use site observation data to calculate the temperature in the depth of the soil， but the model effect weakens as the soil depth increases.While deep learning model has a structure that can extract spatial features， it has high requirements on the amount of data， and is only used for remote sensing inversion of surface temperature in current research.In the study of soil temperature time series， the traditional machine learning method with periodic information has better performance， the sequence learning model in deep learning can automatically capture the law of soil temperature changes， and the hybrid model combined with the non-stationary sequence analysis method can fully consider the continuity and periodicity of soil temperature changes.Due to the complexity of land surface processes， there are few studies on the temporal and spatial variation of soil temperature.Based on model characteristics and research status， this paper summarizes the applicability of machine learning in soil temperature prediction， and discusses data selection， model selection， and model evaluation methods.Different data conditions and research purposes determine the choice of data and models.Decision Tree methods can provide a certain degree of interpretability through visualization.Support Vector Machines can be applied to situations with a small amount of data， and Extreme Learning Machines can meet the needs of fast computing.Due to the lack of physical constraints in machine learning， model testing and comparison of results should be emphasized when applied to soil temperature prediction.In view of the challenges of current research， the future work is prospected.The use of machine learning methods to predict soil temperature can be carried out from three aspects in the future： integrating prior scientific knowledge into the learning model， combining remote sensing data and multi-layer observations for soil temperature three-dimensional spatial modeling， and using convolutional recurrent neural networks for spatio-temporal modeling.

In this study, a novel temporal convolutional neural network (TCNN) model is developed for long-term streamflow projection in California within the Catchment Attributes for Large-Sample Studies (CAMELS) watershed regions. The TCNN model consists of several convolution blocks and causal convolution is used as physical constraint. The ensemble performance of the model is first compared with other machine learning models for streamflow prediction. The model is further assessed through comparison with reduced models and using different hyperparameters, with results suggesting that this model correctly ascertains the physical relationship between input variables and streamflow. The stability of the model and its behavior in the extrapolated regime is assessed through an idealized extreme test with quadruple precipitation and 5°C higher temperature. Future streamflow projections are then developed using daily high-resolution Localized Constructed Analogs dataset (LOCA). To understand the importance of the nonlinear machine learning approach, we estimate the degree of nonlinearity in the streamflow response among input variables. Our work shows the ability and potential for TCNNs to perform future hydrology projections.

Abstract Recent debate over the scope of the U.S. Clean Water Act underscores the need to develop a robust body of scientific work that defines the connectivity between freshwater systems and people. Coupled natural and human systems (CNHS) modeling is one tool that can be used to study the complex, reciprocal linkages between human actions and ecosystem processes. Well‐developed CNHS models exist at a conceptual level, but the mapping of these system representations in practice is limited in capturing these feedbacks. This article presents a paired conceptual–empirical methodology for functionally capturing feedbacks between human and natural systems in freshwater lake catchments, from human actions to the ecosystem and from the ecosystem back to human actions. We address extant challenges in CNHS modeling, which arise from differences in disciplinary approach, model structure, and spatiotemporal resolution, to connect a suite of models. In doing so, we create an integrated, multi‐disciplinary tool that captures diverse processes that operate at multiple scales, including land‐management decision‐making, hydrologic‐solute transport, aquatic nutrient cycling, and civic engagement. In this article, we build on this novel framework to advance cross‐disciplinary dialogue to move CNHS lake‐catchment modeling in a systematic direction and, ultimately, provide a foundation for smart decision‐making and policy.