Shujie Cheng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Shujie Cheng has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Lei Cheng
State Key Laboratory of Water Resources Engineering and Management Wuhan University Wuhan China
4 shared publications - Lu Zhang
State Key Laboratory of Water Resources Engineering and Management Wuhan University Wuhan China
3 shared publications - Pan Liu
State Key Laboratory of Water Resources Engineering and Management Wuhan University Wuhan China
3 shared publications - Shujing Qin
State Key Laboratory of Water Resources Engineering and Management Wuhan University Wuhan China
3 shared publications - Liu Liu
State Key Laboratory of Efficient Utilization of Agricultural Water Resources China Agricultural University Beijing China
2 shared publications - Akash Koppa
Department of Environmental Science and Technology, University of Maryland, College Park, MD 20742, USA
1 shared publication
Latest Publications
- Open Access
- 21 February 2024
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 60 (2)
Improving the Applicability of Lumped Hydrological Models by Integrating the Generalized Complementary Relationship
Lumped hydrological models (LHMs) are indispensable for water resource planning and environmental studies due to simple structures and robust performances. LHMs commonly focus on runoff processes with crude representations for other hydrological processes, such as evapotranspiration (E). Therefore, these models may yield unrealistic water balance partitioning. The challenge is to enhance the LHMs performance while retaining simplicity. The generalized complementary relationship (GCR) is a simple and robust method for estimating E. This study attempted to incorporate GCR into four widely used LHMs (Australian water balance model, GR2M, SIMHYD, and TANK) to test whether the integrated models (GCR‐LHMs) can improve runoff simulation at little cost to the model structure and data requirement. Original LHMs and integrated GCR‐LHMs were tested in 2112 catchments over various climatic conditions. Results show that the GCR‐LHMs outperform original LHMs in most catchments (77.7 ± 5.0%). In addition, the number of catchments that GCR‐LHMs have qualified performance (i.e., Kling‐Gupta coefficient [KGE] more than 0.5) increased by 10.7 ± 3.6% compared with LHMs. The performance of original and integrated models is dependent on the aridity index and normalized vegetation index. However, the improvement in model performance is less catchment characteristics dependent. These results indicate that incorporating GCR into the LHMs improves the model performance under different climatic and vegetation conditions and justifies the integration. GCR integration with LHMs can improve runoff estimation ability (with higher KGE and R‐Square) while retaining model simplicity and readily available input. These findings are valuable for improving the applicability and accuracy of LHMs.
- 21 February 2024
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 60 (2)
Open AccessImproving the Applicability of Lumped Hydrological Models by Integrating the Generalized Complementary Relationship
Lumped hydrological models (LHMs) are indispensable for water resource planning and environmental studies due to simple structures and robust performances. LHMs commonly focus on runoff processes with crude representations for other hydrological processes, such as evapotranspiration (E). Therefore, these models may yield unrealistic water balance partitioning. The challenge is to enhance the LHMs performance while retaining simplicity. The generalized complementary relationship (GCR) is a simple and robust method for estimating E. This study attempted to incorporate GCR into four widely used LHMs (Australian water balance model, GR2M, SIMHYD, and TANK) to test whether the integrated models (GCR‐LHMs) can improve runoff simulation at little cost to the model structure and data requirement. Original LHMs and integrated GCR‐LHMs were tested in 2112 catchments over various climatic conditions. Results show that the GCR‐LHMs outperform original LHMs in most catchments (77.7 ± 5.0%). In addition, the number of catchments that GCR‐LHMs have qualified performance (i.e., Kling‐Gupta coefficient [KGE] more than 0.5) increased by 10.7 ± 3.6% compared with LHMs. The performance of original and integrated models is dependent on the aridity index and normalized vegetation index. However, the improvement in model performance is less catchment characteristics dependent. These results indicate that incorporating GCR into the LHMs improves the model performance under different climatic and vegetation conditions and justifies the integration. GCR integration with LHMs can improve runoff estimation ability (with higher KGE and R‐Square) while retaining model simplicity and readily available input. These findings are valuable for improving the applicability and accuracy of LHMs.
- Open Access
- 25 May 2023
- Published by Wiley
Global runoff partitioning based on Budyko-constrained machine learning
Understanding the partitioning of runoff into baseflow and quickflow is crucial for informed decision-making in water resources management, guiding the implementation of flood mitigation strategies, and supporting the development of drought resilience measures. Methods that combine the physically-based Budyko framework with machine learning (ML) have shown promise in estimating global runoff. However, such ‘hybrid’ approaches have not been used for baseflow estimation. Here, we develop a Budyko-constrained ML approach for baseflow estimation by incorporating the Budyko-based baseflow coefficient (BFC) curve as a physical constraint. We estimate the parameters of the original Budyko curve and the newly developed BFC curve based on 13 climatic and physiographic characteristics using boosted regression trees (BRT). BRT models are trained and tested in 1226 catchments worldwide and subsequently applied to the entire global land surface at a 0.25° grid scale. The catchment-trained models exhibit strong performance during the testing phase, with R2 values of 0.96 and 0.88 for runoff and baseflow, respectively. Results reveal that, on average, 30.3% (spatial standard deviation std=26.5%) of the continental precipitation is partitioned into runoff, of which 20.6% (std=22.1%) is baseflow and 9.7% (std=10.3%) is quickflow. Among the 13 climatic and physiographic characteristics, topography and soil-related characteristics generally emerge as the most important drivers, although significant regional variability is observed. Comparisons with previous datasets suggest that global runoff partitioning is still highly uncertain and warrants further research.
- 25 May 2023
- Published by Wiley
Open AccessGlobal runoff partitioning based on Budyko-constrained machine learning
Understanding the partitioning of runoff into baseflow and quickflow is crucial for informed decision-making in water resources management, guiding the implementation of flood mitigation strategies, and supporting the development of drought resilience measures. Methods that combine the physically-based Budyko framework with machine learning (ML) have shown promise in estimating global runoff. However, such ‘hybrid’ approaches have not been used for baseflow estimation. Here, we develop a Budyko-constrained ML approach for baseflow estimation by incorporating the Budyko-based baseflow coefficient (BFC) curve as a physical constraint. We estimate the parameters of the original Budyko curve and the newly developed BFC curve based on 13 climatic and physiographic characteristics using boosted regression trees (BRT). BRT models are trained and tested in 1226 catchments worldwide and subsequently applied to the entire global land surface at a 0.25° grid scale. The catchment-trained models exhibit strong performance during the testing phase, with R2 values of 0.96 and 0.88 for runoff and baseflow, respectively. Results reveal that, on average, 30.3% (spatial standard deviation std=26.5%) of the continental precipitation is partitioned into runoff, of which 20.6% (std=22.1%) is baseflow and 9.7% (std=10.3%) is quickflow. Among the 13 climatic and physiographic characteristics, topography and soil-related characteristics generally emerge as the most important drivers, although significant regional variability is observed. Comparisons with previous datasets suggest that global runoff partitioning is still highly uncertain and warrants further research.
- 27 April 2022
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 58 (4)
Improved Understanding of How Catchment Properties Control Hydrological Partitioning Through Machine Learning
Long‐term hydrological partitioning of catchments can be well described by the Budyko framework with a parameter (e.g., Fu's equations with parameter ω). The Budyko framework considers aridity index as the dominant control on hydrological partitioning, while the parameter represents integrated influences of catchment properties. Our understanding regarding the controls of catchment properties on the parameter is still limited. In this study, two machine learning methods, that is, boosted regression tree (BRT) and CUBIST, were used to model ω. Interpretable machine learning methods were adopted for better physical understanding including feature importance, accumulated local effects (ALE), and local interpretable model‐agnostic explanations. Among the 15 properties of 443 Australian catchments, analysis of feature importance showed that root zone storage capacity (SR), vapor pressure, vegetation coverage (M), precipitation depth, climate seasonality and asynchrony index (SAI), and water use efficiency (WUE) were the six primary control factors on ω. ALE showed that ω varied nonlinearly with all factors, and varied non‐monotonically with M, SAI, and WUE. LIME showed that the importance of the six dominant factors on ω varied between regions. CUBIST was further used to build regionally varying relationships between ω and the primary factors. Continental scale ω and evapotranspiration were further mapped across Australia based on the most robust BRT‐trained parameterization scheme with a resolution of 0.05°. Instead of using the machine learning method as a black box, we employed interpretability approaches to identify the controls. Our findings not only improved the capability of the Budyko method for hydrological partitioning across Australia, but also demonstrated that the controls of catchment properties on hydrological partitioning vary in different regions.
- 27 April 2022
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 58 (4)
Improved Understanding of How Catchment Properties Control Hydrological Partitioning Through Machine Learning
Long‐term hydrological partitioning of catchments can be well described by the Budyko framework with a parameter (e.g., Fu's equations with parameter ω). The Budyko framework considers aridity index as the dominant control on hydrological partitioning, while the parameter represents integrated influences of catchment properties. Our understanding regarding the controls of catchment properties on the parameter is still limited. In this study, two machine learning methods, that is, boosted regression tree (BRT) and CUBIST, were used to model ω. Interpretable machine learning methods were adopted for better physical understanding including feature importance, accumulated local effects (ALE), and local interpretable model‐agnostic explanations. Among the 15 properties of 443 Australian catchments, analysis of feature importance showed that root zone storage capacity (SR), vapor pressure, vegetation coverage (M), precipitation depth, climate seasonality and asynchrony index (SAI), and water use efficiency (WUE) were the six primary control factors on ω. ALE showed that ω varied nonlinearly with all factors, and varied non‐monotonically with M, SAI, and WUE. LIME showed that the importance of the six dominant factors on ω varied between regions. CUBIST was further used to build regionally varying relationships between ω and the primary factors. Continental scale ω and evapotranspiration were further mapped across Australia based on the most robust BRT‐trained parameterization scheme with a resolution of 0.05°. Instead of using the machine learning method as a black box, we employed interpretability approaches to identify the controls. Our findings not only improved the capability of the Budyko method for hydrological partitioning across Australia, but also demonstrated that the controls of catchment properties on hydrological partitioning vary in different regions.
- Open Access
- 18 August 2021
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 57 (8)
An Analytical Baseflow Coefficient Curve for Depicting the Spatial Variability of Mean Annual Catchment Baseflow
Catchment baseflow is jointly controlled by climate and landscape properties. Previous studies have recognized that spatial variability of mean annual baseflow coefficient (BFC = , ratio of baseflow to precipitation) is primarily controlled by aridity index and storage capacity. However, an analytical solution of BFC in terms of the dominant controlling factors has not yet been established. The objective of this study was to develop an analytical BFC curve to depict spatial variability of BFC based on the “limit” concept of the Budyko framework. The BFC curve relates the baseflow coefficient to aridity index and storage capacity without resolving complex interactions between evapotranspiration and baseflow generation. The proposed BFC curve showed that, in the arid catchments, baseflow coefficient was primarily limited by available water (precipitation, P) and, in the humid catchments, was jointly controlled by both the available energy (potential evapotranspiration, Ep) and catchment retention capability (ratio of catchment storage capacity to P, i.e., Sp/P). Observed hydrological data from 950 catchments in Australia, the conterminous United States and the United Kingdom with diverse hydro‐climatic conditions (BFC = 0.001–0.650) were collected to demonstrate the capability of the developed curve. Results showed that the BFC curve captured the spatial variability of observed BFC in the 950 study catchments (R2 = 0.75, RMSE = 0.058). Mean annual baseflow estimated by the BFC curve agreed well with observed baseflow (R2 = 0.86, RMSE = 0.19 mm). The developed analytical curve provides an analytical solution for understanding how aridity index and storage capacity control mean annual catchment baseflow, and will improve predictability of baseflow at ungauged basins.
- 18 August 2021
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 57 (8)
Open AccessAn Analytical Baseflow Coefficient Curve for Depicting the Spatial Variability of Mean Annual Catchment Baseflow
Catchment baseflow is jointly controlled by climate and landscape properties. Previous studies have recognized that spatial variability of mean annual baseflow coefficient (BFC = , ratio of baseflow to precipitation) is primarily controlled by aridity index and storage capacity. However, an analytical solution of BFC in terms of the dominant controlling factors has not yet been established. The objective of this study was to develop an analytical BFC curve to depict spatial variability of BFC based on the “limit” concept of the Budyko framework. The BFC curve relates the baseflow coefficient to aridity index and storage capacity without resolving complex interactions between evapotranspiration and baseflow generation. The proposed BFC curve showed that, in the arid catchments, baseflow coefficient was primarily limited by available water (precipitation, P) and, in the humid catchments, was jointly controlled by both the available energy (potential evapotranspiration, Ep) and catchment retention capability (ratio of catchment storage capacity to P, i.e., Sp/P). Observed hydrological data from 950 catchments in Australia, the conterminous United States and the United Kingdom with diverse hydro‐climatic conditions (BFC = 0.001–0.650) were collected to demonstrate the capability of the developed curve. Results showed that the BFC curve captured the spatial variability of observed BFC in the 950 study catchments (R2 = 0.75, RMSE = 0.058). Mean annual baseflow estimated by the BFC curve agreed well with observed baseflow (R2 = 0.86, RMSE = 0.19 mm). The developed analytical curve provides an analytical solution for understanding how aridity index and storage capacity control mean annual catchment baseflow, and will improve predictability of baseflow at ungauged basins.