Hydrological Inference Research Articles

Achieving water budget closure through physical hydrological process modelling: insights from a large-sample study

Abstract. Modern hydrology is embracing a data-intensive new era, with information from diverse sources currently providing support for hydrological inferences at broader scales. This results in a plethora of data-reliability-related challenges that remain unsolved. The water budget non-closure is a widely reported phenomenon in hydrological and atmospheric systems. Many existing methods aim to enforce water budget closure constraints through data fusion and bias correction approaches, often neglecting the physical interconnections between water budget components. To solve this problem, this study proposes a Multisource Dataset Correction Framework grounded in Physical Hydrological Process Modelling to enhance water budget closure, termed the PHPM-MDCF. The concept of decomposing the total water budget residuals into inconsistency and omission residuals is embedded in this framework to account for different residual sources. We examined the efficiency of the PHPM-MDCF and the distribution of residuals across 475 contiguous United States (CONUS) basins selected by hydrological simulation reliability. The results indicate that the inconsistency residuals dominate the total water budget residuals, exhibiting highly consistent spatiotemporal patterns. This portion of residuals can be significantly reduced through PHPM-MDCF correction and achieved satisfactory efficiency. The total water budget residuals decreased by 49 %, on average, across all basins, with reductions exceeding 80 % in certain basins. The credibility of the correction framework was further verified through noise experiments and comparisons with existing methods. In the end, we explored the potential factors influencing the distribution of residuals and found notable scale effects, along with the key role of hydro-meteorological conditions. This emphasizes the importance of carefully evaluating the water balance assumption when employing multisource datasets for hydrological inference in small and humid basins.

Paradoxes and hydrological inferences for the desert valleys extending between Haditha and Al-Ramadi using remote sensing data and geographic information systems.

Paradoxes and hydrological inferences for the desert valleys extending between Haditha and Al-Ramadi using remote sensing data and geographic information systems.

Hydrological Inferences of Morphometric Parameters for Management of River Sub-basin in Semi-arid Tropics of India using Geospatial Technique

The runoff and sediment produced from a river basin due to effective precipitation depends on the characteristics of its catchment. The catchment characteristic is site-specific and varies from basin to basin. It depends upon the topographic characteristics, climatic parameters, soil properties, geology, and human-induced activities of the catchment. Knowledge of these factors is essential for sustainable management of natural resources, i.e. soil and water and site-specific conservation measures of the basin or watershed. Morphometric parameters of the basin truly reflect the combination of all characteristics, and this is one of the major indicators for the hydrological behaviour of any watershed or catchment. For the present study, Pairi River sub-basin, situated between 81°51'58.7"- 82°26'16.04" E longitude to 19°55'51.9"- 20°55'6.9" N latitude, has been considered for the morphometric analysis using geospatial technique . Results of morphometric analysis revealed that the Pairi sub-basin has seventh order drainage network with a dendritic pattern. The dendritic drainage pattern of the subbasin shows that the basin has homogeneous lithology and a lack of structural control. The drainage density and stream frequency indicates that the sub-basin has a gentle slope and good permeability of sub-surface material. So, this sub-basin is suitable for constructing artificial groundwater recharge structures to improve the groundwater resource. The length of overland flow revealed that surface runoff may reach the stream more quickly within the sub-basin. The bifurcation ratio varies different successive stream orders. The mean bifurcation ratio is low, indicating that the geologic structure does not distort the drainage pattern and produces low to moderate peak flow in the subbasin. The shape parameters such as form factor, circulatory ratioand elongation ratio indicate that the sub-basin is elongated in shape, high infiltration capacity, and flatter hydrograph with low to moderate runoff. This study can be useful for developing and managing natural resources of the Pairi sub-basin, like site-specific soil and water conservation measures such as check dam, percolation tank, water harvesting structure and development and management of groundwater, etc

Hydrological Inferences from Watershed Analysis for Sustainable Water Management Using Geoinformatics Approach

The rapid growth of urbanization, industrialization, an exponential increase in population and change of climate along with uneven distribution of rainfall make proper water management and planning of storage is very challenging. Satellite-based geoinformatics technology has supported to be an efficient tool in the analysis of drainage networks, surface morphological features, and interrelation with groundwater management at the watershed level. Geoinformatics approach such as remote sensing and Geographic information system has used for extraction of watershed analysis using Cartosat Digital Elevation Model (DEM) satellite images for assessment of drainage and extraction of their relative parameters For the Darna watershed in Nashik district, Maharashtra. In the study Hydrological parameters drainage analysis, topographical parameters, and land-use patterns has evaluated and interpreted for sustainable watershed management. The results reveals that Cartosat Data Dem based hydrological evaluation of watershed-scale is applied and accurate compared to available various technique. Keywords: CARTOSAT-DEM, Geoinformatics, LISS­IV, Darna Basin, Watershed

Development of a physics-informed data-driven model for gaining insights into hydrological processes in irrigated watersheds

• Process-based baseflow separation method can help trace the streamflow origins from the perspective of sub-hydrological processes. • Multilevel factorial analysis is applied to support the hydrological inference through stepwise clustered ensemble. • Hydrologicla inferences achieved from WFI method can lead to physically consistent results. Data-driven hydrological modeling has seen rapid development in recent years owing to its flexibility to approximate the complex relationships between driving forces and hydrological fluxes. However, traditional data-driven models typically cannot simultaneously capture the processes that pose both chronic and acute impacts on streamflow, thus impeding further inference. Therefore, this study presents a baseflow-filtered hydrological inference model to gain insights into hydrological processes in irrigated watersheds. The proposed model starts with separating the streamflow process into two sub-processes using a process-based baseflow separation method. Each sub-process is simulated through a new interpretable data-driven model. The resulting hydrological inferences facilitate the identification of the dominant factors influencing flows in saturated and unsaturated zones. The proposed model is applied to three irrigated watersheds, and the evaluation metrics show that the proposed model outperforms two conventional data-driven models. Our findings reveal that predictors associated with air temperature and long-term (i.e., monthly) irrigation are mainly responsible for characterizing baseflow dynamics, while precipitation and short-term (i.e., semi-weekly or weekly) irrigation are primarily responsible for describing overland flow and interflow dynamics. The fidelity of the derived hydrological inference is further demonstrated through sensitivity analysis. The results show that the relative importance of predictors not only reflects their significance on model performance, but also influence the changes on streamflow.

Assessment of Morphometric Parameters as the Basis for Hydrological Inferences in Water Resource Management: A Case Study from the Sinú River Basin in Colombia

The geomorphology of a basin makes it possible for us to understand its hydrological pattern. Accordingly, satellite-based remote sensing and geo-information technologies have proven to be effective tools in the morphology analysis at the basin level. Consequently, this present study carried out a morphological analysis of the Sinú river basin, analyzing its geometric characteristics, drainage networks, and relief to develop integrated water resource management. The analyzed zone comprises an area of 13,971.7 km2 with three sub-basins, the upper, the middle, and the lower Sinú sub-basins, where seventeen morphometric parameters were evaluated using remote sensing (RS) and geographical information system (GIS) tools to identify the rainwater harvesting potential index. The Sinú basin has a dendritic drainage pattern, and the results of the drainage network parameters make it possible for us to infer that the middle and lower Sinú areas are the ones mainly affected by floods. The basin geometry parameters indicate an elongated shape, implying a lesser probability of uniform and homogeneous rainfall. Additionally, the hypsometric curve shape indicates that active fluvial and alluvial sedimentary processes are present, allowing us to conclude that much of the material has been eroded and deposited in the basin’s lower zones as it could be confirmed with the geological information available. The obtained results and GIS tools confirm the basin’s geological heterogeneity. Furthermore, they were used to delimit the potential water harvesting zones following the rainwater harvesting potential index (RWHPI) methodology. The research demonstrates that drainage morphometry has a substantial impact on understanding landform processes, soil characteristics, and erosional characteristics. Additionally, the results help us understand the relationship between hydrological variables and geomorphological parameters as guidance and/or decision-making instruments for the competent authorities to establish actions for the sustainable development of the basin, flood control, water supply planning, water budgeting, and disaster mitigation within the Sinú river basin.

Development of a Wilks feature importance method with improved variable rankings for supporting hydrological inference and modelling

Abstract. Feature importance has been a popular approach for machine learning models to investigate the relative significance of model predictors. In this study, we developed a Wilks feature importance (WFI) method for hydrological inference. Compared with conventional feature importance methods such as permutation feature importance (PFI) and mean decrease impurity (MDI), the proposed WFI aims to provide more reliable variable rankings for hydrological inference. To achieve this, WFI measures the importance scores based on Wilks Λ (a test statistic that can be used to distinguish the differences between two or more groups of variables) throughout an inference tree. Compared with PFI and MDI methods, WFI does not rely on any performance measures to evaluate variable rankings, which can thus result in less biased criteria selection during the tree deduction process. The proposed WFI was tested by simulating monthly streamflows for 673 basins in the United States and applied to three interconnected irrigated watersheds located in the Yellow River basin, China, through concrete simulations for their daily streamflows. Our results indicated that the WFI could generate stable variable rankings in response to the reduction of irrelevant predictors. In addition, the WFI-selected predictors helped random forest (RF) achieve its optimum predictive accuracy, which indicates that the proposed WFI could identify more informative predictors than other feature importance measures.

Hydrologically informed machine learning for rainfall–runoff modelling: towards distributed modelling

Abstract. Despite showing great success of applications in many commercial fields, machine learning and data science models generally show limited success in many scientific fields, including hydrology (Karpatne et al., 2017). The approach is often criticized for its lack of interpretability and physical consistency. This has led to the emergence of new modelling paradigms, such as theory-guided data science (TGDS) and physics-informed machine learning. The motivation behind such approaches is to improve the physical meaningfulness of machine learning models by blending existing scientific knowledge with learning algorithms. Following the same principles in our prior work (Chadalawada et al., 2020), a new model induction framework was founded on genetic programming (GP), namely the Machine Learning Rainfall–Runoff Model Induction (ML-RR-MI) toolkit. ML-RR-MI is capable of developing fully fledged lumped conceptual rainfall–runoff models for a watershed of interest using the building blocks of two flexible rainfall–runoff modelling frameworks. In this study, we extend ML-RR-MI towards inducing semi-distributed rainfall–runoff models. The meaningfulness and reliability of hydrological inferences gained from lumped models may tend to deteriorate within large catchments where the spatial heterogeneity of forcing variables and watershed properties is significant. This was the motivation behind developing our machine learning approach for distributed rainfall–runoff modelling titled Machine Induction Knowledge Augmented – System Hydrologique Asiatique (MIKA-SHA). MIKA-SHA captures spatial variabilities and automatically induces rainfall–runoff models for the catchment of interest without any explicit user selections. Currently, MIKA-SHA learns models utilizing the model building components of two flexible modelling frameworks. However, the proposed framework can be coupled with any internally coherent collection of building blocks. MIKA-SHA's model induction capabilities have been tested on the Rappahannock River basin near Fredericksburg, Virginia, USA. MIKA-SHA builds and tests many model configurations using the model building components of the two flexible modelling frameworks and quantitatively identifies the optimal model for the watershed of concern. In this study, MIKA-SHA is utilized to identify two optimal models (one from each flexible modelling framework) to capture the runoff dynamics of the Rappahannock River basin. Both optimal models achieve high-efficiency values in hydrograph predictions (both at catchment and subcatchment outlets) and good visual matches with the observed runoff response of the catchment. Furthermore, the resulting model architectures are compatible with previously reported research findings and fieldwork insights of the watershed and are readily interpretable by hydrologists. MIKA-SHA-induced semi-distributed model performances were compared against existing lumped model performances for the same basin. MIKA-SHA-induced optimal models outperform the lumped models used in this study in terms of efficiency values while benefitting hydrologists with more meaningful hydrological inferences about the runoff dynamics of the Rappahannock River basin.

Atmospheric vapour and precipitation are not in isotopic equilibrium in a continental mountain environment

AbstractIsotopic exchange with atmospheric vapour can strongly influence the isotopic values of evaporating surface water bodies (e.g., lakes), influencing our understanding of hydrological processes across aquatic and terrestrial environments. Rather than measure the isotopic values of the atmosphere directly, it is much more common to estimate values by assuming equilibrium with local precipitation. This assumption may introduce large errors, thereby biasing hydrological inferences and understanding. The pattern and magnitude of this error has been quantified only in a few circumstances. We compared observations of vapour and precipitation isotope values over a four‐year record collected in a montane environment in the central Rocky Mountains of North America. We further investigated factors and conditions promoting disequilibrium. Scenario comparisons assessed the impact of theoretical and methodological elements on the accuracy of the equilibrium assumption. We found that the equilibrium assumption was not well supported by direct and continuous observations of vapour isotopes using tower‐based laser isotope spectroscopy, particularly during the summer months. Across all scenarios, errors associated with the equilibrium assumption were high, credibly ranging from 14 to 154 ‰ for δ2H and 1.5 to 16.3 ‰ for δ18O. Environmental covariates (e.g., vapour pressure deficit, air pressure) helped explain some of the apparent disequilibrium. Although the equilibrium assumption for estimating atmospheric vapour isotope values may not be applicable in a continental montane environment, our findings highlight opportunities for using direct vapour isotope measurements to better understand vapour sources, air mass mixing, and phase changes over complex mountainous terrain, which in turn may better constrain regional‐ to global‐scale hydrological process estimates, such as evapotranspiration rates and the water budgets of mountain lakes.

Challenges in Applying Machine Learning Models for Hydrological Inference: A Case Study for Flooding Events Across Germany

AbstractMachine learning (ML) algorithms are being increasingly used in Earth and Environmental modeling studies owing to the ever‐increasing availability of diverse data sets and computational resources as well as advancement in ML algorithms. Despite advances in their predictive accuracy, the usefulness of ML algorithms for inference remains elusive. In this study, we employ two popular ML algorithms, artificial neural networks and random forest, to analyze a large data set of flood events across Germany with the goals to analyze their predictive accuracy and their usability to provide insights to hydrologic system functioning. The results of the ML algorithms are contrasted against a parametric approach based on multiple linear regression. For analysis, we employ a model‐agnostic framework named Permuted Feature Importance to derive the influence of models' predictors. This allows us to compare the results of different algorithms for the first time in the context of hydrology. Our main findings are that (1) the ML models achieve higher prediction accuracy than linear regression, (2) the results reflect basic hydrological principles, but (3) further inference is hindered by the heterogeneity of results across algorithms. Thus, we conclude that the problem of equifinality as known from classical hydrological modeling also exists for ML and severely hampers its potential for inference. To account for the observed problems, we propose that when employing ML for inference, this should be made by using multiple algorithms and multiple methods, of which the latter should be embedded in a cross‐validation routine.

Assessing the Spatio-Temporal Pattern of Land Use and Land Cover Changes in Osun Drainage Basin, Nigeria

Abstract Over the years, Osun drainage basin has witnessed tremendous increase in population, and urbanization that have changed the landscape of the area. This study evaluated the spatio-temporal pattern of land use/land cover change (LULC) in the study area, and made hydrological inferences. Landsat imageries were acquired from USGS-EROS satellite image database for the period 1984, 2000 and 2015, while the Digital Elevation Model (DEM) was obtained from Shuttle Radar Topography Mission (SRTM) of the National Aeronautics and Space Agency (NASA). Supervised image classification using the Maximum Likelihood Algorithm in Erdas Imagine was adopted to classified the land use/land cover of the study area into seven classes. Elevation, aspect and slope of the study area were processed from DEM using ArcGIS. Modules for Land Use Change Evaluation (MOLUSCE) plugin in QGIS was used to simulate the basin future LULC change, using change driving factors of population, elevation, aspect and slope of the study area. There was about 234% increase in built up areas and 89.22% in crop/shrubs between 1984 and 2015. The most significant decrease in LULC occurred in forest (58.75%) and wetland (84.69%) during this period. The predicted future LULC change suggests that only about 12% of the basin will remain under forest cover by the year 2046. The results underscored the increasing anthropogenic activities in the basin that influenced recharge rate, surface runoff, incidences of soil erosion, etc., in Osun drainage basin. The planting of the lost native trees was recommended for the sustainability of the basin’s ecosystem.

Morphometric Analysis and Hydrological Inference for Water Resource Management in Atrai-Sib River Basin, NW Bangladesh Using Remote Sensing and GIS Technique

ABSTRACT Demand for irrigation water increases day by day along with meteorological vagaries and extension of irrigated area in the drought-prone Barind area of Bangladesh. This increasing stress on water resource is gradually making the area water scare. The study is aimed at studying the morphometric parameters of the Atrai-Sib river basin in the Barind area and on their relevance in water resource management based on satellite images and SRTM DEM. Computation and delineation of linear and areal aspects of the river basin and its morphometric components reveals that stream order ranges from first to eighth order showing dendritic drainage pattern. The basin represents homogeneity of soil texture; possibility of flash flood after heavy rainfall with low discharge of runoff; and is not largely affected by structural disturbance. Moderate drainage density of the river basin area indicates semi-permeable soil lithology with moderate vegetation. Mean bifurcation ratio of the basin is calculated as 3.92 and elongation ratio 0.75, which indicate elongated shape of the river basin with low to moderate relief bounded in the east and west by ‘moderate to steep’ sloping land area. It reveals a flatter peak of runoff flow for longer duration and gravity flow of water. The gentle but undulating slope of the basin represents ‘excellent’ category for groundwater management as the site is favorable for infiltration due to maximum time of runoff water percolation. The east facing slopes of the basin show higher moisture content and higher vegetation than the west-facing slope. The land use pattern of the area shows that major part (95.29%) comes under the cultivated land which will support future river basin development and management. Results obtained from the study would be useful in categorization of river basins for future water resource development and management, and selection of suitable sites for water conservation structures such as check dam, percolation tank, artificial recharge of groundwater through MAR technique etc.

Hydrological inferences through morphometric analysis of lower Kosi river basin of India for water resource management based on remote sensing data

Satellite based remote sensing technology has proven to be an effectual tool in analysis of drainage networks, study of surface morphological features and their correlation with groundwater management prospect at basin level. The present study highlights the effectiveness and advantage of remote sensing and GIS-based analysis for quantitative and qualitative assessment of flood plain region of lower Kosi river basin based on morphometric analysis. In this study, ASTER DEM is used to extract the vital hydrological parameters of lower Kosi river basin in ARC GIS software. Morphometric parameters, e.g., stream order, stream length, bifurcation ratio, drainage density, drainage frequency, drainage texture, form factor, circularity ratio, elongation ratio, etc., have been calculated for the Kosi basin and their hydrological inferences were discussed. Most of the morphometric parameters such as bifurcation ratio, drainage density, drainage frequency, drainage texture concluded that basin has good prospect for water management program for various purposes and also generated data base that can provide scientific information for site selection of water-harvesting structures and flood management activities in the basin. Land use land cover (LULC) of the basin were also prepared from Landsat data of 2005, 2010 and 2015 to assess the change in dynamic of the basin and these layers are very noteworthy for further watershed prioritization.

Prospective Interest of Deep Learning for Hydrological Inference.

Article impact statement: Deep learning may assist hydrological inference through its increased capacities to interpolate, classify, and hierarchically model databases.

Analysis of Hydrological Inferences through Morphometric Analysis, A Remote Sensing- GIS Based Study of Kanakanala Reservoir Subwatershed

Analysis of Hydrological Inferences through Morphometric Analysis, A Remote Sensing- GIS Based Study of Kanakanala Reservoir Subwatershed

Describing the catchment‐averaged precipitation as a stochastic process improves parameter and input estimation

AbstractRainfall input uncertainty is one of the major concerns in hydrological modeling. Unfortunately, during inference, input errors are usually neglected, which can lead to biased parameters and implausible predictions. Rainfall multipliers can reduce this problem but still fail when the observed input (precipitation) has a different temporal pattern from the true one or if the true nonzero input is not detected. In this study, we propose an improved input error model which is able to overcome these challenges and to assess and reduce input uncertainty. We formulate the average precipitation over the watershed as a stochastic input process (SIP) and, together with a model of the hydrosystem, include it in the likelihood function. During statistical inference, we use “noisy” input (rainfall) and output (runoff) data to learn about the “true” rainfall, model parameters, and runoff. We test the methodology with the rainfall‐discharge dynamics of a small urban catchment. To assess its advantages, we compare SIP with simpler methods of describing uncertainty within statistical inference: (i) standard least squares (LS), (ii) bias description (BD), and (iii) rainfall multipliers (RM). We also compare two scenarios: accurate versus inaccurate forcing data. Results show that when inferring the input with SIP and using inaccurate forcing data, the whole‐catchment precipitation can still be realistically estimated and thus physical parameters can be “protected” from the corrupting impact of input errors. While correcting the output rather than the input, BD inferred similarly unbiased parameters. This is not the case with LS and RM. During validation, SIP also delivers realistic uncertainty intervals for both rainfall and runoff. Thus, the technique presented is a significant step toward better quantifying input uncertainty in hydrological inference. As a next step, SIP will have to be combined with a technique addressing model structure uncertainty.

Impacts of future climate change on river discharge based on hydrological inference: A case study of the Grand River Watershed in Ontario, Canada

Over the recent years, climate change impacts have been increasingly studied at the watershed scale. However, the impact assessment is strongly dependent upon the performance of the climatic and hydrological models. This study developed a two-step method to assess climate change impacts on water resources based on the Providing Regional Climates for Impacts Studies (PRECIS) modeling system and a Hydrological Inference Model (HIM). PRECIS runs provided future temperature and precipitation projections for the watershed under the Intergovernmental Panel on Climate Change SRES A2 and B2 emission scenarios. The HIM based on stepwise cluster analysis is developed to imitate the complex nonlinear relationships between climate input variables and targeted hydrological variables. Its robust mathematical structure and flexibility in predictor selection makes it a desirable tool for fully utilizing various climate modeling outputs. Although PRECIS and HIM cannot fully cover the uncertainties in hydro-climate modeling, they could provide efficient decision support for investigating the impacts of climate change on water resources. The proposed method is applied to the Grand River Watershed in Ontario, Canada. The model performance is demonstrated with comparison to observation data from the watershed during the period 1972–2006. Future river discharge intervals that accommodate uncertainties in hydro-climatic modeling are presented and future river discharge variations are analyzed. The results indicate that even though the total annual precipitation would not change significantly in the future, the inter-annual distribution is very likely to be altered. The water availability is expected to increase in Winter while it is very likely to decrease in Summer over the Grand River Watershed, and adaptation strategies would be necessary.

Development of a Stepwise-Clustered Hydrological Inference Model

AbstractFlow prediction is one of the most important issues in modern hydrology. In this study, a statistical tool, stepwise-clustered hydrological inference (SCHI) model, was developed for daily streamflow forecasting. The SCHI model uses cluster trees to represent the nonlinear and complex relationships between streamflow and multiple factors related to climate and watershed conditions. It allows a great deal of flexibility in watershed configuration. The proposed model was applied to the daily streamflow forecasting in the Xiangxi River watershed, China. The correlation coefficient for calibration (1991–1995) was 0.881, and that for validation (1996–1998) was 0.771. Nash–Sutcliffe efficiencies for calibration and validation were 0.768 and 0.577, respectively. The results were compared to those of a conventional process-based model, and it was found that the SCHI model had a superior performance. The results indicate that the proposed model could provide not only reliable and efficient daily flow predic...

Hydrological inferences from watershed analysis for water resource management using remote sensing and GIS techniques

The present study highlights the importance of Digital Elevation Model (DEM) and satellite images for assessment of drainage and extraction of their relative parameters for the Orr watershed Ashok Nagar district, M.P., India. Hydrological parameters such as drainage analysis, topographic parameters and land use pattern were evaluated and interpreted for watershed management of the area.Hydrological module of ARC GIS software was utilized for calculation and delineation of the watershed and morphometric analysis of the watershed using SRTM DEM. The stream order of watershed ranges from first to sixth order showing dendritic type drainage network which is a sign of the homogeneity in texture and lack of structural control of the watershed. The drainage density in the area has been found to be low to medium which indicates that the area possesses highly permeable soils and low relief. The bifurcation ratio varies from 4.74 to 5 and the elongation ratio is 0.58 which reveals that the basin belongs to the elongated shaped basin category. The mean Rb of the entire basin is 4.62 which indicates that the drainage pattern is not much influenced by geological structures. Land use map of the watershed was generated from latest available multispectral satellite data and whole watershed covers under agricultural land, settlement, fallow land, forest, mining areas and water body.The present study reveals that SRTM DEM based hydrological evaluation at watershed scale is more applied and precise compared to other available techniques.

On red herrings and real herrings: disinformation and information in hydrological inference

On red herrings and real herrings: disinformation and information in hydrological inference