Farhad Bahmanpouri

Discharge is a fundamental hydrological variable, yet often unavailable under hard‐to‐access or critical conditions, hindering understanding of ungauged catchment hydrology. Unoccupied Aerial Systems (UAS) offer promising contactless approaches for monitoring inland waters under such conditions. Numerous studies have assessed UAS discharge estimation, but few have focused on evaluating fully contactless approaches across diverse river conditions. To address this gap, we present a UAS‐based framework for estimating discharge using contactless measurements of water surface velocity, water surface elevation, water surface slope and river bathymetry, supplemented by satellite observations and external DEMs where needed. Four surface‐to‐bulk velocity conversion methods: 0.85 coefficient, log law, power law and entropy method were applied. UAS discharge estimation accuracy was benchmarked at 46 cross‐sections across four river systems with gauge records. We quantified uncertainty propagation from each input variable, assessed their relative contributions, and performed an error propagation experiment to simulate high‐bias measurement scenarios. To identify factors controlling performance, we calculated correlations between discharge error, uncertainty metrics, and cross‐sectional properties. Across most cross‐sections, at least one method achieved discharge error less than 30%, though no single method consistently outperformed the others. The 0.85 method was robust with respect to input uncertainty, with performance patterns across methods linked to morphology: 0.85 and entropy methods performed generally better at larger rivers, while high water surface slopes reduced the accuracy of the log law and power law methods. Correlation analysis revealed that surface velocity uncertainty dominated discharge uncertainty in deep rivers, whereas water surface elevation was more influential in shallow ones.

Sediment transport governs the dynamic exchange of sediment and flow that underpins channel–floodplain interactions and is a primary controller of fluvial morphology in large alluvial braided rivers. Quantitative analysis of sediment flux is essential to determine whether the river bed is in a dynamic equilibrium or undergoing sedimentation and/or erosion under varying flow conditions. This work depicts the representation and understanding of the physical processes governing river morphodynamics by developing a fully coupled hydro-morphodynamic model for a braided reach of the Brahmaputra River at Nimatighat, India. The Van Rijn sediment transport model is coupled with the 2D Shallow Water Equations (SWEs) to calculate the hydraulic parameters and the bedload flux. For validation, a few benchmark case studies were taken from published literature. A good agreement with R2 = 0.99, RMSE = 0.002 m, and Pbias = 1.067% was found between the simulated results and experimental data across different scenarios. The framework was applied to the Brahmaputra River to investigate the equilibrium geomorphic response to anthropogenic interventions, including mid-channel sandbar dredging. The study further evaluated the model's ability to mitigate bank erosion in the vulnerable reach upstream of a bull-head structure along the outer bank. We found that dredging had minimal effect in controlling the erosion in the eroded bank due to the immediate sedimentation of the dredged bed to attain equilibrium.

Floods are among the most destructive natural hazards, posing a significant risk to lives and infrastructure worldwide. Effective flood risk management demands precise, multidimensional approaches. In this direction, the current study aims to evaluate and compare two methods for flood-risk assessment: the scenario-based Ordered Weighted Averaging (OWA) and the data-driven Random Forest (RF) method. To this end, the Great Karun watershed in Iran was chosen due to its complex hydrological and climatic conditions. Hydro-climatic, hydrological, topographic, land-cover datasets, and actual flood observations were applied and analyzed based on fifteen influencing factors recommended by expert opinion. In the OWA approach, while factor weights were determined using the Best-Worst Method, flood-risk maps were produced based on five scenarios: very optimistic, optimistic, intermediate, pessimistic, and very pessimistic. In the RF approach, factor importance index was calculated via the mean decrease impurity algorithm, and the model was trained to generate flood-risk maps. Results showed that distance from rivers and slope were the most influential factors in OWA, while precipitation and flow accumulation dominated in RF. Prediction rate for OWA scenarios ranged from 1.4 to 3.5%, while RF achieved 16.0%, and the Area Under the Curve (AUC) was 0.984. Optimistic scenarios overestimated, and pessimistic scenarios underestimated risk, with the OWA intermediate scenario most closely matching RF results. RF demonstrated superior performance for flood-risk classification, highlighting its applicability for precise flood management. Overall, this study presents a novel comparative framework by integrating scenario-based OWA-BWM and Random Forest approaches to investigate the effects of decision-maker preferences and data-driven learning on flood susceptibility mapping.

Non-contact image-based velocimetry methods provide a flexible, cost-effective, and safe alternative to standard river discharge measurement techniques, but require prior knowledge of the bathymetry and suitable tracer distributions. Image wave velocimetry (IWV) approaches estimate the flow velocity and water depth simultaneously without artificial tracers, but have a low spatial resolution and lack robustness. Here, we extend a previous IWV approach to reconstruct the spatial transverse distribution of the surface velocity and depth across a river and determine the water discharge. We introduce a novel physics-informed approach that integrates the IWV analysis in an entropy-based framework, improving the resolution and robustness of the reconstructions. Testing in a shallow stream showed a 10% error in discharge for the extended IWV method, which improved to 4% with the combined IWV-entropy method. The results demonstrate the unprecedented capability of IWV methods to measure the discharge of shallow rivers without requiring any contact with water.

The southern rim of the Indian Himalayas is highly susceptible to floods during the summer monsoon, making accurate streamflow modelling critical yet difficult due to complex terrain, climate variability, and sparse ground observations. This study uses a conceptual, semi-distributed hydrological model – enhanced with both static and dynamic glacier modules – to reproduce streamflow into the Alaknanda River at Rudraprayag gauge (∼8600 km ² ), a representative basin in northern India. The model was calibrated using multi-variable data, including satellite-based glacier water loss and actual evapotranspiration in addition to streamflow, also to address bias in the precipitation input. Despite inherent data uncertainties and simplified process conceptualization, the tailored hydrological modelling captured key features of observed streamflow and produced internally consistent water balance estimates. Multi-variable calibration provided a more plausible representation of hydrological processes and highlighted the value of using complementary satellite-based information in data-poor mountain regions. Parsimonious precipitation adjustment approaches are proven effective for hydrological applications. However, input data errors such as unaccounted-for heavy precipitation events limited short-term streamflow prediction accuracy. The study demonstrates that a viable, parsimonious modelling strategy can still be developed for data-scarce, monsoon-dominated Himalayan basins, offering insights into the spatiotemporal dynamics of streamflow generating processes, the inter-seasonal redistribution of precipitation, the role of cryosphere contributions, and flood simulation. The approach is transferable to other monsoon-dominated, glacier-influenced, and data-limited mountain catchments facing increasing hydroclimatic risks.

According to existing procedures for defining the velocity distribution across cross profile sections of watercourses (e.g., Entropy theory and Power Law theory), surface velocity is a key input parameter, together with cross-sectional bathymetry. Field measurements to obtain velocity values and their distributions are often difficult due to limited equipment, unreliable data, missing data, or hazardous conditions such as flooding and inaccessible locations. This creates a strong need for alternative approaches to measuring surface velocities in rivers. The application of unmanned aerial vehicles (UAVs), mobile phones, and traditional field instruments such as the Propeller Current Meter (PCM) can significantly improve measurement efficiency, especially in situations where conventional methods are not feasible. This paper presents an algorithm for comparing these measurement approaches and quantifying their differences. The methodology is demonstrated using a real case study on the Bednja River in Croatia, which flows through alluvial deposits. The results show that video-based surface velocity estimation using UAV and mobile phone imagery is feasible under real river conditions. Still, its accuracy depends strongly on flow conditions and surface characteristics. While UAV recordings provide reliable results in fast and turbulent flows, mobile phone videos yield more stable performance in smoother flow conditions, where additional surface texture is available from natural tracers.

This study investigates the spatial and temporal variations in salinity in the Jarahi River and its traditional channels using field measurements and numerical simulations. The primary objective is to assess the effectiveness of different management strategies for salinity reduction under minimum-discharge conditions. Salinity dynamics were analyzed through electrical conductivity (EC) measurements collected over a one-year period and simulated using the MIKE 11 hydrodynamic model. Model performance was evaluated by comparing simulated and observed EC values at key monitoring stations. The results indicate that maximum salinity levels occur during March and April in both the main river and traditional channels, while the highest temporal variability in EC was observed in October. The comparison between observed and simulated data showed a relative error of less than 10%, confirming the reliability of the model simulations. Four management scenarios were evaluated: (1) preventing inflow from the Motbeg drainage, (2) blocking non-centralized drainage inputs, (3) removing all inlet drains, and (4) increasing discharge releases from the Ramshir Dam. The first and third scenarios led to the highest salinity reductions, reaching up to 39% (approximately 1266 µS/cm) in the Gorgor channel, while reductions of up to 53% were observed in traditional streams such as Mansuri and Omal-Sakher under the third scenario. Increasing dam releases resulted in a maximum reduction of 23% (724 µS/cm) at the Gorgor station. Finally, the proposed management strategies significantly reduced salinity levels along the river system, particularly at the entrance of the Jahangiri traditional stream, providing practical insights for salinity control and river basin management.

The southern rim of the Indian Himalayas is highly susceptible to floods during the summer monsoon, making accurate streamflow modelling critical yet difficult due to complex terrain, climate variability, and sparse ground observations. This study uses a conceptual, semi-distributed hydrological model – enhanced with both static and dynamic glacier modules – to reproduce streamflow into the Alaknanda River at Rudraprayag gauge (~8600 km²), a representative basin in northern India. The model was calibrated using multi-variable data, including satellite-based glacier water loss and actual evapotranspiration, also to address bias in the precipitation input. Despite inherent data uncertainties and simplified process conceptualization, the tailored hydrological modelling captured key features of observed streamflow and produced internally consistent water balance estimates. Multi-variable calibration improved the simulation of hydrological fluxes and highlighted the value of using complementary satellite-based information in data-poor mountain regions. Parsimonious precipitation adjustment approaches are proven effective for hydrological applications. However, input data errors such as unaccounted-for heavy precipitation events limited short-term streamflow prediction accuracy. The study demonstrates that a viable, parsimonious modelling strategy can still be developed for data-scarce, monsoon-dominated Himalayan basins, offering insights into the spatiotemporal dynamics of streamflow generating processes, the inter-seasonal redistribution of precipitation, the role of cryosphere contributions, and flood simulation. The approach is transferable to other monsoon-dominated, glacier-influenced, and data-limited mountain catchments facing increasing hydroclimatic risks.