The geomorphic flood index (GFI) method provides a good representation of flood‐prone areas. However, the method does not account for floodwater transfers in undefined interbasins (UIBs), which represent intercluded small basins along the coastline likely to be flooded by adjacent major rivers. The present work addresses this shortcoming by complementing the GFI approach with an iterative procedure that considers UIBs and water transfers between basins. The methodology was tested on a coastal basin in southern Italy and the outcome was compared with a flood map obtained by a two‐dimensional hydraulic simulation. GFI performance as a morphological descriptor improved from 74% (standard method) to 94% with the addition of the iterative procedure. The proposed methodology, with the same parameterization, was applied on a second adjacent coastal basin obtaining improvements both in terms of true positive (from 56 to 79%) and false negative rates (from 44 to 21%). Finally, a sensitivity analysis to the flood return periods highlighted a strong influence on model parameterization for return periods below 20 years. This achievement represents a new development in the application of the GFI method, which can help stakeholders in a more time‐ and cost‐effective flood risk management in hazard‐prone areas.
In the present work, the use of optical cameras for turbidity measurements is tested on the Bode River in Germany, which is one of the best-instrumented catchments in Central Germany with a long-term time series on water quantity and quality. Four trap cameras have been installed on monitored cross-sections with the aim to explore the potential of RGB indices for the description of water turbidity. A description of the experimental setup and some preliminary results are introduced.
8 September 2025 in Hydrology and Earth System Sciences
The development of continuous river turbidity monitoring systems is essential since this is a critical water quality metric linked to the presence of organic and inorganic suspended matter. Current monitoring practices are mainly limited by low spatial and temporal resolutions and costs. This results in the huge challenge of providing extensive and timely water quality monitoring at the global scale. In this work, we propose an image analysis procedure for river turbidity assessment using different camera systems (i.e. fixed-trap camera, camera on board an uncrewed aerial vehicle and a multispectral camera). We explored multiple types of camera installation setups during a river turbidity event artificially re-created on site. The outcomes prove that processed digital-camera data can properly represent the turbidity trends. Specifically, the experimental activities revealed that single-band values were the most reliable proxy for turbidity monitoring in the short term, more so than band ratios and indexes. The best camera positioning, orientation and lens sensitivity, as well as daily and seasonal changes in lightning and river flow conditions, may affect the accuracy of the results. The reliability of this application will be tested under different hydrological and environmental conditions during our next field experiments. The final goal of the work is the implementation of this camera system to support existing techniques and to help in finding innovative solutions to water resource monitoring.
Using large image data sets has been the conventional strategy to improve object detection, but it increases annotation effort and training cost and does not guarantee robust transfer to new sites. Here we quantify the value of a small, diverse training set for floating macroplastic detection by jointly evaluating performance, computational cost, annotation effort, and cross‐site transferability. We compile four river‐camera data sets from Indonesia, The Netherlands, and Vietnam (training/internal validation) and Italy (external validation, single site and single day data from a long‐term camera monitoring system), harmonized into 13 litter classes and a five‐level tiering scheme (progressive class aggregation). We train YOLOv7 and YOLOv8 models and compare site‐specific data sets with a merged “Mixed” data set (999 images) spanning heterogeneous environmental conditions. Results show that data sets with more diverse backgrounds (Type II; e.g., The Netherlands) achieve higher performance per annotation than homogeneous data sets (Type I; e.g., Indonesia, Vietnam), whereas naïvely merging data sets can degrade internal validation unless accompanied by feature‐aware filtering. Class aggregation substantially increases overall detection skill, with gains consistent across data sets when moving from fine (Tier 4) to coarse (Tier 0) label spaces. Finally, internal validation does not reliably predict external‐site performance, underscoring the need for transferability‐aware data set design and evaluation. Overall, our findings emphasize that data diversity and curation , rather than data set size alone, are key levers to scale river plastic detection toward broader deployment.
The design of hydraulic structures needs to account for a trade-off between implementation costs and flood damages, as well as for the impacts on basins hydrological responses over a wide spectrum of events. In this work, a new methodology for dimensioning an in-line detention dam that integrates geomorphic, probabilistic and economic modeling is proposed. It is formulated as an economic optimization problem aimed at minimizing the sum of the construction cost and the cost of the residual flood risk on residential buildings. The optimization procedure was applied to a hypothetical in-line detention dam located upstream of the urban area of Castellammare di Stabia (Naples, Italy).