About the Project
Background:
The EU Water Framework Directive (WFD) identifies the characterisation of river hydromorphology as a critical component for assessing good ecosystem quality (2000/60/EC). In 2012, 56% of European rivers were of a moderate or lesser ecological stat, and all 22 EU member states reported hydromophology as a major pressure. Fundamental to this is the determination of geomorphological variables including channel width, slope, depth; bed and bank characteristics; sediment transport (erosion and deposition); and riparian vegetation. Such data is not routinely collected beyond academic studies (with partial exception of the River Habitat Survey conducted by the Environment Agency). This information enables typology of river physical habitat, including habitat quality (e.g. via biotope mapping), identifying pool-riffle sequences, backwaters, gravel bars, woody habitat, etc. which provides the foundation for habitat suitability classification, and underpins river restoration initiatives. Traditionally, this information is derived via conventional aerial photography, direct measurement of river cross-sections, and walk-over surveys for identification of habitat indicators. However, high resolution geospatial information is of increasing relevance, and offers an intelligent and cost-effective means of deriving the above variables at high spatial resolution and through non-contact means. Unmanned aerial vehicle (UAV) platforms, coupled with developments in photogrammetric computer vision algorithms are of particular relevance, offering the potential to provide a flexible and low-cost approach for characterisation at the reach scale. This is of importance for detailed assessment for a range of applications such as hydraulic modelling, habitat suitability and longer-term monitoring of ecosystem quality.
UAV-based survey techniques have been implemented for a range of environmental applications, most commonly utilising compact digital cameras to generate digital elevation models (DEMs) and ortho-imagery for mapping purposes. The potential of UAV imagery for remote measurement of channel bathymetry (depth) has also been demonstrated, making this technique especially attractive for fluvial applications. UAVs enable provision of imagery at very high spatial resolution, typically ~4 cm. However, intelligent processing of such imagery is relatively immature, and challenges exist in terms of i) efficient processing of massive data volumes; ii) automated image segmentation and classification at high spatial resolution; and iii) data collection strategies for optimising image quality. In the context of hydromorphological characterisation, photogrammetric computer vision approaches (image segmentation, feature and attribute extraction, integration of 2D and 3D, multi-temporal analysis, etc.) offers potential for significantly enhancing efficiency and provision of relevant information relating to physical processes. This project will address the above challenges, developing computer vision methods for intelligent, quantitative characterisation of channel hydromorphology with potential to support various catchment science challenges.
Aim and Objectives:
This project aims to fully exploit the potential of UAVs for hydromorphological characterisation at the reach scale, delivering advances in automated feature extraction and classification from high resolution UAV imagery. This will be achieved through the following objectives:
1. evaluate existing status of geomatics and computer vision approaches for hydromorphological and related environmental applications, reviewing methods for feature extraction and image classification;
2. optimise UAV data collection, and develop intelligent algorithms for automated extraction and classification of key hydromorphological variables at a selected reach-scale test site;
3. demonstrate and refine the developed methodology, assessing precision through multi-temporal assessment of hydromorphology for river restoration, validated through conventional approaches;
4. investigate broader transferability through implementation at a second test site under differing hydromorphological conditions, deriving and communicating stakeholder-relevant outcomes.
The research will be undertaken at the James Hutton Institute (JHI, Aberdeen) and Newcastle University (Newcastle upon Tyne) with time split between both organisations. The PhD will be supervised by Dr Pauline Miller ([Email Address Removed]) and Dr Mark Wilkinson at JHI, and Dr Claire Walsh ([Email Address Removed]) and Prof Jon Mills at Newcastle University.
Funding Notes
The studentship is funded under the James Hutton Institute/University Joint PhD programme, in this case with Newcastle University. Candidates are strongly urged to apply as soon as possible so as to stand the best chance of success. A more detailed plan of the studentship is available to suitable candidates upon application. Funding is available for UK and European applicants, but international applicants who possess suitable self-funding are also invited to apply.
References
1) Gneeniss, A.S, Mills, J.P., and Miller, P.E., 2015. In-flight photogrammetric camera calibration/validation via complementary lidar. ISPRS Journal of Photogrammetry and Remote Sensing, 100(2015): 3-13.
2) Orr H.G., Large ARG, Newson MD, Walsh CL. 2008. A predictive typology for characterising hydromorphology. Geomorphology 100, 32–40.
3) Walsh CL, Kilsby CG. 2007. Implications of climate change on flow regime affecting Atlantic salmon. Hydrology and Earth System Sciences 11(3), 1127-1143.
4) Tamminga, A.D., Hugenholtz, C., Eaton, B.C., Lapointe, M.F., 2015. Hyperspatial remote sensing of channel reach morphology and hydraulic fish habitat using an unmanned aerial vehicle (UAV): a first assessment in the context of river research and management. River Research and Applications, 31(2015): 379-391;
5) Report from the Commission on the implementation of the Water Framework Directive, 2012. European Commission Report, COM(2012) 670 final, 14 pp;