A 3.5-year PhD research project in the Department of Geography and the Institute for Hazard Risk and Resilience at Durham University. Supervised by Dr. Sim M. Reaney, Prof. Richard Hardy and Dr Isabella Bovolo

Background Flood hazards in Nepal have resulted in significant regional impacts in recent years. Nepalese floods in August 2017 affected over 1.7 million people; 461,000 people were displaced, 159 people died and 65,000 houses were destroyed (UN RC 2017) making the event one of the worse in recent years. One widely employed method of risk transfer is the use of insurance to cover the losses incurred during flood events. The normal approach is to determine the probability of a flood event occurring at a location through hydraulic modelling and the associated damages from a depth-damage curves developed for that specific location.  However, in data poor regions (such as Nepal), there has not been the information required to develop and test these models. Recent innovations in remote sensing and simulation modelling have enabled existing quantitative approaches to be rolled out more widely. The results have the potential to be coupled with new approaches to flood insurance to enable the sale of risk transfer insurance products to new communities at affordable prices. To operationalize this aspect of the research, there are still several challenges associated with determining the flood hazard and the potential for errors which can significantly affect people’s lives.

Communities at risk of flooding near Pokhara, Nepal, in the valley bottom and potential flood water sources areas in the high mountains.

Aims and objectives The aim of this PhD research project is to determine flood hazard patterns in Nepal using remote sensing and simulation modelling. This research aim will be addressed with the following objectives: O1: To undertake spatial flood water source mapping and identification of vulnerable communities to determine case study catchments. O2: To determine temporal changes in flood hazard magnitude and frequency for a set of case study catchments. O3: To produce spatial patterns of flood risk determined by various terrain datasets and performance assessment approaches. The project will take a wide view across the different catchments in Nepal to determine the spatial pattern of flood water generation at the national scale. Findings from this research will feed into new technologies for financing disasters with colleagues in Durham University Business School.

Methods The broad scale mapping work will use the SCIMAP-Flood approach to determine the likely origin of the flood waters. This approach uses rainfall information, such as from TRMM coupled with on the ground rainfall records, topographic data to determine the hydrological connectivity and land cover information to determine the runoff coefficient from each location. This information set is integrated to give the catchments with the greatest amount of flood water generation across Nepal. This information will then be combined with information on communities at risk from flooding to determine a set of three case study catchments for more detailed analysis. To determine the flood hazard generation for each of the identified catchments, the project will apply hydrological simulation modelling techniques. The research will consider the trade-off between conceptual lumped catchment simulation models, such as HBV, and full spatially distributed process-based models, such as CRUM3. Each approach will be limited by the availability and quality of the rainfall and river discharge records and hence novel approaches to model assessment may need to be developed make best use of the datasets. Targeted field work to install monitoring equipment and undertake key hydrological measurements may be undertaken to support this work. The models will be run for both current climate conditions based on observed data and for projected climate change to determine the magnitude and frequency of flood events in the future.

Example aerial image and detailed topography that can be used for flood modelling

The final section of work will consider the inundation pattern within effected communities based on an understanding of the uncertainties in the estimation of the magnitude frequency relationship and the topographic dataset used. The inflow hydrographs will be based on the different catchment models and an ensemble of flood events will be used to drive a 2D hydraulic flood simulation model, such as HEC RAS 2D or Flood Modeller Pro.  The topographic data is key in determining the performance of the flood inundation model and the PhD will test the effectiveness of a range of data products. These will range from published datasets, such as the 30m ALOS data, the commercially available AW3D 5m data and detailed data based on photogrammetry since LiDAR is not available. These detailed datasets will be created from the Structure from Motion approach using source images from existing helicopter surveys and targeted drone flights. The helicopter sourced data can give ground resolutions of ca. 0.5m and the drone data can give resolutions of 0.1m. To determine the benefit of the more detailed, but more expensive data, the flood inundation model performance will be assessed against known events to determine the fit between the observed and the simulated flood extents. The database of known events will be determined using geomorphic evidence from past flood events using remote sensing products and on survey data in collaboration with other researchers at Durham. The before and after data sets will be based on open access Landsat and Sentinel data and the commercial Planet Labs daily imagery. The Landsat and Sentinel data may be able to give information on the maximum extent of the flood events but the daily Planet Labs data has the advantage of being able to potential capture the flood in progress, giving more information to assess the performance of the flood model.

Geomorphic changes near Lothar, Nepal, before (Jul 18 2017) and after (Aug 29 2017) the Summer 2017 flood event. Images from Planet Labs.

Research outcomes At the end of the research project, there will be the understanding and evidence to recommend a spatial flood hazard mapping method for Nepal. This approach will include recommendations as to how to make the best use of the available data and which additional datasets would have the greatest improvement on the hazard mapping. The climate change work will allow for an understanding of the potential changes in flood hazards over the 21st century. This information will be integrated into future work on providing flood insurance to the at-risk communities in Nepal and beyond.

The Studentship: This studentship will cover the UK / EU tuition fees at Durham University, a stipend of £14,777 per year and a Research Training and Support Grant (RTSG) to cover field work in Nepal, data costs and attendance at conferences including an international event, such as the American Geophysical Union meeting.

To apply:  Please send the following documentation by email to geog.pgadmissions@durham.ac.uk with the reference “NepalFloodsIHRR2”, by the deadline of 31st May, 2019 (5pm BST): 1) A current CV; 2) A cover letter (2 pages A4 max) which describes your motivation for applying for the project and your previous research experience; 3) Letters from two references (to be sent directly to geog.pgadmissions@durham.ac.uk by referees); 4) Transcripts of your previous qualifications. This opportunity covers the UK or EU fees. If you are interested in doing a PhD with IHRR and are from outside the EU, then please consider applying to the Christopher Moyes Memorial Foundation Studentship.

Your application will be reviewed and maybe shortlisted for interview. Interviews will take place in June 2019. The candidate will be expected to start in October 2019. If you have any questions about the research project, please get in touch with Dr. Sim Reaney (sim.reaney@durham.ac.uk). If you have questions about the application procedure, please contact geog.pgadmissions@durham.ac.uk.

Tagged with →  
Share →