Drone Watch

Near real-time (48h) accurate mapping of flood extent, property and infrastructure damage in urban areas is required for the estimation of insurance loss and insurance claim validation. With flood damage estimated to be above £3billion/event (based on 2007 flooding events) and an average house claim of £40k, flood mitigation and damage assessment are a priority in the Governmental agenda. Current methodologies rely on the use of satellite data (SAR or optical) or aerial imagery collected from aircraft. However, several limitations within the underlying technology curtail the current effectiveness of the strategy: (i) optical imagery cannot provide information under low-cloud cover presence and (ii) satellite (SAR) data, which can penetrate cloud cover, is challenging in urban areas due to its oblique viewing angle and the difficulty of separating the water signature from other urban features. Hence, at present, we do not have sufficient data to robustly calculate flood impact in urban areas.

Small-scale Vertical Take Off (VTO) Unmanned Aerial Vehicles (UAVs) are emerging as a key engineering tool for future environmental monitoring tasks. Within the context of flood extent mapping and damage estimation, UAV aerial imagery offers both timely (on-demand) and increasingly detailed (higher resolution) information than comparable satellite or aircraft imagery. Here, we contribute to address the gap in the provision of immediate post-event flood extent and visible damage information by coupling knowledge from four core research areas:- (i) emerging technologies (Dr. Rivas Casado), (ii) flood engineering (Dr. Butler and Dr. Yu), (iii) remote sensing (Dr. Johnson) and (iv) catastrophe modelling (Dr. Irvine).

The objectives are achieved via six consecutive steps that integrate the four core disciplines mentioned.
Step 1 (O1). UAV data will be collected to obtain a representation of the flooded area at fine resolution. This will require site selection, flight design and data acquisition. Independent frames will be mosaicked together to produce an accurate orthoimage of the area of interest (maximising resolution).

Step 2. Satellite imagery will be collected for the area of interest to generate an additional layer of information (maximising coverage). The FloodMap model will be used to map flood extent using the digital elevation model generated from the UAV imagery. As far as possible, in limited time, existing damage data matching the selected event will be sought from Oasis or Ambiental partners.

Step 3. The data sets from (1) and (2) will be integrated in a Geographical System Information (GIS) environment (maximising modelling efficiency).

Step 4. Key parameters informing on flood extent will be extracted from (3) using established approaches. Outputs will be validated via ground truth data (if flooding occurs). In a flood constraint scenario, a comparison between flood derived parameters from UAV and satellite imagery will be provided (maximising modelling robustness).

Step 5. Key to the overall success of the project is a strong focus on the correct understanding of the market requirements. For this purpose, a small-scale, exploratory market definition will be undertaken to gather the type of products different end user communities will require to visualise and successfully use the integrated product (maximising product uptake).

Step 6. Step 6 will deliver the proof of concept through the integration of results from (4) and (5). A service definition will be provided to conceptualise the product the end user requires. At this stage, the service definition and integrated outcome should align to guarantee further allocation of funding resources on a cost-efficient tool for rapid assessment of damage from floods and storms that increases the accuracy in estimates of insurance loss and has the potential to enable the validation of insurance claims.

The proposal (and follow ups) will aim to facilitate maximal impact over a number of key themes:- industrial, societal and academic.

Impact to industry and society: The project has been designed from the outset to consider the information needs of the user communities detailed in the beneficiaries section. In particular, Step 5 of the project involves a market analysis to further engage with user communities to understand what information they currently lack, what information would help them respond more efficiently and rapidly and via what mechanism they would want to receive this information. The project team will draw heavily on its existing expertise and contacts through the OASIS+ project (Innovation of the year Industry award 2014) and potentially through the project's commercial partner, Ambiental, to engage with these communities.

The service definition activity (Step 6) will by its very nature involve feeding back to the user communities to demonstrate the proof of concept and seek their views on the potential service offering. It will also provide a clear opportunity to take their views on board in identifying further research needs and shaping the next stage of service development.

The impact of the project has far wider reach than the user communities identified above and will impact significantly on the general public. Making sure that insurance claims are handled rapidly and accurately will make a significant real world impact on the lives of those whose homes and businesses have been flooded. It will also help utilities companies to continue providing services to the public with reduced outages. Hence, not only does the proposed service support significant economic benefits in terms of reduced costs or improved efficiencies for insurance companies, as well as for utilities companies and local authorities but it supports significant social and societal good to the general population living in flood risk areas.

Academic Impact: In the short term, cross-collaboration between team members will enable the development of new ideas for the submission of future RCUK, Innovate UK and Horizon 2020 proposals. It will also consolidate collaboration with industry (Ambiental) and open pathways to further enable the transfer of knowledge between industry and academia. The 3 month project will help establish a closer link with the Catapult Satellite Applications and help Ambiental explore commercialization routes through different industry partners. This proof of concept project is sufficiently novel for the results to be published in high impact journals. All publications will be available open access through golden route. In the long term and as a result of follow up proposals, outcomes will be presented at key national and international conferences. The multidisciplinary nature of the project will enable the reporting of both individual state-of-the art advances relevant to each discipline and key advances across-disciplines. This will contribute to consolidate the PI and Co-Is track records with NERC.

The project aligns to the priorities of the space innovation and growth strategy, especially on the "use of climate and environmental monitoring from Space facility (CEMS) to process and merge streams into climate quality and other required data sets" and contributes to "assess the validity and integrity of observations, data and products". In addition, it aligns with the research priorities and specific challenges highlighted by the Roadmap for Climate Services (i.e. growing the climate services market and demonstrating the added value). The cross-disciplinary approach to service provision links to cross-council initiatives such as the "Living with Environmental Change" theme ("Coastal and waterway engineering", "Water engineering", "sustainable land management") targeting flood understanding, monitoring and defence.