Ambiental in partnership with universities in pioneering research with UAVs
When flooding occurs the priority response is saving lives and ensuring the safety of inhabitants – as was the case during the devastating floods in Cumbria last December. But as events unfold it is the role of insurance companies to assist their policy holders, and insurers face challenges in accurately assessing the extent of flooding. A clear and immediate understanding of the scale and likely impact upon insured assets is required in order to inform shareholders of expected losses, make logistical preparations and to prevent fraudulent claims. This is where the use of flood models and aerial field surveys can help.
Ambiental has been involved in pioneering research which uses unmanned aerial vehicles (UAVs) to image areas at risk of flooding. Aerial surveys can occur pre-emptively, as well as during and directly after a flood event. By using UAVs, as opposed to light aircraft, it is possible to capture high resolution images of the areas affected for flood extent delineation and damage assessments. Furthermore, elevation models can be generated through the use of stereo images. These images which are created by measuring the displacement of neighbouring images of the same building.
In December, 2015, Justin Butler of Ambiental attended a National Environmental Resources Council [NERC] Climate Change Sandbox meeting. These events facilitate collaborations between private companies and university researchers, and Justin assembled a consortium of industry specialists. From this fateful meeting a partnership was formed between Ambiental and several Universities. The project aimed to explore the feasibility of using UAVs to image flood zones, and then using this data for flood science research and to develop commercial usage cases. UAV data collection can also be used to validate the accuracy of Ambiental’s UKFloodMap4™ and FloodScore™ products.
This project was fortuitously initiated in the weeks before the December 2015 flooding, caused by the unprecedented rainfall from storm Desmond. In reaction to this a pilot study was initiated which focused on Cockermouth, a town particularly seriously affected by the flooding. Ambiental contributed a targeted flight-plan for the UAV, which was directed by UKFloodMap4™ and a host of other flood related datasets including flood defence location data.
Despite taking flight as the floodwaters had started to recede, the high resolution imagery captured by the team allowed accurate assessment of the areas impacted by the peak flooding as debris and saturated ground were easy to observe. This data was corroborated with media reports and other sources to build a detailed picture of how many properties had been affected by the flooding.
Whilst this technology is still in the development stages, it is hoped that in the future data collected in this way can help predict flood risk and pinpoint which properties are most likely to be affected should floods hit Cockermouth again. Such information will be hugely beneficial to the flood response community, and allow insurance companies to mobilise their staff as early as possible.
Following this initial survey the project team have also gained permission to fly a UAV over Carlisle to capture images for a baseline survey of the area. The outcomes of this study have also been discussed at a recent productive workshop held with the first responder community and representatives of ResilienceDirect, who were very positive about the potential benefits of UAV technology. As such there are now plans underway to develop this work further as part of other innovative projects which Ambiental are involved in.
The work of the project team has been showcased in the March/April edition of GeoConnexion UK. The article entitled ‘UAVs in flood disaster response: a case study from Cumbria’ was written by Ambiental’s GIS & Data Manager Paul Drury. The article can be read in full click here and on the GeoConnection website.
The following video was produced as an output of the project and shows the photogrammetrically derived terrain model which was generated from a composite image mosaic formed from the overhead aerial photography acquired. The computer animation takes you on a virtual flyover of the town which shows much of the physical evidence of the flooding.