PhD Opportunity at Imperial College London: Automated monitoring of rainforest biophysical structure from aerial platforms
Supervisors: Dr Mirko Kovac (email@example.com), Dr Robert Ewers (Life Sciences)
Department: Department of Aeronautics
One of the most common field methods in all of forest ecology is the use of the tape measure to record the diameter of trees. Such a simple process underlies global initiatives in tropical ecology such as the Centre for Tropical Forest Science, RAINFOR, AFRITRON and GEM forest plots networks, the most widely sources of data on the carbon balance of tropical forests.
While simple, using a tape measure is time-consuming and therefore involves considerable labour costs. A typical 1 ha forest plot in the tropics contains >5,000 trees that have stems >10 cm diameter at breast height (dbh), the standard cut off size for measurement, and surveying that number of trees takes a team of five people an average of eight days. Based on the labour costs for staff working in Malaysia (£5,000 per person per year), this equates to a cost of ~£900 per hectare, and thus an annual recurring cost of £18,000 to maintain a relatively small plot network of just 20 ha such as at the SAFE Project (www.safeproject.net).
This project will develop a method for automating forest plots surveys, with the potential to drastically reduce the cost of doing the research. Our proposed system will involve the use of twin cameras attached to an unmanned aerial vehicle (UAV) that will fly unaided through the forest understory generating a 3-D map of forest structure. From this map we will be able to automate the recording of important biophysical attributes of the forest, including the size structure of trees (important for modelling forest dynamics), the woody volume of trees (and hence forest biomass and carbon stock), the proportion of canopy cover (which determines the amount of light reaching the forest floor), and the leaf area index (a key determinant of forest microclimate).
Although a variety of UAVs are available commercially, none are able to move robustly inside the forest due to the complexity in flight planning and crash susceptibility that is typically encountered when operating close to trees. In this PhD project we will develop novel UAV technologies and vision based measurement principles to enable automated tree structure monitoring. We have implemented first “aerial 3D scanners” in previous projects allowing for 3D mapping of indoor environments. In this project we will extend these platforms for flight in forests and tailor its control, navigation and platform design for ecological applications.
We will field test the UAV system at the SAFE Project experimental site in Borneo, where Dr. Ewers maintains a team of locally employed research assistants that is constantly monitoring a 20 ha tropical Forest plot network. This project will use the data generated from these manual surveys to calibrate and validate the accuracy of our new approach.
To apply for this PhD opportunity, please send a CV and cover letter to the project supervisor, Dr Mirko Kovac (firstname.lastname@example.org). The deadline for applications is 16 January 2017.
This is a paid position.