Efficient agricultural practices is a global concern. Improved agricultural productivity and sustainable practices are needed if we want to meet the UN sustainable development goal to end hunger by 2030. Underpinning investment in infrastructure, technology and research into improved crop monitoring capabilities to drive the development of agricultural information and crop yield forecasting system is urgently needed.
To ensure good agricultural productivity, the health of crops as well as potential crop infestations needs to be assessed. Stress can be the result of water deficiencies or of infestation of insects, fungi etc. which need to be detected to allow early mitigation by farmers. Productivity will also vary due to local growth differences, for example from soil nutrient deficiencies.
Remote sensing is one of the key technologies that can help to allow farmers to identify areas which are experiencing difficulties and to decide on the best intervention to improve crop yield but also to reduce the impact on the environment. Remote sensing provides also important data needed for the development and testing crop yield models as well as for research on the functioning of crops and impact of climatic variations on plants.
Hyperspectral imaging from airborne (aircraft and drone) platforms is one of the key technologies for detecting, identifying and characterising surface features providing a wealth of information on crops and crops health. These instruments are highly flexible and cost efficient and the acquired information is on a sub-meter scale not achievable with satellites.
However, the performance of current hyperspectral imaging systems is limited and there is an urgent need to advance such systems to produce more accurate, repeatable surface reflectance and subsequently more reliable information on surfaces and crops. This could be achieved by measuring downwelling solar irradiance which can then be used in conjunction with the imaging system to infer reflectances. This would be a significant advancement over existing systems.
We will work closely with 2Excel Aviation Ltd., a leader in the field of airborne hyperspectral imaging and the NERC Field Spectroscopy Facility who are experts in radiometric and spectral calibration. We expect that this new, combined up/downwelling system will enable improved applications of hyperspectral imaging which we will demonstrate making use of the field site network and ongoing activities of 2Excel Aviation. Our focus here will be on crop (and tree) health information which will be extracted from the improved hyperspectral reflectances using state-of-the art machine learning approaches.
Partners and collaboration (including CASE)
This project will be a cooperation between University of Leicester and 2Excel Aviation Ltd. who have offered to provide CASE support for this studentship. The partnership between University of Leicester and 2Excel Aviation Ltd. has been facilitated by the Leicester Innovation Hub and the studentship project has been jointly developed by Leicester and 2Excel Aviation Ltd. 2Excel Aviation Ltd. will also play an important role in the delivery of the project providing access to facilities, instruments and expertise in hyperspectral imaging and the student is expected to work closely with the team at 2Excel Aviation Ltd.
This project is well aligned with the Industrial Strategy and is directly relevant for the grand challenge of ‘Clean growth’ which strives to put the UK at the forefront of advanced sustainable agriculture with innovative technologies and techniques. The project is also addressing the grand challenge of ‘Growing AI & Data-Driven Economy’ by contribution to observing systems that collect substantial remotely-sensed data volumes that are analysed using innovative machine learning methods.