Unmanned aerial vehicles for environmental applications

Recent developments in unmanned aerial vehicles (UAVs), platforms, positional and attitudinal measurement sensors, imaging sensors, and processing approaches (https://en.wikipedia.org/wiki/Unmanned_aerial_vehicle) have opened up a vast new area of opportunities in remote sensing for observation, measurement, mapping, monitoring, and management in various environments, natural (e.g. forests), production (e.g. agriculture), and built (e.g. urban). UAVs are also commonly referred to as drones, unmanned aerial systems (UASs) or remotely piloted aerial systems (RPASs), and form part of an on-going development process in Earth observation (EO).

Initially, there was air photography, which many decades ago was a relatively expensive operation and which tended to be carried out by national mapping agencies, while more recently it is the domain of private companies using digital systems and often with, or being replaced by, airborne lidar. Then there was the advent of Earth-orbiting satellites about 30–40 years ago and the invention of the term ‘remote sensing’ with this now progressing to include smaller and higher-spatial-resolution cube-sats. From the point of view of developing and operating the technology this too was expensive, although acquiring the data and using it was generally not too expensive. But the cost of developing, building, launching, and operating the spacecraft was generally borne by governments or more recently by private companies. Consequently, the new technology brought great advantages over aerial photography. In the 1970s–1990s, the data were made available to scientists in some cases for a fee, while since 2000 we have seen coarse- and moderate-spatial-resolution multispectral images being provided as open-access data in rapidly growing global public archives. Thus it became possible to study enormous areas of the surface of the Earth, far beyond what was possible previously. Mapping and resource monitoring, etc., became possible at regional, national, continental, and global scales, however we still lacked the ability to image and match the spatial scale at which most field- or ground-based measurements are needed, namely <1 m pixels. In the last few years we have seen the development of drones for the mass market as well as highly specialized and task specific drones. What began as a radio-controlled model aircraft hobby several decades ago has now developed far beyond that. In the early days it involved fixed-wing model aircraft and small petrol engines. Now, with the development of light low-power electric motors and modern rechargeable lightweight batteries, the motive power has become much more flexible and controllable. Helicopter or rotor-winged drones have become increasingly popular due to their ability to take off and land vertically (VTOL) and to maintain position. The UAV flight control of the systems, incorporating global navigation satellite systems (GNSSs) such as GPS, have become more stable and reliable. Cameras and other instruments have become much smaller and lighter and able to be carried on small UAVs and operated remotely. By being under the control of a local operator, within line of sight, the constraints imposed by orbital considerations of satellites have been removed. Having acquired a UAV, an observer can go out and acquire data to a bespoke specification. Some flight planning software, along with geometric and radiometric correction software, is available through code-repositories and available at low costs allowing users to collect data and process it to supposedly ortho-rectified image mosaics. However, these methods remain in their infancy, and therefore the methods for collecting appropriate datasets and turning them into useable, accurate, and repeatable geo-spatial information is an on going and evolving area of research. The exceptionally low altitude of drone operations and improved optical imaging systems means that the spatial resolution achievable is significantly improved over that obtained with satellites or manned aircraft, and at a scale matched to traditional field survey and observations in many disciplines and professions.

The other side of the coin, of course, is that line-of-sight UAVs are only suitable for studies over rather limited areas, typically of the order of several hectares up to several km². Another aspect is that using a UAV to gather data for a project places extra responsibilities on the project team to have all of the correct equipment and other requirements: platform, sensors, measurement systems, controllers, software, qualifications, operating protocols, and experience. This also includes experience in airborne image data processing to deliver geospatial data and derived products. Ordering airborne or spaceborne EO data or downloading existing data from publicly available archives may be a lot more straightforward. If a project involves using a UAV to gather the data then the project acquires responsibilities for complying with all the legal requirements associated with operating a UAV and in many countries these are not trivial. One paper in this special issue attempts to address the legal questions (Cracknell 2017). It is probably fair to say that the development of UAVs has been so rapid that many countries have not been able to keep their legislation updated fast enough to deal with the changing situation. Another aspect a project manager or the pilot of a UAV needs to take account of is the responsibility to members of the public and the need to have third-party insurance in case of damage or injury caused while operating the UAV. This point is particularly important for academic operators of UAVs, since academics often seem to blur the boundaries between their hobbies and their work; this will not be able to continue under UAV legislation, which has now been developed in a number of countries.

We thought it would be valuable to collect a number of papers on the use of UAVs for environmental research. This special issue of the International Journal of Remote Sensing was stimulated by two recent key conferences, a two-day conference on ‘UAS for Remote Sensing Applications’ held at the University of Queensland, Brisbane (Australia), in February 2016 (http://conf2016.uas4rs.org.au/) and a two day conference on ‘Small Unmanned Aerial Systems (sUAS) for Environmental Research’, held at the University of Worcester (UK) in June 2016 (http://www.worcester.ac.uk/discover/uav-conference.html). These are now annual conferences (http://conf2017.uas4rs.org.au/and http://uas4enviro2017.utad.pt). However, we have gone outside these two conferences to solicit additional papers for this special issue.

The focus of this special issue is on UAVs for environmental research and monitoring. With expertise from a range of international experts we have selected novel papers that have come out of the two recent conferences and additional international research in this field. The papers in this special issue are concerned in some way or another with solving real scientific or environmental problems. This does not necessarily exclude papers where the main thrust is methodological or concerned with the techniques of data analysis and interpretation, so long as it is ultimately relevant to scientific or environmental study objectives. Very many of these papers in this special issue describe work which could not have been done without a UAV, while some are concerned to show that the work that was able to be done more easily, cost-effectively, or better in some other way than by more conventional methods. The reader should not be looking for brilliant new methods of analysis and interpretation of remotely sensed data in these papers. Most of what is being done is applying established digital photogrammetric methods or remote-sensing analysis and interpretation to data acquired in a new way and to problems that in many cases could not have been tackled previously. The applications of the new opportunities provided by UAVs that are described in this special issue are many and varied. We have attempted to classify them according to the environments, sensor type, and specific techniques they used (Table 1).

Table 1. Classification of papers by topics.

Agriculture, vegetation, land cover
Ahmed et al. (2017)	Bagheri (2017)	Cummings et al. (2017)	Hill et al. (2017)	Li et al. (2017b)
Liu, Zhu, and Wang (2017)	Massarelli et al. (2017)	Matese, Di Gennaro, and Berton (2017)	Mesas-Carrascosa et al. (2017)	Müllerová et al. (2017)
Samiappaan et al. (2017a)	De Souza et al. (2017)	Szantoi et al. 2017	Tamouridou et al. (2017)	Yu et al. (2017a)
Forestry
Huang et al. (2017)	Kang et al. (2017)	Khan, Aragão, and Iriarte (2017)	Marx, McFarlane, and Alzahrani (2017)	Pádua et al. (2017)
Panagiotidis et al. (2017)	Thiel and Schmullius (2017)	Torresan et al. (2017)	Watt et al. (2017)
Geology
Barlow, Gilham, and Ibarra Cofrã (2017)	Bernard et al. (2017)	Bi et al. (2017)
Image processing
Kakooei and Baleghi (2017)	Kalantar et al. (2017)	Kim et al. (2017)	Li et al. (2017a)	Lizarazo, Angulo, and Rodríguez (2017)
Longmore et al. (2017)	Arif et al. (2017)	Qayyum et al. (2017)	Yi (2017)
Instruments
Babapour, Mokhtarzade, and Valadan Zoej 2017	Crusiol et al. (2017)	Cunliffe et al. (2017)	Tetuko et al. (2017)	Tripolitsiotis et al. (2017)
Lakes and rivers
Aguirre-Gómez et al. (2017)	Detert, Johnson, and Weitbrecht (2017)	Lloyd, Lepper, and Pomeroy (2017)	Lyu et al. (2017)	Nobajas et al. (2017)
Samiappan et al. (2017b)	Shintani and Fonstad (2017)	Seier et al. (2017)
Lidar
Bakuła et al. (2017)	Goodbody et al. (2017)	Guo et al. (2017)	Koska et al. (2017)	Wei et al. (2017)
Miscellaneous
Brenner et al. (2017)	Castillo-Carrion and Guerro-Ginel (2017)	Cracknell (2017)	Pei et al. (2017)	Simic Milas et al. (2017)
Yu et al. (2017b)
Photogrammetry
Ajayi et al. (2017)	Bühler et al. (2017)	Coveney and Roberts (2017)	Duarte et al. (2017)

The EditorsFebruary 2017