TY - GEN
T1 - Advances in multi-sensor systems for in situ remote sensing of Australian forest canopy processes
AU - Hughes, D. E.
AU - Culvenor, D.
AU - van Dijk, A. I.J.M.
AU - van Gorsel, E.
AU - Woodgate, W.
N1 - Publisher Copyright:
© 2017 Proceedings - 22nd International Congress on Modelling and Simulation, MODSIM 2017. All rights reserved.
PY - 2017
Y1 - 2017
N2 - Remote sensing has been the mainstay of much environmental research for many years. Data acquired from remote sensing systems deployed on spacecraft and aircraft have provided insight into environmental processes that are difficult or impossible to obtain otherwise. However, traditional remote sensing systems are expensive, highly specialised and may lack the ability to be moved or quickly deployed. Fortunately, the cost of sensors like those used in existing remote sensing systems has decreased over time, and their availability and capability has increased. It is now possible to build and deploy ground-based remote sensing systems with capabilities similar to those on traditional platforms. In this context we define ‘in situ remote sensing systems’ to be within, above, or otherwise adjacent to the vegetation under study but not in direct contact with it. Such systems typically have a measurement range of a few tens of metres, can be deployed quickly and easily moved to new areas of study when required. This paper describes the development of in situ remote sensing systems as has occurred in Australia. The paper also outlines some of the challenges that were met during development and use of several generations of in situ remote sensing systems and presents contemporary work being done for the next generation of systems that will further expand our scientific measurement capability. Initial development of in situ remote sensing systems started in about 2003 with the deployment of the first single pixel multi-angle spectrometer at the CSIRO Tumbarumba field site in south-east New South Wales. Though this was a basic system, it provided much useful data, both in a scientific sense as well as providing information that would be useful in engineering the next generation of sensor systems. Early work in comparing acquired in situ data to satellite sensed data, and radiative modelling was complicated; difficulties included the effects of a non-homogenous forest canopy and the angular distribution of leaves. In 2013 a much more sophisticated system capable of hyperspectral and thermal imaging was installed at the same site, and this system continues to provide detailed and calibrated spectral and thermal time series images of forest canopy dynamics. The ability to examine specific and more uniform regions of interest within the forest canopy is the real value of imaging systems, as opposed to single footprint area-averaged measurements. Because of the high spatial resolution of the acquired data, it is now possible to examine ‘within-tree’ variability as well as ‘between-tree’ variability. The system provides much data for scientific analysis, and points to some of the engineering and data handling issues that influence the way future systems are developed. Building upon these earlier systems, the Australian National University has embarked on a project to build a ‘state of the art’ sensor system that will extend the range of observed wavelengths and be deployable on various platforms for either short-term or extended observation campaigns. This new system uses a variety of sensors that cover the visible, short- and long-wave infrared wavelengths. In the first instance, scientific applications will focus on in situ remote sensing at the ANU Forest Research Facility at the National Arboretum in Canberra. Some important lessons have been learnt over the past 15 years. In situ remote sensing systems require careful design and engineering if they are to reach their full potential. Instrument calibration, handling the vast amount of acquired data and efficient data reduction are significant factors in the successful use of such systems. This contribution concludes with a speculative view of the future direction of science infrastructure in this field.
AB - Remote sensing has been the mainstay of much environmental research for many years. Data acquired from remote sensing systems deployed on spacecraft and aircraft have provided insight into environmental processes that are difficult or impossible to obtain otherwise. However, traditional remote sensing systems are expensive, highly specialised and may lack the ability to be moved or quickly deployed. Fortunately, the cost of sensors like those used in existing remote sensing systems has decreased over time, and their availability and capability has increased. It is now possible to build and deploy ground-based remote sensing systems with capabilities similar to those on traditional platforms. In this context we define ‘in situ remote sensing systems’ to be within, above, or otherwise adjacent to the vegetation under study but not in direct contact with it. Such systems typically have a measurement range of a few tens of metres, can be deployed quickly and easily moved to new areas of study when required. This paper describes the development of in situ remote sensing systems as has occurred in Australia. The paper also outlines some of the challenges that were met during development and use of several generations of in situ remote sensing systems and presents contemporary work being done for the next generation of systems that will further expand our scientific measurement capability. Initial development of in situ remote sensing systems started in about 2003 with the deployment of the first single pixel multi-angle spectrometer at the CSIRO Tumbarumba field site in south-east New South Wales. Though this was a basic system, it provided much useful data, both in a scientific sense as well as providing information that would be useful in engineering the next generation of sensor systems. Early work in comparing acquired in situ data to satellite sensed data, and radiative modelling was complicated; difficulties included the effects of a non-homogenous forest canopy and the angular distribution of leaves. In 2013 a much more sophisticated system capable of hyperspectral and thermal imaging was installed at the same site, and this system continues to provide detailed and calibrated spectral and thermal time series images of forest canopy dynamics. The ability to examine specific and more uniform regions of interest within the forest canopy is the real value of imaging systems, as opposed to single footprint area-averaged measurements. Because of the high spatial resolution of the acquired data, it is now possible to examine ‘within-tree’ variability as well as ‘between-tree’ variability. The system provides much data for scientific analysis, and points to some of the engineering and data handling issues that influence the way future systems are developed. Building upon these earlier systems, the Australian National University has embarked on a project to build a ‘state of the art’ sensor system that will extend the range of observed wavelengths and be deployable on various platforms for either short-term or extended observation campaigns. This new system uses a variety of sensors that cover the visible, short- and long-wave infrared wavelengths. In the first instance, scientific applications will focus on in situ remote sensing at the ANU Forest Research Facility at the National Arboretum in Canberra. Some important lessons have been learnt over the past 15 years. In situ remote sensing systems require careful design and engineering if they are to reach their full potential. Instrument calibration, handling the vast amount of acquired data and efficient data reduction are significant factors in the successful use of such systems. This contribution concludes with a speculative view of the future direction of science infrastructure in this field.
KW - Forest
KW - Hyperspectral
KW - Infrared
KW - LIDAR
KW - Remote sensing
KW - Thermal
UR - http://www.scopus.com/inward/record.url?scp=85080910596&partnerID=8YFLogxK
M3 - Conference contribution
T3 - Proceedings - 22nd International Congress on Modelling and Simulation, MODSIM 2017
SP - 1236
EP - 1242
BT - Proceedings - 22nd International Congress on Modelling and Simulation, MODSIM 2017
A2 - Syme, Geoff
A2 - MacDonald, Darla Hatton
A2 - Fulton, Beth
A2 - Piantadosi, Julia
PB - Modelling and Simulation Society of Australia and New Zealand Inc (MSSANZ)
T2 - 22nd International Congress on Modelling and Simulation: Managing Cumulative Risks through Model-Based Processes, MODSIM 2017 - Held jointly with the 25th National Conference of the Australian Society for Operations Research and the DST Group led Defence Operations Research Symposium, DORS 2017
Y2 - 3 December 2017 through 8 December 2017
ER -