Periodic Report Summary 1 - EXTREWAFOR (Acquiring extremely high resolution maps of water use efficiency of Australian forests to assess the effects of drought, species composition and stand structure.)
• A summary description of the project objectives
Instantaneous water use efficiency (WUE) of ecosystems, the ratio of gross primary production (GPP) and evapotranspiration (ET), is an important integral part of plant and ecosystem fitness and productivity. It is a key variable for modelling and understanding the impact of climate change, on terrestrial ecosystems. Unfortunately, the underlying mechanisms driving WUE variation at different scales are largely unknown, mostly because of the poor knowledge of differences in WUE between plants (trees) within the same ecosystem.
The objective of this project is to develop a method to measure within-ecosystem WUE using thermal, RGB and near infrared imagery obtained with unmanned aerial vehicle (UAV). Separate algorithms estimating evapotranspiration and GPP are developed and combined to yield WUE information. This method is used to study WUE of selected Australian forests.
• a description of the work performed since the beginning of the project
Flight campaigns
In the first two years of the research, measurements were performed at flux tower sites across three biomes three Australian forest biomes. The flights were performed with a UAV equipped with thermal, RBG and multispectral camera. The biomes were:
- Sclerophyll Eucalypt woodland at Cumberland Plains forest (NSW). This threatened ecosystem was followed up with fourteen UAV flight days over 16 months.
- Savannah forests along the Northern Australian Tropical Transect (NATT, NT). We performed a flight campaign in March 2013, where we applied the UAV in nine different locations, ranging from the very wet Top End part of the NATT transect (Howard Springs) to the driest part of the transect (Alice Springs supersite at Ti Tree (grassy mulga woodland and Corymbia/Triodia savannah). On this trip, measurements were performed at five locations with flux towers.
- Tropical rainforests (Queensland). In August 2015, we performed a flight campaign at the flux sites of the tropical rainforests of Australia, Cape Tribulation and Robson Creek. In this trip, we also applied hyperspectral sensors to measure sun-induced fluorescence and programmed our UAV so we can measure the BRDF (bidirectional reflectance distribution function) of natural ecosystems – we are the first researchers to do this.
Data processing and analysis
Processing and analysing the data is a very time-consuming part of the research. Specialized software is used to mosaic the images and to create RGB, thermal and multispectral orthophotos as well as 3D models of the canopy. These maps are co-registered and used for further calculations of ET, GPP and WUE. The ET, GPP and WUE algorithms are under development. Ground truth data of the tree species and their location will allow investigating these variables per species and height class.
Fig 1 RGB Orthophoto (left) and thermal (right) of the Alice Springs mulga flux site. Footage taken from flight at 28 March 2014.
• a description of the main results achieved so far
The data processing at the dry sclerophyll Eucalypt woodland site and the savannah biomes has been performed; the processing of the data of the tropical rainforest is nearly finished.
In the sclerophyll Eucalypt woodland, we studied the influence of mistletoe infection on the growth, performance and survival of Eucalyptus moluccana trees. Among the results, we found that mistletoe temperature is always lower than the temperature of infected trees, indicative of higher transpiration, and that this difference increases with average Eucalypt canopy temperature (Fig 2, left). As a consequence, infected Eucalypt trees have higher canopy temperatures than uninfected trees, indicative of reduced transpiration, particularly when canopy temperature (ie evaporative demand) is high (Fig 2, right).
Fig 2 the effect of mistletoe infection on canopy temperature of Eucalyptus mollucama trees. See text for details (in preparation for Global Change Biology). Dots indicate average temperatures per measurement day.
• the expected final results and their potential impact and use (including the socio-economic impact and the wider societal implications of the project so far).
Our project will develop a methodology to estimate ET, GPP and WUE at an unprecedented spatial (and temporal) scale using UAV remote sensing. This will allow us to investigate how the water use efficiency of single plants and entire ecosystems is affected by drought, and to what extent species composition and vegetation structure affect this relation.
In addition, the developed methodology will be generally applicable to all terrestrial ecosystems.
Contact details:
Postdoctoral researcher
Dr. Wouter Maes, wh.maes@ugent.be
Supervisors
Professor dr. Kathy Steppe, Kathy.steppe@ugent.be
Returning host, Laboratory of Plant Ecology, Ghent University
Professor dr. Alfredo Huete, alfredo.huete@uts.edu.au
Outgoing host, Remote Sensing, C3, University of Technology Sydney (UTS)
Instantaneous water use efficiency (WUE) of ecosystems, the ratio of gross primary production (GPP) and evapotranspiration (ET), is an important integral part of plant and ecosystem fitness and productivity. It is a key variable for modelling and understanding the impact of climate change, on terrestrial ecosystems. Unfortunately, the underlying mechanisms driving WUE variation at different scales are largely unknown, mostly because of the poor knowledge of differences in WUE between plants (trees) within the same ecosystem.
The objective of this project is to develop a method to measure within-ecosystem WUE using thermal, RGB and near infrared imagery obtained with unmanned aerial vehicle (UAV). Separate algorithms estimating evapotranspiration and GPP are developed and combined to yield WUE information. This method is used to study WUE of selected Australian forests.
• a description of the work performed since the beginning of the project
Flight campaigns
In the first two years of the research, measurements were performed at flux tower sites across three biomes three Australian forest biomes. The flights were performed with a UAV equipped with thermal, RBG and multispectral camera. The biomes were:
- Sclerophyll Eucalypt woodland at Cumberland Plains forest (NSW). This threatened ecosystem was followed up with fourteen UAV flight days over 16 months.
- Savannah forests along the Northern Australian Tropical Transect (NATT, NT). We performed a flight campaign in March 2013, where we applied the UAV in nine different locations, ranging from the very wet Top End part of the NATT transect (Howard Springs) to the driest part of the transect (Alice Springs supersite at Ti Tree (grassy mulga woodland and Corymbia/Triodia savannah). On this trip, measurements were performed at five locations with flux towers.
- Tropical rainforests (Queensland). In August 2015, we performed a flight campaign at the flux sites of the tropical rainforests of Australia, Cape Tribulation and Robson Creek. In this trip, we also applied hyperspectral sensors to measure sun-induced fluorescence and programmed our UAV so we can measure the BRDF (bidirectional reflectance distribution function) of natural ecosystems – we are the first researchers to do this.
Data processing and analysis
Processing and analysing the data is a very time-consuming part of the research. Specialized software is used to mosaic the images and to create RGB, thermal and multispectral orthophotos as well as 3D models of the canopy. These maps are co-registered and used for further calculations of ET, GPP and WUE. The ET, GPP and WUE algorithms are under development. Ground truth data of the tree species and their location will allow investigating these variables per species and height class.
Fig 1 RGB Orthophoto (left) and thermal (right) of the Alice Springs mulga flux site. Footage taken from flight at 28 March 2014.
• a description of the main results achieved so far
The data processing at the dry sclerophyll Eucalypt woodland site and the savannah biomes has been performed; the processing of the data of the tropical rainforest is nearly finished.
In the sclerophyll Eucalypt woodland, we studied the influence of mistletoe infection on the growth, performance and survival of Eucalyptus moluccana trees. Among the results, we found that mistletoe temperature is always lower than the temperature of infected trees, indicative of higher transpiration, and that this difference increases with average Eucalypt canopy temperature (Fig 2, left). As a consequence, infected Eucalypt trees have higher canopy temperatures than uninfected trees, indicative of reduced transpiration, particularly when canopy temperature (ie evaporative demand) is high (Fig 2, right).
Fig 2 the effect of mistletoe infection on canopy temperature of Eucalyptus mollucama trees. See text for details (in preparation for Global Change Biology). Dots indicate average temperatures per measurement day.
• the expected final results and their potential impact and use (including the socio-economic impact and the wider societal implications of the project so far).
Our project will develop a methodology to estimate ET, GPP and WUE at an unprecedented spatial (and temporal) scale using UAV remote sensing. This will allow us to investigate how the water use efficiency of single plants and entire ecosystems is affected by drought, and to what extent species composition and vegetation structure affect this relation.
In addition, the developed methodology will be generally applicable to all terrestrial ecosystems.
Contact details:
Postdoctoral researcher
Dr. Wouter Maes, wh.maes@ugent.be
Supervisors
Professor dr. Kathy Steppe, Kathy.steppe@ugent.be
Returning host, Laboratory of Plant Ecology, Ghent University
Professor dr. Alfredo Huete, alfredo.huete@uts.edu.au
Outgoing host, Remote Sensing, C3, University of Technology Sydney (UTS)