Novels

A national overview of airborne lidar application in Australian forest agencies

Description
A national overview of airborne lidar application in Australian forest agencies
Categories
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  SilviLaser 2011, Oct. 16-19, 2011 Ð Hobart, TAS, AU " A national overview of airborne lidar application in Australian forest agencies R.Turner  1 , N.Goodwin 2 , J. Friend 3 , D.Mannes 4 , J.Rombouts 5  and A.Haywood 6   1 Forest Science Centre, Department of Primary Industries New South Wales, Sydney,  NSW Russell.Turner@industry.nsw.gov.au  2 Remote Sensing Centre, Department of Environment and Resource Management, Ecosciences Precinct, Dutton Park. QLD  Nicholas.Goodwin@derm.qld.gov.au  3 Planning, Environment and Silviculture, Forest Products Commission, Bunbury, WA  jeremy.friend@fpc.wa.gov.au  4 Resource Information, Forestry Tasmania, Hobart, TAS david.mannes@forestrytas.com.au  5 Resource Planning, Forestry South Australia, Mount Gambier, SA Rombouts.Jan@forestrysa.com.au  6 Resource Planning, Department of Sustainability and Environment, VIC Andrew Haywood@dse.vic.gov.au  Abstract: This paper provides a narrative of airborne lidar application across Australian forest agencies. It includes a brief history of early lidar research and operational trials, as well as current programs and future directions on a state by state basis. This review demonstrates a diverse range of lidar applications and increasing adoption of lidar technology within state agencies across Australia. Keywords:  Airborne lidar, remote sensing, national review, forestry 1. Introduction It is now ten years since the first lidar trials were conducted in Australian forests and, assisted  by the growing accessibility of lidar datasets and the development of new processing procedures and software tools, there has been a dramatic escalation in lidar use in forest agencies. Today most forest agencies have experienced a paradigm shift from explorative research to large scale operational programs. Lidar technology is having a significant impact on Australian forest management, and continues to revolutionise wood inventory programs and harvest planning  processes. The lidar forestry community in Australia is relatively small but active; and with representatives in every state. Several national lidar forestry forums have been held across Australia to share ideas and experience. The first was held in Brisbane (Queensland) in 2002, the second was in Hobart (Tasmania) in 2007, and most recently another workshop was again held in Hobart in 2010. A consistent issue raised at these workshops is the importance of disseminating  SilviLaser 2011, Oct. 16-19, 2011 Ð Hobart, TAS, AU # information on significant lidar research and potential operational applications to the general forestry community within Australia. The aim of this report is to provide a useful summary of  past and present lidar work in each state forest agency. Of course, a comprehensive review of every research trial and operational program is well beyond the scope of this paper, however, a general overview will give a sense of the wealth of information and experience that has emerged over the past decade, and will serve to guide future research directions. 2. State overview This overview concentrates on forest agencies within the six states of Australia (i.e. Queensland (QLD), New South Wales (NSW), Victoria (VIC), Tasmania (TAS), South Australia (SA) and Western Australia (WA)). As far as the authors are aware, airborne lidar has not yet been utilised for forestry purposes in the two territories; the Australian Capital Territory (ACT) and the  Northern Territory (NT). This paper focuses on public commercial forests and plantations, but it should also be noted that lidar use in private plantations and public national parks and reserves has also increased significantly. A basic overview of lidar application is presented on a state by state basis. 2.1 Queensland The remote sensing centre within the Queensland Department of Environment and Resource Management (DERM) has used lidar over the last decade for quantifying a range of biophysical attributes and to support the implementation of DERM's vegetation management policy and  programs. The Injune Landscape Collaboration Project (ILCP) (Lucas et al. 2010a) has been an important lidar research site for Queensland since 2001. Injune is located in the Brigalow Belt of central Queensland and the vegetation consists mainly of open poplar box (  Eucalyptus  populnea ) woodland with patches of denser white cypress pine ( Callitris glaucophylla ) regeneration. Lidar was used to evaluate its utilisation in estimating biomass and a set of forest structural attributes (Tickle et al.  2001, and Lucas et al  . 2006) and results showed that stand-based lidar derived biomass models were highly correlated with field data (R  2  = 0.92, SE = 12 Mg/ha). This site was reflown with lidar in 2009 and is now the focus of further research into detecting forest structural change over time. DERM has been involved in several research studies to better understand the relationship  between field data, terrestrial laser scanning, sensor configuration, multi-temporal lidar, and forest structure. For example, a series of field monitoring plots throughout Queensland have had several repeat airborne lidar acquisitions between 2000 and 2009 (Lucas et al  ., 2010b) to better understand the impacts of drought and land management practices on forest structure and species composition. In 2005 DERM, in collaboration with the University of Newcastle, investigated the use of lidar intensity and crown transparency to distinguish between forest species in white cypress woodland (Moffiet et al  . 2005). This study showed that vegetation types could be discriminated using the proportion of singular returns and a "porosity" index  based on the proportion of lidar points penetrating the canopy. Forestry Plantations Queensland (FPQ), now owned by Hancock Queensland Plantations Pty Ltd, has incorporated lidar into their operational program. Over a three year period (2005 Ð 2007) FPQ captured lidar data across their entire softwood plantation estate (188,000 ha) (http://www.fpq.net.au/). This information was used mostly for digital terrain model (DTM) derived products (e.g. slope, hillshade, and contours) to assist with classifying harvesting terrain classes in pre-harvest planning (e.g. above and below 24 o  slope). Lidar was also used to characterise forest structure including stratification of stands based on height (equivalent to site index) to assist pre-harvest inventory.  SilviLaser 2011, Oct. 16-19, 2011 Ð Hobart, TAS, AU $  The largest airborne lidar acquisition to date has been the Protecting Our Coastal Communities (POCC) project which covers an area of around 60,000 km 2  along the Queensland coast (see Figure 1). Commencing in 2009, this project, jointly funded by the state and local government, involved multiple providers/sensors for different regions. Data was captured with an average sampling density of 2 returns per sq.m. Although primarily intended for planning flood risk and urban development along the coastline, it also provided an excellent baseline for monitoring coastal forests and an opportunity for research such as the calibration/validation of spaceborne sensors, input for forest monitoring applications, as well as the extraction of many spatial  products. Figure 1: Current lidar coverage (in black) within Queensland. Note: this is only the lidar data held by DERM with the total area exceeding 68000 km 2 . Lidar data has since been utilised in over 50 sites strategically located to capture the variability in forest structure and composition. In addition to developing new relationships between field data and lidar, this data has been used operationally to calibrate and validate statewide methods and products (Armston et al  . 2009). For example, lidar derived layers have been used to validate foliage projective cover (FPC) and woody extent which were derived from Landsat (25m) and MODIS (250m) products and to explore the relationships between canopy variables (Armston et al  ., 2008; Gill et al  . 2009; Scarth et al  ., 2008 and Witte et al.  2000). More recently, the southeast Queensland region of the coastal capture was used to produce a calibrated FPC layer using 20 field sites located across the 8500 km 2  area and results showed that field measurements of FPC were highly correlated with first return lidar data (R  2 =0.92, RMSE=5%). In addition to  providing a baseline of FPC for southeast Queensland, it may help to improve the calibration of FPC using satellite imagery in forested areas with steep topography. 2.1 New South Wales In NSW, around 2.2 million hectares of state-owned commercial forests are managed by Forests  New South Wales (FNSW). With 230,000 ha of softwood plantation, FNSW is also the largest softwood plantation owner in Australia. The first lidar trial began in 2001 when 1,000 ha of  SilviLaser 2011, Oct. 16-19, 2011 Ð Hobart, TAS, AU % eucalypt forest were flown on the Central Coast. The focus of this trial was above-ground  biomass assessment and an automated canopy segmentation process was developed (Turner 2006). In the same year, the first large scale trial occurred as a spin-off from a much larger (1.8 million hectare) catchment study by the Murray Darling Basin Commission (Liu et al. 2003). This project provided data across 90,000 ha of river red gum forests (  Eucalyptus camaldulensis ) along the southern border. The data was used to remap road and drainage networks, identify and estimate thinning resources, and plan harvesting events. In addition, a small (450ha) wood inventory trial showed it was possible to predict maximum height (R  2  = 0.9), mean dominant height (R  2  = 0.76), basal area (R  2  = 0.72) and gross volume (R  2  = 0.79) (Turner & Webster 2005, and Turner 2007). In 2004 a study funded by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) acquired lidar data across 1,813 ha of native forest near Coffs Harbour on the North Coast. The project investigated the influence of scanning at different altitudes (1000, 2000, and 3000 m), footprint sizes (0.2, 0.4, and 0.6 m), scan angles (10¡ and 15¡ angle off nadir) and  point sampling densities (0.18 to 1.9 m) on forest structure assessment (Goodwin et al.  2006). By 2006, another regrowth forest site, covering an area of 12,800 ha, had been flown on the Central Coast. The Jilliby Catchment Area (JCA) was a multiagency collaborative research  project and data was used for two studies running simultaneously. The first focused on the application of airborne lidar for spatially mapping forest fuel characteristics (Roff et al   2006 and Turner 2007), while the other study explored the potential for forest health monitoring by mapping Bell Miner Associated Dieback (BMAD), (Haywood & Stone, 2011). Another 6,000 ha trial in coastal eucalypt forests was completed in 2008. A number of compartments were selected to investigate the benefits of lidar data in harvest planning and field supervision (Turner 2007 & 2008). It was estimated that lidar data reduced the planning effort  by 2 to 3 person days per compartment plan and provided a 10% time saving in field supervision during harvesting, with less walking required to locate exclusion zones and merchantable trees. Automated drainage maps from lidar-derived DTMs were also tested. Precision surveys along two creek lines indicated an excellent correlation between survey and DTM elevation (R  2  = 0.99) with a mean elevation error of 0.6 m, while automated drainage networks had a centreline mean error of 1.65 m. A second large scale study was implemented in early 2008 when 240,000 ha of native forest were flown in North Central NSW near the town of Baradine. The Pilliga Remote Sensing (PILRES) project covered cypress pine/eucalypt woodland on predominantly flat terrain. A wood resource inventory utilised airborne lidar (to provide height and stocking) and multispectral digital photography (to define stands of commercial forest types). Lidar-derived canopy height models (CHMs) were used for a strategic thinning program and the high resolution DTMs assisted the update of road and drainage networks. The project was expanded  by another 129,000 ha during 2009 across 93 state forests scattered throughout western NSW. This was also the first time that simultaneous lidar and digital photography (colour infra-red) was acquired, making it easier to combine attributes at crown level from both sensors (i.e. data fusion). The first case study in a NSW softwood plantation was initiated in July 2008. The Plantation Airborne Resource Inventory Appraisal (PARIA) project was undertaken in a 5,000 ha  Pinus radiata  plantation in south-central NSW. The study evaluated airborne lidar and digital multispectral aerial photography for wood resource inventory, structure stratification and forest health monitoring (Stone et al.  2008, Stone et al.  2010 and Turner et al   2011). A remote sensing guide for softwood plantation managers (Turner and Stone 2010) was also produced as well as a new in-house ArcGIS lidar toolbox.  SilviLaser 2011, Oct. 16-19, 2011 Ð Hobart, TAS, AU & In late 2009, FNSW and the University of New South Wales (UNSW) obtained an Australian Research Council (ARC) grant to investigate full-waveform lidar applications and to develop a commercial software package for on-screen interpretation. In 2010, full-waveform lidar was acquired at three sampling densities (i.e. 2, 5 and 10 pulses per sq.m) over a 140 ha pine  plantation west of Sydney. The data is being used to develop a new waveform processing technique to minimise lidar negative tree height bias (Park et al. 2011) and to model 3D structure with both airborne and terrestrial laser scanner (TLS) data (Park et al.  2010). Development of the lidar interpretation software is well underway and a working prototype should be ready for field testing by December 2011. FNSWÕs largest operational program is currently in progress throughout the northern tablelands and mid-north coast of NSW. The project is capturing almost 296,000 ha of state forest, including around 17,500 ha of pine plantation. The data will be used for general forest management, forest inventory, stand structure assessment and drainage mapping. Increasingly, lidar data is becoming more accessible from other sources. Local shire councils are acquiring lidar for flood mitigation and urban planning, and data covering around 198,000 ha of state forest has been made available. In 2008, the NSW State Government also purchased a Leica ALS50-II laser scanner that is operated by the Land and Property Management Authority (LPMA) based in Bathurst. LPMA has an ongoing lidar acquisition program for their state mapping needs and where data is captured over state forests it is also provided to FNSW. Since 2001, the accumulative total of all airborne lidar coverage in NSW state forests has reached 943,000 ha, which represents 43% of the total native forest estate. In addition, around 22,000 hectares of softwood plantation have been captured covering almost 10% of the total plantation area. Figure 2 illustrates the widespread coverage of this data across NSW. Figure 2. NSW map showing the location of lidar coverage in state forests (in black). 2.3 Victoria The Victorian Department of Sustainability and Environment (DSE) is responsible for the sustainable management of 7.8 million hectares of public native forests (DSE 2008). DSE has  been investigating the use of lidar as a land management tool since 2001 (Choma et al  ., 2005).
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks