Entity

Time filter

Source Type

Melbourne, Australia

Mellor A.,RMIT University | Boukir S.,University of Bordeaux 1 | Haywood A.,Land | Jones S.,RMIT University
ISPRS Journal of Photogrammetry and Remote Sensing | Year: 2015

Studies have demonstrated the robust performance of the ensemble machine learning classifier, random forests, for remote sensing land cover classification, particularly across complex landscapes. This study introduces new ensemble margin criteria to evaluate the performance of Random Forests (RF) in the context of large area land cover classification and examines the effect of different training data characteristics (imbalance and mislabelling) on classification accuracy and uncertainty. The study presents a new margin weighted confusion matrix, which used in combination with the traditional confusion matrix, provides confidence estimates associated with correctly and misclassified instances in the RF classification model. Landsat TM satellite imagery, topographic and climate ancillary data are used to build binary (forest/non-forest) and multiclass (forest canopy cover classes) classification models, trained using sample aerial photograph maps, across Victoria, Australia. Experiments were undertaken to reveal insights into the behaviour of RF over large and complex data, in which training data are not evenly distributed among classes (imbalance) and contain systematically mislabelled instances. Results of experiments reveal that while the error rate of the RF classifier is relatively insensitive to mislabelled training data (in the multiclass experiment, overall 78.3% Kappa with no mislabelled instances to 70.1% with 25% mislabelling in each class), the level of associated confidence falls at a faster rate than overall accuracy with increasing amounts of mislabelled training data. In general, balanced training data resulted in the lowest overall error rates for classification experiments (82.3% and 78.3% for the binary and multiclass experiments respectively). However, results of the study demonstrate that imbalance can be introduced to improve error rates of more difficult classes, without adversely affecting overall classification accuracy. © 2015 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Source


Amundson R.,University of California at Berkeley | Berhe A.A.,University of California at Merced | Hopmans J.W.,Land | Olson C.,U.S. Department of Agriculture | And 2 more authors.
Science | Year: 2015

Human security has and will continue to rely on Earth's diverse soil resources. Yet we have now exploited the planet's most productive soils. Soil erosion greatly exceeds rates of production in many agricultural regions. Nitrogen produced by fossil fuel and geological reservoirs of other fertilizers are headed toward possible scarcity, increased cost, and/or geopolitical conflict. Climate change is accelerating the microbial release of greenhouse gases from soil organic matter and will likely play a large role in our near-term climate future. In this Review, we highlight challenges facing Earth's soil resources in the coming century. The direct and indirect response of soils to past and future human activities will play a major role in human prosperity and survival. © 2015, American Association for the Advancement of Science. All rights reserved. Source


Burmeister N.,Kellogg Brown and Root | James D.,Land | Victory S.,KBR
Australian Coasts and Ports 2015 Conference | Year: 2015

On 24 June 2014, a storm event caused wave damage to numerous coastal structures in Port Phillip, Victoria, Australia. The storm event was modelled to quantify the prevailing extreme (storm tide) water level and wave conditions. An Average Recurrence Interval (ARI) of between 50 and 100 years was estimated for the event. Post-storm damage condition assessments of pier and seawall structures in Port Phillip were conducted by members of the study team and six structures were selected for discussion. The six structures were estimated to cost approximately $1m to repair and resulted in facility closures, some of which lasted several months. Uplift pressures of up to 10 kPa were estimated to have impacted low lying pier decks. A review of the failure mechanisms revealed a common issue that is relatively easy to rectify. Overtopping at Black Rock Seawall during the event was estimated to exceed recommended limits. It was found that Black Rock Seawall, perhaps like many historic seawalls in Port Phillip, underperforms by today's standards. This investigation sheds light on the management of coastal and maritime structures in Port Phillip, some of which were first constructed over a century ago. The investigation findings can be used for risk profiling of structures in Port Phillip and to guide future design, investment and decision-making. Source


Appleby G.,95 Broadway | O'Brien M.,Land
Australian Field Ornithology | Year: 2015

The Victorian distribution of the Brown Gerygone Gerygone mouki was previously thought to include only East Gippsland. Since the late 1970s, this species has been recorded at an increasing number of sites to the west, between East Gippsland and Melbourne. Throughout this period and region, nesting has also been noted but with breeding success uncertain. Brown Gerygones have been recorded in increasing numbers and in most months of the year, suggesting that they are probably now resident well to the west of their historical range in Victoria (i.e. East Gippsland). The habitats used by the Gerygones in the newly occupied range are similar to those used in East Gippsland but may differ in floristic composition, structure and naturalness. Reasons for and mechanisms of this species' range extension are unclear and would be worthy of further research. © 2015 Australian Field Ornithology. Source


Lacey G.,University of Melbourne | O'Brien M.,Land
Australian Field Ornithology | Year: 2015

Breeding of Fairy Terns Sternula nereis at the two current breeding sites in Western Port, both on French Island, Victoria, has been documented over more than four decades. At the main site, Rams Island, there has been great variability annually in the numbers of Terns breeding but no significant trend. This breeding colony has persisted for over 40 years, which has not happened at any other Fairy Tern colony in Victoria. There are only three breeding records for Tortoise Head. Current management of the breeding sites appears to be successful and this pattern should be maintained. Priorities include continued monitoring of breeding, and control of feral cats and weeds. © 2015 Australian Field Ornithology. Source

Discover hidden collaborations