 |
||
|
||
 |
||
|
• Sponsorship & Exhibitor Information
• Registration coming soon • Conference Dinner and Entertainment
• About Oregon State University
|
ForestSAT 2012 • Invited Presentation Sessions
Below are descriptions for several of the invited presentation sessions that will take place at ForestSAT 2012.
1. Assessing the uncertainty of remote sensing-based estimates Session Organizers: Ron McRoberts (US Forest Service), Pontus Olofsson (Boston University) For remote sensing applications that focus on constructing land cover or land change maps or estimating the extent of land cover or land change classes, uncertainty assessment often ends with construction of an error or confusion matrix. These matrices reveal information regarding the overall degree to which the map or predicted classes correspond to observed classes in an accuracy assessment sample, but they do not produce inferences in the form of confidence intervals for parameters of the population that the map depicts. The presentations in this session address issues related to estimating variances and constructing inferences for remote sensing-based estimates of land cover parameters for both categorical response variables such as the extent of land covers or land change (including its impact on estimated terrestrial carbon dynamics) and continuous response variables such as tree volume or biomass. 2. Region to global forest monitoring programs & strategies Session Organizer: Scott Goetz (Woods Hole Research Center)
Forest monitoring has taken on considerably greater importance in recent years as a means to potentially mitigate climate change by maintaining forest stocks that sequester carbon while also avoiding carbon emissions associated with deforestation and forest degradation. Any related carbon management actions will require repeatable and accurate methods of monitoring, reporting and verifying estimates of forest extent, change and carbon stock. Monitoring implies repeated measurements through time that, in turn, require estimates of uncertainty and error. Remote sensing will continue to play a critical role in monitoring forest over large areas, but coupling with forest inventory and related field measurements is essential to ensuring a uniform, stable and transparent monitoring framework. This session will focus on the application of remote sensing and forest inventory in support of activities that enhance our ability to map, monitor, model and manage forest carbon. A series of invited talks will provide an overview of various programs designed to address these objectives at the regional, national and global scales. Additional contributed talks are welcomed to focus on quantifying changes in forest area, structure, and biomass dynamics; establishing baseline measurements; integrating remote sensing with analyses of historical data and field measurements; coupling remote sensing measurements with ecosystem and integrated assessment models; development of frameworks that incorporate airborne and satellite observations to help define and support emerging policies and programs; and additional topics related to the session theme.
3. Insect and wildfire disturbance ecology
Session Organizer: Garrett Meigs (Oregon State University)
In western North America, two of the most prevalent forest disturbances are insect outbreaks and wildfires. Both disturbance types have influenced forest ecosystems for millennia, both have been widespread in recent years, and both are projected to increase with climate change. Because insects and fires are particularly important in the same dry conifer forests, there is acute concern that insect-caused tree mortality increases forest susceptibility to extreme, fuel-driven fire behavior and effects. Studies across a variety of systems and spatiotemporal scales have produced mixed results, however, suggesting that there is no simple link between insect and wildfire damage. This session focuses on the causal factors and ecosystem impacts of these two important agents, as well as their interactions in space and time. Invited speakers will present innovative approaches to quantify and map the effects—both independent and interactive—of insects and wildfires. The session will draw from remote sensing, modeling, and field-based approaches, linking theoretical and empirical studies in a spatial analysis framework.
4. Using multi-source remotely sensed data and nearest neighbor techniques to improve inference for natural resource applications
Session Organizers: Temesgen Hailemariam (Oregon State University) and Ron McRoberts (US Forest Service)
Sustainable forest management decisions are rarely based on single objectives, and hence, strategic and operational forest assessments require information to support timber production, wildlife habitat, fire hazard mitigation, biodiversity, and carbon balance goals. Thus, timely, accurate, and precise information about the entire forest resource is needed. However, lack of data for some areal units limits the sustainable management of forests. Multi-source remotely sensed data and nearest neighbor techniques offer unprecedented opportunities to augment data, quantify forest attributes, handle missing data, and tackle emerging challenges. The purpose of the session is to bring together researchers, scientists, and practitioners with common research interests and to exchange ideas related to some of the challenges and opportunities in using multi-source remotely sensed data and nearest neighbor techniques to improve inference for sustainable forest management.
5. Forest ecosystem process modeling
Session Organizers: Dan Hayes (Oak Ridge National Lab) and David Turner (Oregon State University)
The incorporation of accurate and detailed land surface characteristics and biophysical variables in ecosystem biogeochemical models – needed for reliable assessments of climate variability within coupled Earth System Model frameworks and diagnosis of contemporary patterns in carbon flux – remains an on-going challenge. Carbon cycle process models, when combined with distributed climate data and remote sensing based observations, provide a means of scaling carbon stocks and fluxes over large spatial and temporal domains. Remote sensing of forest cover, LAI, FPAR, phenology, stand height, biomass, disturbance and successional stage are used for model drivers, model calibration, and model validation in various applications. This session will highlight case studies representing the state-of-the-art in forest ecosystem process model applications driven and evaluated by remote sensing data and products at scales ranging from the watershed to the region.
6. Recent advances in forest disturbance detection
Session Organizer: Sean Healey, US Forest Service
While the roots of modern satellite-based forest change detection go back more than two decades, the past few years have seen exponential growth in the extent and detail of available change maps. A well-publicized reason for this expansion has been the opening of the Landsat archive, but there have been other catalysts: cross-pollination between MODIS and Landsat communities; increased use of ancillary information for calibration and validation; and dramatically improved computing resources. The disturbance-detection work presented in this session spans multiple scales, sensors, and disturbance types. Presenters will introduce both new techniques and new applications, and the session should provide a good update on where this rapidly evolving field is headed.
7. Remote sensing education
Session Organizer: Peder Nelson (Oregon State University)
Remotely sensed images and the scientific products derived from them are important for understanding the Earth. As scientists discover and model the complexity of forests, there is a desire by non-scientists to incorporate this information into their understanding of the natural world. Therefore, this session focuses on the application of remote sensing in formal and informal education. Invited speakers will provide example tools, techniques, and programs developed for translating scientific work into effective teaching elements for everyday learning and decision-making.
8. Remotely controlled aerial systems
Session Organizer: Michael Wing (Oregon State University)
Increasingly, remotely controlled aircraft are being used to capture data and imagery. Airspace, however, has been severely restricted within the U.S. for remotely controlled aircraft flights. Recent activity by the U.S. Federal Aviation Association (FAA) suggests that remotely controlled aircraft will be gradually allowed to fly over a broader portion of the U.S. This increased capability will drive efforts at platform and sensor development, with significant implications for remote sensing applications. This session will include presenters that address the current flight capabilities of remotely controlled aircraft in the U.S. and other countries. Presenters will also include material regarding current and anticipated payload, sensor, and platform technologies that could significantly transform the efficiency of remote sensing.
9. Precision forestry and its potential benefits for forest management
Session Organizer: Juan Suárez (UK Forest Research)
This session focuses on the capability of current sensors to map forest stand variability and suggests how this information can benefit forest management. Variability in structural characteristics is a feature of natural systems. Even in systems designed to be as uniform as possible, such as planted monoculture forests, structural differences are inevitable. Understanding the factors controlling variability is important in developing our knowledge of how ecosystems operate and behave. Forest observation with remote sensing can produce abundant and detailed information at increasingly finer scales that can inform the development of new methods in forest management. In a context where foresters have to balance timber supply with carbon monitoring, the provision of recreation facilities, sustainable ecosystems and land conservation, the requirement for abundant and timely information is paramount. Therefore, this session will consist of presentations that not only address methods for mapping complexity but also outline the benefits of integrating this new information into the planning and decision making processes.
10. Imaging spectrometry of forested environments
Session Organizer: Phil Townsend (University of Wisconsin)
Please check back for additional and updated session descriptions. |
|
|
|
||