Mission | People | Organization | Research | Education | Project | Data | Tools | Links
Remote Sensing for Precision Forest Management
Project Leader: Dr. Randolph Wynne (L), VT
Background:
Rapid increases in human population and economic development stress the natural resource base on which we rely for food, fiber, and a wide array of public goods. As such, we are asked to produce more food and fiber on less land, demanding precision utilization of fertilizers, herbicides, pesticides, and improved genotypes on our farms and forests. However, maximizing farm and forest productivity while simultaneously minimizing inputs that are potentially expensive to both the landowner (direct costs) and general public (indirect costs arising from deleterious environmental effects) requires that the land manager have access to precise and up to date information on the condition of their fields and forests. Systematic remote sensing surveys can provide the ability for society to monitor and assess patterns of urban and suburban growth in relation to established patterns of natural resources and agricultural land. Such surveys can assist communities to identify, in advance, regions that might experience losses to productive capability, areas that might be subjected to environmental risks, or places where incompatible land uses might occur. Broad scale surveys provide information resources that allow people to anticipate and accommodate impacts of such processes.Appropriate data from air- and spaceborne sensors are increasingly available to forest land managers, but simple and reliable tools to extract the needed information (i.e., the amount and location of nutrient deficiency, drought stress, and unwanted competition) are lacking. CEARS researchers are already active in this area through a variety of competitive grants from federal agencies (NASA, NSF, USDA) and through cooperative efforts with Virginia Cooperative Extension and forest industry.
Most of the tools that we produce are applicable to both farm and forest management. However, we see the primary needs in precision forestry, as it is less funded and developed than precision agriculture. Our research will help us provide the remote sensing tools necessary to answer the following questions regarding the Commonwealth’s and Nation’s forest resources: (1) Where are the forests and how are they changing? (2) How much fertilizer and/or herbicide should plantation owners apply to their forests, and where? (3) Can we produce information for managers more quickly and reliably ? (4) Can we provide the information necessary for sustainable forest management while concurrently keeping Virginia’s and the Nation’s forest industry competitive in a global economy?
Technical Approach:
Our component objective is to begin to facilitate the adoption of advanced remote sensing and related geospatial information technologies to enable precision forest management by forest managers in the public and private sector. Our specific objectives, developed in discussions with partners in the public and private sector, include the development of the following tools and protocols:A prototype approach to delineation of sub stand treatment units that allows users to select from a set of options to accommodate uncertainty in the delineation process
Improvements in merchantable volume and biomass estimation by species using high-density multiple-return small-footprint Lidar data fused with co-registered hyperspectral data
Protocols to afford the use of commonly available remotely-sensed data to parameterize models that estimate carbon sequestration in urban forests and loblolly pine plantations
Improving the precision of forest area estimates derived using medium-resolution earth resource satellite data
Summary of Work PlanTwo initial study areas have been identified in consultation with partners in federal and state forest management agencies as well as forest industry. They comprise portions of the Appomattox-Buckingham State Forest in Virginia and the USDA Forest Service’s Southeast Tree Research and Education Site (SETRES). Lidar and hyperspectral data will be acquired in the summer of 2002 over Appomattox-Buckingham State Forest; data to support the creation of high-resolution (25 cm) digital color-infrared orthophotos will be collected over SETRES in the same general period. Extensive in situ data, including field spectra, field Lidar, and more traditional forest measurements will be collected in conjunction with both remote sensing data acquisitions. Two graduate students and one postdoctoral research associate will be assigned to the analysis of the resulting data sets that will allow us to meet our specified objectives according the quarterly milestones specified below.
Goddard’s Remote Sensing Information Partner (RSIP) program will provide MODIS data products that, when properly tailored, will support this activity’s purposes. James McManus, GMU, will provide that support.
Project Milestones1st Quarter
Compilation of GIS data for error propagation study area
Annotated bibliography on overlay of uncertain data
Completion of volume/biomass estimation algorithm using high-resolution data from airborne sensors
Initial results from sub-study to determining whether or not creating a bare-earth DEM is a necessary precursor to measuring important forest biophysical parameters using Lidar data2nd Quarter
Specification of options for handling uncertainty in sub stand delineation
Initial prototype of sub-stand delineation user interface
Establish C:N ratio of canopy and litter at SETRES (NC) study site
Initial results from exploration of utility of lidar intensity data for identifying land cover type of Lidar return
Data collected for study comparing data from ground- and airborne Lidar sensors to traditional forest measurements3rd Quarter
Development of optimal treatment unit algorithm(s)
Documentation of optimal treatment unit procedures
Parameterization of Biome-BGC for loblolly pine
Completion of new stratification method to improve precision of forest area estimates4th Quarter
Completion of first draft of prototype approach to delineating treatment units
Completion of urban forest carbon model and initial sequestration estimates
Initial empirical model relating lidar time-intensity domain to important forest biophysical parameters
Initial results of protocol using Lidar/hyperspectral data fusion to extract forest volume by species
Initial results from study comparing data from ground- and airborne- Lidar sensors to traditional forest measurements.
VAccess-MAGIC © 2001-2002. All rights reserved
