This first application in Ohio of remote sensing and GIS technology to inventory a forest species for forest management and utilization purposes was successful, providing useful information and forming the foundation for further refinements of the technology. The spectral signature for conifers produced a 100% accuracy rule in identifying the conifer stands that were visited.
This initial classification of land area into conifer or non-conifer saved much time compared to working with aerial photographs and performing the field work that would have been required with more traditional inventory methods. No attempt was made in this initial study to use spectral signature to distinguish among the various conifer species found in Ohio.
Fifty-six percent of the conifer stands identified were white pine. The other 44 percent were a variety of conifers found in natural stands and plantations throughout the study area. These species include Norway spruce (Picea abies L.), red pine (Pinus resinosa Ait.), Virginia pine (Pinus virginiana), pitch pine (Pinus rigida Mill.), scotch pine (Pinus sylvestris L.), and shortleaf pine (Pinus echinata Mill.).
A more efficient and accurate inventory of the white-pine resource in Ohio could be obtained with a spectral classification scheme refined to identify only white pine. Karteris (1990) used a six-band combination of Landsat TM data to achieve classification accuracies for individual stands of conifers ranging from 79.7 percent for Scotch pine (Pinus sylvestris L.) to 88.7 percent for Jack pine (Pinus banksiana Lamb.).
Information and data from our study can provide an important foundation for future efforts with white pine. This study estimated the white-pine resource for the 21-county study area in eastern Ohio as 24,147 acres containing 570.5 million board feet of volume (International 1/4-Inch Log Rule), with standard errors of 3,150 acres (13.05%) and 81.3 million board feet (14.24%), respectively.
These estimates differed substantially from those released in the 1991 Ohio Forest Inventory and Analysis (Griffith etal., 1993) which reported more than 51 thousand acres of white pine containing nearly 335 million board feet for the same geographic area. Standard errors for acreage and volume estimates in the Forest Inventory and Analysis, most of which were calculated on geographic subunits of the study area, all exceeded 34%.
Differences in results and precision between this study and the 1991 Ohio Forest Inventory and Analysis may be partially explained by temporal changes
and differences in sampling methods. At the time of this study, the average stand age within the study area was 37 years; at the time of the Forest Inventory, average stand age was 28 years. White-pine volume per acre can be expected to increase dramatically during this nine-year age period (Leak et. al., 1970).
This study was specifically designed to sample a single species within a specific geographic region; the Ohio Forest Inventory and Analysis examined all species and forest types over the entire state.
Statistically, as one examines smaller and smaller parts of a larger, broader study, precision is generally reduced and standard errors increase. This is true with the Forest Inventory and Analysis data. While standard errors statewide were exemplary 1.2 percent for total forest acreage and 2.4 percent for total volume, for example standard errors for smaller geographic areas or individual species or groups of species were much greater, often ranging up to 100 percent.
Finally, the average stand volume of 23,625 board feet per acre provided a suitable estimate on which to base the estimate of total white-pine volume in the 21-county study area.
However, additional detailed information on age distribution by acreage would be invaluable in assessing the availability of the resource over time under alternative utilization and management scenarios. This would more definitively answer questions concerning the amount and types of industry and harvesting the resource could support long-term. Obtaining such data using the remote sensing techniques and GIS technology utilized in this study would require developing separate spectral signatures for individual white-pine age classes.
Back | Forward | Table of Contents
All educational programs conducted by Ohio State University Extension are available to clientele on a nondiscriminatory basis without regard to race, color, creed, religion, sexual orientation, national origin, gender, age, disability or Vietnam-era veteran status.
Keith L. Smith, Associate Vice President for Ag. Adm. and Director, OSU Extension.
TDD No. 800-589-8292 (Ohio only) or 614-292-1868