|Section II --
Map Development (continued)
|Local Table of Contents:|
|From: Influences on
Wetlands and Lakes in the Adirondack Park of New York State:
A Catalog of Existing and New GIS Data Layers for the
400,000 Hectare Oswegatchie/Black River Watershed, 1997
|II.K. 1916 Fire Protection Map||OB2 Contents|
One of the objectives of this study was to determine the recent history of land use for the study area and major disturbances to the landscape. Such known and documented disturbances include a series of major forest fires that occurred at the turn of the last century. Other disturbances such as major storm blowdowns are discussed in other sections of this report.
According to McMartin (1994), in the Adirondacks "before 1890, fires had been small and scattered...in the period between 1890 and 1910, however, fires, mostly caused by railroad locomotives, plagued landowners, seriously affecting timber tracts for the first time". Estimates of the acreage burned parkwide during this period are over one million (McMartin 1994).
Description of existing data
Prior to this project, a digital map showing recent major fire disturbances in the Adirondack Park did not exist. The best available paper map is one prepared in 1916 by K.Schmitt of the State of New York Conservation Commission in Albany, NY. It is called Fire Protection Map of the Adirondack Forest and is drawn at a scale of 1 inch = 2 miles. There is no known documentation of how the information on this map was collected.
The polygon data were digitized directly from the map and are shown (and labelled as):
Historical landscape disturbances, such as fires, influence nutrient cycling and determine plant composition. In the boreal forest ecosystem, which is an important constituent of the Oswegatchie and Black River basins, fire is believed to be the primary altering force in forest successional processes and many of these species which currently exist on these sites are fire tolerant (Barbour and Billings 1988). A forest fire can significantly affect nutrient cycling in several ways. First, the organic matter at the surface of the forest floor is oxidized eliminating ion exchange sites that otherwise held nutrients in the most reactive part of the soil. Also, the exposed soil surface is reduced to ash offering a high erosion potential from uplands to lowlands. Wetlands provide the first sink of nutrients in this net loss of nutrients from the upland. Much of the Adirondacks was burned by devastating fires in 1903 and 1908 (Barrett et al. 1961) and the effects of those fires, as described above, are still apparent today (Kudish 1992).
|II.L. November 1950 Blowdown Map||OB2 Contents|
One of the objectives of this study was to determine the recent history of land use and major disturbances for the study area. Such documented disturbances included a major windstorm that occurred on November 25, 1950. Other disturbances are discussed in other sections of this report.
McMartin (1994) writes about this storm:
" known as the big blowdown, it struck in November of 1950, and did the most damage in the western and central Adirondacks. This powerful northeaster affected 420,000 acres and was said to have caused a loss that ranged from a quarter of the trees to the entire forest cover. The loss was estimated at two million cords of softwood and forty million board feet of hardwood. The Cold River country between Seward and Santanoni experienced the greatest destruction, followed closely by the Moose River Plains, but private tracts such as 60,000-acre Whitney Park, parts of the Adirondack League Club property, or Finch Pruyn's 183,000-acre holdings were also severely damaged.
The storm was particularly devastating to old-growth forests. The wind came from the east and northeast, while the mature trees had grown wind-firm in the direction of more normal west winds. Because spruce had shallow roots making it prone to wind damage and because it towered above the canopy of hardwoods, this most prized species was stripped from most virgin stands as well as some from State land that had been logged a half century or more earlier."
Description of the existing data
We found no existing digital map of the November 1950 storm. The best available map was a paper blueprint prepared by the New York State Conservation Department called the Map of Storm Damage in the Adirondack Area November 1950 drawn at a scale of 1 inch = 4 miles. The legend on this map reads:
- 25-50% Blowdown
- 50-100% Blowdown
- County lines
- Adirondack blueline
This map provides the following information: "From aerial reconnoissance data furnished by District Offices of the Conservation Department. Compiled and drawn by Gerald J. Rides, Forest General Foreman, Conservation Department. December 1950. Map No.36".
The polygon data were transferred on to a cartographically correct base map (1:250,000 NYS DOT) and then digitized and labelled as:
- 50-100% Blowdown
- 25-50% Blowdown
- Oswegatchie/Black Boundary
|II.M. State Forest Acquisition Map||OB2 Contents|
One of the objectives of this study was to determine recent land use history for the study area and documented natural disturbances such as major windstorms and forest fires. In addition, we expected to discuss the potential locations of virgin stands (never harvested).
According to Barbara McMartin (1994), "In order to understand what happened historically to different parts of the Adirondack forest, a good map is essential for locating forest tracts, comprehending their history, and determining their owners". This book documents the history of Adirondack logging and includes a discussion of old growth forest occurrences.
Description of existing data
Prior to this project, only a paper map belonging to B.McMartin based on her Adirondack logging history research was available. This map shows parcels of land acquired over different periods by the State of New York to the Adirondack Park. McMartin used a number of sources including a series of books on acquisition published by DEC (1920). This acquisition information was color coded and transferred to a base map (Great Lot Map at a scale of 1:250,000+). Much field checking was done by McMartin and others over the years, particularly in areas where virgin stands were expected to occur.
For this project the existing McMartin map was recoded and redrawn onto a cartographically correct 1:250,000 scale map (NYS DOT) using a light table. This re-coding included the following categories which were used as a legend:
- Virgin stands; never harvested.
- Parcels acquired in 1871 and 1877 tax sales, or owned by the State prior to 1871.
- Parcels acquired in 1881 and 1885 tax sales.
- Parcels acquired in 1890 or later via tax sale or purchase.
- Parcels not in Forest Preserve.
This final paper map product was then digitized and labelled using GIS.
|II.N. July 1995 Blowdown (Micro-Burst Storm) Map||OB2 Contents|
One of the objectives of this study was to document recent major landscape disturbances to the study area. About half-way through this project, a major storm event swept through northern New York across a significant portion of the study area. This section of the report describes the additional study of this storm conducted in cooperation with the NYS Department of Conservation and St. Lawrence University.
The Oswegatchie/Black study area has been the scene of major natural disturbances documented in the proposal and incorporated in the basic data for the project. On July 15, 1995, a storm event impacted thousands of acres of forest, concentrated in the Five Ponds Wilderness area, but also distributed throughout the region, including significant areas of private forest. As we contemplated working with both aerial photography and satellite data for analysis of ground conditions, this presented a valuable opportunity to extend these analyses and provided a foundation for managing the response to the catastrophe.
The Agency proposed to evaluate the spatial extent of the damage and use of remote sensing techniques for quantitative analysis of the degree of damage in cooperation with Dr. William Elberty, chair of the Geography Department at St. Lawrence University. The project was amended to incorporate cooperative acquisition of Thematic Mapper images covering this area for the period August 1994 and August 1995. Acquisition was made possible by financial contributions from the NYS Department of Environmental Conservation (the lead agency for disaster response), St. Lawrence University Department of Geography, Niagara Mohawk Power Corporation and the Adirondack Park Agency.
The primary products of this effort are: 1) classified images using two different techniques to map the storm damage, also available in digital and printed formats; and 2) secondary analyses and documentation by Dr. Elberty as the data are used for teaching and student projects at St. Lawrence University. The primary data sets are owned by the Adirondack Park Agency and maintained at the Geography Department of St. Lawrence University.
The cooperating parties evaluated the satellite scenes and acquired geometrically corrected imagery resampled to a 25 meter standard grid cell size. The raw satellite image was preprocessed by the EOSAT Corporation to provide an image with standard map coordinates suitable for comparison with other similarly processed images, or other GIS maps. While resampling simplified the analytic process, particularly for the multi-year change analysis, the additional processing slightly degraded the data, an issue that may be addressed in future academic analyses if copies of the raw data can also be provided by EOSAT.
The first step in the utilization of the data was subsetting the images to the Oswegatchie/Black study area and its vicinity for each date by St. Lawrence University. The data were next processed using a modification of change detection protocols as described in National Oceanic and Atmospheric Administration Technical Report (Dobson 1995). This involved using band 7 data from each of two dates on separate color guns in a color display. Areas with no change in data values produce a uniform green color of varying intensity, while areas of change produce shades of red, yellow and orange depending on the nature of change. This provided a map of storm damage that could be scaled and visually compared with other GIS data, but as the data were unclassified, it was not suitable for tabulation or quantitative analysis.
Dr. Elberty prepared a classified data set using the same data based on standard remote sensing classification methods. Two LANDSAT Thematic Mapper (TM) images captured August 8, 1994 and August 10, 1995 were analyzed to classify forest blowdown in the area of the Five Ponds Wilderness, Adirondack State Park. The 1994 image was classified into forest-only, agricultural/cultural, and water using a multivariate image analysis (MIA) strategy. A forest-only mask prepared from this classification was applied to the 1995 image and MIA (principal component analysis) performed on the forest area most severely affected by the windstorm. The image analysis strategy uses scatter plots generated from a principal component image that permit differentiation of non parametric classes in "statistical space". IMAGINE 8.2 software by ERDAS Inc. was used to generate the MIA. Seven classes were differentiated representing severity of damage assessment as: >HARSH, <HARSH, and LOW; and the type of forest cover: Hardwood; Mixed, and Softwood. Areas identified as "LOW" damage were not differentiated as to cover-type affected. Class descriptions are intentionally ambiguous because of a lack of detailed "ground truth" data. The damage categories are roughly equivalent to: >HARSH as major damage signifying more than 60% blowdown; <HARSH as significant damage between 30 to 60% blowdown; and LOW as occasional damage of less than 30% blowdown. Although the classes are highly inferential, comparison of the classification with a change detection image prepared by write function memory insertion using TM Band 7 of the 1994 and 1995 images indicates a high reliability of damage identification. Limited Global Positioning Systems (GPS) traverses through heavy blowdown south of the hamlet of Star Lake, N.Y. (Tamarack Creek Trail) and narrative verification from other areas suggest that the assessment is well within the framework of its ambiguity.
Both the Park Agency and the Elberty work resulted in printed and digital maps in a raster format for use with the other GIS data sets in the Oswegatchie/Black study. The Agency product is a graphic product that preserves the kind of detail found in air photography and provides opportunities for rapid assessment of blowdown conditions in comparison to other secondary maps, either in the computer or using scaled paper prints. It has been used to verify private forest inventory maps where ground access was difficult or impossible, to assess damage in proximity to developed areas, and for limited assessment of damage assessments hand digitized by the Department of Environmental Conservation from a 50-acre sample grid applied to air photography from August of 1995.
The St. Lawrence University product is a fully classified raster image suitable for quantitative analysis of the storm damage. As additional ground truth data has been gathered, the initial classification has proven to have good reliability, and is a very cost effective alternative to air photo analysis. The Elberty classified raster image is provided in digital form as a part of the digital data set for the Oswegatchie/Black project (see graphic in report attachment).
Additional technical documentation of this data set is available with the digital data copy.
|II.O. Landscape Disturbance Composite Map||OB2 Contents|
One of the objectives of this study was to test the use of an interactive tabular database, such as the Adirondack Lakes Survey Corporation water quality database, with an ArcInfo or ArcView spatial database over a large region.
Using existing GIS data layers, the pattern of major disturbance episodes to the natural communities of the 400,000-hectare Oswegatchie/Black watershed was depicted. The purpose of the analysis is to show, based upon the best available information, areas of land with and without major landscape level disturbances. The following data layers were used: State Forest Acquisition Map; 1916 Fire Protection Map of the Adirondack Forest; and 1950 Blowdown Map.
Our first assumption is that all forest lands within the Oswegatchie/Black study area, now or recently in private ownership, have been logged or cut over for some purpose and therefore support only second growth forests. Only those forest lands owned by the State of New York as Forest Preserve lands are expected to harbor old growth forest stands.
Old growth forests of the Adirondacks are defined by Leopold et al. (1988) as including, at a minimum, the following: an area of 8 hectares; no evidence of human influence; an intact and undisturbed groundlayer; old-aged canopy-dominant trees with a stand age of half the average maximum life span of the dominant tree species; a community structure with a mosaic of canopy gaps of variable size with corresponding woody debris and tree regeneration within the gaps. These criteria are designed to incorporate "a long period of growth and change of the forest without human interference".
The State Forest Acquisition Map is based upon research by Barbara McMartin (1994) documenting the year of acquisition of parcels of land by New York State and subsequently their addition to the Forest Preserve. Once added to the Forest Preserve, removal of the timber is prohibited by law. Some parcels acquired by the State had never been logged and these tracts have been documented (McMartin 1994). Further, based upon the research by McMartin, the parcels acquired by New York State before 1890 which had been logged only for softwoods are in near pristine condition. The logging techniques employed on these lands limited the amount and kind of logs that were cut primarily to spruce over 10 inches in diameter, and therefore the residual stands were left somewhat intact. Their recovery in the past hundred years leaves them remarkably similar to stands that have never been cut. For this reason, they are worth identifying as potentially unique natural old growth communities.
The 1916 Fire Protection Map shows the state of the forests after several decades of logging and widespread wildfires, most notably in the years 1903, 1908 and 1913. Stands of forests which might have been in State ownership, and presumed to be in pristine condition prior to the forest fires, might have been burned. Large areas of burns are shown on the Fire map.
The Blowdown Map of 1950 is a final layer showing disturbance of the forest stands. Moderate to severe levels of blowdown are depicted. This map does not have very fine resolution but does illustrate the extent of this disturbance to the once forested landscape.
We note that the 1995 storm blowdown disturbance was not included in this composite. The complete coverage for that storm was unavailable for the entire study area.
By combining each of these layers, it can be illustrated that much of the Oswegatchie/Black watershed has been disturbed in some way in the last century. This, however, emphasizes the fact that the undisturbed areas are potentially highly significant old growth forest communities. This new coverage is called the Landscape Disturbance Map and is illustrated on the poster insert at the back of this report.
Before assuming the veracity of these data, however, the age of canopy-dominant trees and species composition of these areas should be examined in the field. This will help to verify what are believed, at this point, to be regional or landscape-level patterns both of disturbed lands and of persisting old growth communities.
|Continue reading next section of OB2 Report -- Section II.P.||OB2 Contents|