Submerged aquatic vegetation (SAV) is an important habitat and site of primary production in many aquatic ecosystems but there was no baseline information on SAV extent or distribution in the tidal freshwater Hudson River. In 1994, a collaboration was initiated between the Institute of Ecosystem Studies (IES), the Hudson River National Estuarine Research Reserve/NYSDEC, the Cornell Laboratory for Environmental Applications of Remote Sensing (CLEARS), now the Cornell Institute for Resource Information Systems (IRIS) <http://www.css.cornell.edu/iris/index.htm>. In addtion, the New York Sea Grant, the Hudson River Estuary Program, and the National Oceanic and Atmospheric Administration (NOAA) are identified as partners. These groups provided diverse expertise to enable the first broad delimitation of SAV in the Hudson.
The project was undertaken in two separate time periods with different sources of funding. In 1995, Phase I (Hyde Park to Castleton) was initiated with National Oceanic and Atmospheric Association and Hudson River Foundation funds. Subsequently in 1997, the remaining portions (Hastings to Hyde Park and Castleton to Troy) were undertaken in Phase II with New York. State Environmental Protection Funds through the Hudson River Estuary Program.
High quality, true color aerial photographs were acquired for Phase I on August 9, 1995 and Phase II on July 20, August 7 and August 18 in 1997 by AeroGraphics, Inc. of Bohemia, New York. Photographs were acquired with 60% forward lap and 30% side lap, at the nominal scale of 1:14,400 under ideal conditions within one hour of low tide and clear atmospheric conditions. Flight and mapping criteria followed the 1994 NOAA C-CAP protocols for SAV mapping. Details are available from <http://www.csc.noaa.gov/>. Stereoscopic aerial photographic coverage was available for the entire study area. The effective area of alternate aerial photographs was established and delineated on 3 mil matte acetate photo overlays. Major tributaries were not mapped with the exception of Schodack Creek. For more information about the project, please visit the IRIS website at: <http://www.css.cornell.edu/iris/rip/hudson/hudsonsav.htm>. Shoreline Coverage-- The four-category classification system was developed through consultations among IES, HRNERR and IRIS. The four categories (numerical codes) are Open Water (VEGCODE=0), Upland/Intertidal (VEGCODE=1); Vallisneria americana (VEGCODE=2), and Trapa natans (VEGCODE=3). Of primary interest is Vallisneria americana, the predominant species of SAV. It is important to note that our Upland/Intertidal category includes intertidal marshes and mudflats which were largely exposed at time of photography. OW - Open Water VEGCODE=0" - Water areas including the Hudson River, tributary rivers, bays, coves with no indicators of submerged rooted vegetation, upland vegetation or dry land. UP/INT - Upland/Intertidal VEGCODE=1 - Dry land with or without upland vegetation or intertidal marsh vegetation. VA - Vallisneria americana (water or wild celery, tape grass) VEGCODE=2 - Beds of Vallisneria americana as well as beds of mixed vegetation including Vallisneria americana. TN - Trapa natans (water chestnut, water nut) VEGCODE=3 - Beds of Trapa natans, rooted with floating leaves. Linear Coverage-- During ground-truthing of polygon beds north of Albany, observations were made of frequent occurrences of very narrow linear beds of Vallisneria located adjacent to the shoreline. These features generally appeared to be in the two-to-three meter width size, and because they did not provide a signature were not visible on the color aerial photographs. In some instances, the plants were over-shadowed by tree cover along the shoreline. These features were subsequently mapped from the ground. Mapping of linear features, grab samples and GPS quality control functions resulted from a collaboration between Cornell IRIS and the Institute of Ecosystems Studies (IES). IES performed digitization; data merging and conversion performed by IRIS. Data have been converted from NAD27 to NAD83.
Coverages are updated as amendments occur, or as errors in the data are discovered and corrected.
All requests for SAV data should be referred to the Cornell University Geospatial Information Repository or refer to the contact information below.
Hudson River NERR c/o Bard College Field Station
Hudson River National Estuarine Research Reserve (NERR) maintains the Hudson SAV data
Delineation of the four land cover categories (Open Water-Vegcode=0; Upland/Intertidal-Vegcode=1; Vallisneria americana-Vegcode=2; Trapa natans-vegcode=3) was accomplished by the use of color pencil lines and labels drawn on the photo overlay as the stereo model was viewed using an Abrams 2X and 4X stereoscope or an Old Delft Scanning stereoscope at 6X.
The inventory included broad delimitation of submerged aquatic vegetation (SAV) beds and T. natans in the Hudson River using aerial photography for primary data collection. Field sampling of random points was used to verify the photo interpretation, generally following established protocols. In the case of SAV linear features, because their signatures were not detectable on the aerial photographs, were mapped in the field using a global positioning system. To verify presence of plants, randomly selected grab samples were taken at 141 field locations.
The image analyst used her experience with key indicator signatures for object recognition and interpretation with true color aerial photographs in the SAV inventory. Vallisneria americana (water celery), for example, varies in color from light green-brown to dark green-brown to dark green-blue. The bed texture is smooth, velvety to fine bumpy; depends on plant density, species composition, and bed location at or below the water surface. The leaves of Trapa natans (water chestnut) are light green to yellow-green to creamy white. It appears to have smooth bed texture due to uniform leaf size and shape.
Coverages are accurate to the quality of the original topographic map
Spatial and attribute information
Base maps for mapping were created through photographic reproductions of the topographic quadrangle maps covering the study area. Each image was masked to include the river and surrounding 2.5 minute latitude/longitude tic marks. Black and white film negatives at 1:48,000 scale were printed onto photographic paper at the 1:14,400 final mapping scale to coincide with aerial photographs. Scale was verified by measurement. Delineation of the four land cover categories (described below) was accomplished by the use of color pencil lines and labels drawn on the photo overlay as the stereo model was viewed using an Abrams 2X and 4X stereoscope or an Old Delft Scanning stereoscope at 6X. The minimum mapping unit for delineation was based upon the smallest polygon that could be drawn with a 0.5 mm lead pencil. A polygon with an average diameter of 1.0 mm could be drawn. At 1:14,400 scale this corresponds to a ground area of 0.016 ha (equivalent spherical diameter roughly 15 m). After initial trials of delineation in areas of known ground conditions, interpretation proceeded from the southern boundary of the study area to the northern boundary. The image analyst did transfer through manual rectification transferring photo delineations to a 4 mil Mylar overlay registered to the photographically reproduced topographic base map. A 0.5 mm polymer lead pencil was used for lines and a colored red pencil for labels. Registration marks were also transferred to the map overlay from the base map. These included 2.5 minute latitude/longitude tic marks and USGS benchmarks. The completed polygon map overlays were edited to ensure that all lines and labels had been transferred accurately and an edge match edit was performed. Prior to the transfer process, interpretation and delineation of photo overlays was reviewed. After transfer, maps were reviewed for clarification on mapping of specific sites. Aerial photographic interpretation and mapping completed by Cornell IRIS, Ithaca, NY. Linear SAV were digitized as line features obtained from the Mylar overlays that were prepared from field observations and spot sampling using a CalComp Drawing Board II digitizing board and mouse with ArcInfo software. Linear features were present on three topographic quadrangles (Albany, Delmar, Troy South). Tolerances (Weed, Arc, Snap) were all set to zero. Four Tic marks were used for Albany (Fuzzy tolerance = 0.08) and eight each for Delmar and Troy South (Fuzzy tolerance = 0.23 and 1.20 respectively). Tics for each map were entered in the NAD27 latitude/longitude coordinates and registered on the map using the digitizing mouse. All linear features on the overlays were large enough for digitization. All other digitization procedures are consistent with digitization of SAV polygons. Visual comparisons were periodically made between Mylar overlays and digitized features. Proof plots of linear features were created at the scale of 1:14,400 for comparison with original map overlays to ensure accurate representation. The digitized linear feature data was edited for cartographic consistency. The lines representing submerged aquatic vegetation were adjusted to reflect the appropriate generalized position of the linear features. Maps were transferred to ArcView for presentation and statistical calculations.
Metadata imported.
Dataset moved.
Internal feature number.
ESRI
Feature geometry.
ESRI
Internal feature number.
ESRI
Feature geometry.
ESRI
Area of feature in internal unites squared.
ESRI
Perimeter of feature in internal units.
ESRI
Internal feature number.
ESRI
User defined feature number.
ESRI
Unspecified.
ESRI
Unspecified.
Unspecified
Cornell University
Cornell University provides these geographic data "as is." Cornell University makes no guarantee or warranty concerning the accuracy of information contained in the geographic data. Cornell University further makes no warranty either expressed or implied, regarding the condition of the product or its fitness for any particular purpose. The burden for determining fitness for use lies entirely with the user. Although these files have been processed successfully on computers at Cornell University, no warranty is made by Cornell University regarding the use of these data on any other system, nor does the fact of distribution constitute or imply any such warranty.
Hudson River NERR c/o Bard College Field Station