SAV Mapping Methods

 Submerged Aquatic Vegetation Mapping
2009
 
2003
 
1996-99
 
1985-87
 
1960's
and 1970's
Gallery
 

2009

Imagery acquisition was undertaken during the summer of 2009 with film aerial photography  collected on June 28, July 7, and August 4, 2009. A number of in situ sites were visited to collect reference information to enable the interpretation of the aerial photography, as well as assess the accuracy of the resulting maps.  A multi-scale image segmentation/object-orient image classification using eCognition software, similar to that employed in the 2003 study, was used. Comparison of the classified seagrass presence/absence map and the in situ validation dataset showed an overall thematic map accuracy of 87%, while the four class seagrass density map (absent, sparse, moderate, dense seagrass cover) has an overall accuracy of 70%.  Results of this work indicate that the overall amount of seagrass beds were similar in 2009 as compared to 2003 (5,122 ha in 2003 vs. 5,260 ha in 2009).  Differences in the seasonal period of image acquisition account for some of the differences in the mapped area and type of seagrass. Due to the later growing season imagery, confirmed by the in situ data, the 2009 survey mapped greater amounts of R. maritima as compared to the 2003 survey.  We do not attribute the mapped increases in R. maritima to a “real” increase of R. maritima acreage but rather as an artifact of the difference in the timing of image acquisition. Examination of the more detailed four class seagrass cover map shows a decline in the area of dense seagrass beds in 2009 vs. 2003 (i.e., 471 ha in 2009 vs. 2,074 ha in 2003).  Whether this apparent thinning in the density of the seagrass beds is real or an artifact of the poorer image quality in the 2009 imagery and the resulting lower accuracy in mapping dense seagrass beds is uncertain. This project was in collaboration with the Jacques Cousteau National Estuarine Research Reserve.

top

2003

The purpose of this study was to map the areal extent and density of submerged aquatic vegetation within the Barnegat Bay and Little Egg Harbor, New Jersey as part of ongoing monitoring for the Barnegat Bay National Estuary Program. We examined the utility of multi-scale image segmentation/object-oriented image classification approaches to seagrass mapping. Our remotely sensed approach allowed for determination of seagrass at 4 levels of density (including shallow sand/mud flats with < 10% seagrass cover), rather than a simple presence/absence, with a comparatively high degree of consistency and accuracy (68% overall accuracy for 4 categories and 83% accuracy for a simpler presence/absence map). While the aerial digital camera imagery employed in this study had the advantage of flexible acquisition, suitable image scale, fast processing return times and comparatively low cost, it had inconsistent radiometric response across the individual images. While we were not successful in using the eCognition software to develop a rule-based classification that was universally applicable across the 14 individual image mosaics that comprised our 73,000 ha study area, the manual classification approach that we developed provided a flexible and time effective approach to mapping seagrass. This multi-scale image segmentation approach coupled with field transect/point surveys has the potential to be more replicable than strictly boat-based surveys and/or visual image interpretation and allow for more robust conclusions regarding change in areal extent, location and spatial pattern of seagrass beds through time. In the present study based on imagery collected in May 2003, we mapped 5,184 ha of seagrass beds. This area is less than the 6,083 ha of seagrass documented from boat-based surveys conducted between 1996-1999. We do not believe that the difference of 899 ha represents a significant change in seagrass extent between the dates of the two studies, but most likely is an artifact of the difference in mapping techniques.

top

1996-99

A field survey of seagrasses was conducted during the summers of 1996, 1997 and 1998 (McLain and McHale 1996) and 1999 (Bologna et al. 2000).  The middle portion of the bay was mapped in 1996, the northern portion in 1997 and the southern portion (e.g., Little Egg Harbor) in 1998. The southern portion of the bay was re-mapped in 1999 using a differentially corrected (post-processing) global positioning system (GPS). During boat-based surveys, SAV beds were identified and boundaries were mapped onto a 1:40,000 scale NOAA nautical chart (Charts # 12324 and 12316).  The dominant species (i.e., Zostera or Ruppia) was noted. These annotated charts were then table digitized and integrated with the GPS mapped data.

top

1985-87

During the years 1985 to 1987, information on eelgrass distribution, water depth and bottom sediments was collected in conjunction with an estuarine shellfish inventory of Barnegat and Little Egg Harbor bays (Joseph et al.,1992).  These surveys spanned the spring, summer and autumn seasons.  Benthic samples were collected at approximately 0.4 km (one-quarter mile) intervals in the deeper waters (greater than 0.9 m; 3 feet) of the bay, for a total of 489 stations. The extreme northern end of the Bay (i.e., Metedeconk River) was not surveyed.  Based on this survey, the distribution of eelgrass beds was then interpolated and mapped onto a nautical chart base map and produced as figures (Joseph et al. 1992).  The resulting map included two areas that were not explicitly sampled (i.e. the boat did not enter due to the shallow depth and no benthic samples were taken) but where visual reconnaissance noted the occurrence of eelgrass beds.  These two areas were included as containing eelgrass  (for a total of 896 ha)  in our analysis. We table digitized photocopies of the maps, using existing shoreline GIS data as a base map to provide ground control points.

top

1960's and 1970's

Mapped information on the spatial distribution of seagrass beds for Barnegat Bay was derived from several sources. The first systematic survey was undertaken in 1968 (U.S. Army Corps of Engineers 1976). The methods for this study were not detailed but are presumed to be a boat-based survey. There were several mapping efforts during the 1970's. The lower portion of Little Egg Harbor was mapped based on spring time aerial photography acquired in 1977, as a pilot project to examine the feasibility of mapping submerged aquatic vegetation (SAV), including seagrass beds, with aerial photography (Good et al. 1978). Based on the success of this project, the Earth Satellite Corporation mapped the state’s entire Atlantic coast and produced a 1:24,000 scale map series for the entire bay based on interpretation of black & white aerial photography and low altitude sea plane reconnaissance during the summer of 1979 (photos taken June and August, field checked July through September) (Macomber and Allen 1979). For both the 1968 and 1979 survey, four general types of SAV communities were mapped (eelgrass Zostera marina, widgeongrass Ruppia maritima, mixed eelgrass and widgeongrass and sea lettuce Ulva lactuca dominated) at various levels of density. We table digitized the above paper maps for later GIS analysis using existing shoreline GIS data as a base map. The U.S. Fish & Wildlife Service (USFWS) incorporated the Earth Satellite Corporation maps into the National Wetland Inventory (NWI) for the state of New Jersey.

top
Sav
Home
 
View
Maps
 
Citations
 
 
GIS Data,
Metadata,
Reports
FAQ /
About SAV
 

CRSSA SAV web site designed and maintained by CRSSA. Original 2002 web site start-up funding by the New Jersey Department of Environmental Protection (NJDEP) through the Jacques Cousteau National Estuarine Research Reserve (JCNERR) for JCNERR Coastal Decision Maker and education outreach programs; continued web updates with support from CRSSA. SAV site is a segment of CRSSA's Coastal Studies website. Page last updated May 4, 2011.. See front/home SAV page for more credit information.