journal article
Download Only Collection
Home
Rosso, P. H.; Ustin, S. L.; Hastings, A.
doi: 10.1080/01431160500218770pmid: N/A
Sustainable management of wetland ecosystems requires monitoring of vegetation dynamics, which can be achieved through remote sensing. This paper assesses the use of hyperspectral imagery to study the structure of wetlands of San Francisco Bay, California, USA. Spectral mixture analysis (SMA) and multiple endmember spectral mixture analysis (MESMA) were applied on an AVIRIS (Airborne Visible and Infrared Imaging Spectrometer) image to investigate their appropriateness to characterize marshes, with emphasis on the Spartina species complex. The role of rms. error as a measure of model adequacy and different methods for image endmember extraction were also evaluated. Results indicate that both SMA and MESMA are suitable for mapping the main components of the marsh, although MESMA seems more appropriate since it can incorporate more than one endmember per class. rms. error was shown not to be a measure of SMA model adequacy, but it can be used to help to assess model adequacy within groups of related models.
doi: 10.1080/01431160500218911pmid: N/A
Multiple endmember spectral mixture analysis (MESMA) was applied to the Airborne Visible and Infrared Imaging Spectrometer (AVIRIS) imagery of a salt marsh in China Camp at San Pablo Bay, California. A nine‐endmember set representing materials within the scene was used including: two Salicornia and two soils, and Grindelia, Spartina, dry grass, water and shade. The resultant abundance maps were used to investigate the spatial distribution of the marsh vegetation species, Salicornia virginica, Grindelia Stricta and Spartina foliosa. The Spartina abundance map exhibited a well‐defined zone bordering the water and the lower marsh, which is in good agreement with the field observations made in 2002. Comparison of the Salicornia map with all six field global positional system (GPS) polygons indicates Salicornia was classified with high accuracy. The proposed approach did a good job in classifying Spartina and Salicornia which cover 93.81% of the total marsh. The Grindelia fraction image underestimates in some areas, while in other areas it shows false detection. This misclassification is attributed to the spectral similarity between Grindelia and Salicornia and to the small patch size of Grindelia. Further work is required to solve this confusion.
doi: 10.1080/01431160500218952pmid: N/A
A study was conducted to map the plant vigour gradient using hyperspectral imagery combined with field‐collected seasonal reflectance spectra of marsh species in a fragmented coastal wetland. Marsh surface types were identified by classifying a low‐tide hyperspectral image of New Jersey Meadowlands acquired by Airborne Imaging Spectroradiometer for Applications (AISA) in October 2000. Reflectance spectra of dominant marsh species and seasonal spectra of Phragmites australis from April to October in 2003 were measured in the field. The separability of field‐collected spectra of marsh species was tested and determined using non‐parametric U‐test. The spectra of Phragmites in greening‐up phases were used as the surrogate to determine the vigour gradient of common reed stands. The results demonstrate that the reflectance spectra of one healthy stand of Phragmites sampled across the growing season did provide equivalent signatures of pure physiognomic types, which can be used to determine the plant vigour gradient.
Morris, James T.; Porter, Dwayne; Neet, Matt; Noble, Peter A.; Schmidt, Laura; Lapine, Lewis A.; Jensen, John R.
doi: 10.1080/01431160500219018pmid: N/A
Vertical elevation relative to mean sea level is a critical variable for the productivity and stability of salt marshes. This research classified a high spatial resolution Airborne Data Acquisition and Registration (ADAR) digital camera image of a salt marsh landscape at North Inlet, South Carolina, USA using an artificial neural network. The remote sensing‐derived thematic map was cross‐referenced with Light Detection and Ranging (LIDAR) elevation data to compute the frequency distribution of marsh elevation relative to tidal elevations. At North Inlet, the median elevation of the salt marsh dominated by Spartina alterniflora was 0.349 m relative to the North American Vertical Datum 1988 (NAVD88), while the mean high water level was 0.618 m (2001 to May, 2003) with a mean tidal range of 1.39 m. The distribution of elevations of Spartina habitat within its vertical range was normal, and 80% of the salt marsh was situated between a narrow range of 0.22 m and 0.481 m. Areas classified as Juncus marsh, dominated by Juncus roemerianus, had a broader, skewed distribution, with 80% of the distribution between 0.296 m and 0.981 m and a median elevation of 0.519 m. The Juncus marsh occurs within the intertidal region of brackish marshes and along the upper fringe of salt marshes. The relative elevation of the Spartina marsh at North Inlet is consistent with recent work that predicts a decrease in equilibrium elevation with an increasing rate of sea‐level rise and suggests that the marshes here have not kept up with an increase in the rate of sea‐level rise during the last two decades.
doi: 10.1080/01431160500219133pmid: N/A
This study was intended to apply the technique of derivative analysis to estimate algal chlorophyll concentration in Pensacola Bay, Florida. The data collection was conducted over the 16 sampling sites on three separate occasions. A portable field spectroradiometer was used to collect the upwelling radiance of water and reference panel at each sampling station. The instrument records a continuous spectrum in 512 bands ranging from 350 nm to 1050 nm, with 1.37 nm spectral resolution. The first derivatives were computed and correlated with the chlorophyll‐a concentration. The results indicated the first derivatives at 630–645 nm, 660–670 nm, 680–687 nm and 700–735 nm were correlated strongly with chlorophyll‐a. The R values reached 0.858 for the wavelength at 686.7 nm. The results support the hypothesis that derivative spectra are less impacted by wave effects and, therefore, are an effective tool for estimating chlorophyll concentration.
Han, Luoheng; Jordan, Karen J.
doi: 10.1080/01431160500219182pmid: N/A
The purpose of this study was to develop algorithms for estimating chlorophyll‐a concentration in Pensacola Bay using Landsat 7 ETM+ data. The techniques used were band ratioing and regression modelling. Pensacola Bay is located on the west end of the Florida panhandle. As one of 39 estuaries located on the Gulf of Mexico, Pensacola Bay is impacted largely by rivers. The Landsat ETM+ data were first geometrically rectified. Then brightness values were converted to reflectance through the radiometric correction process. For the regression models, logarithmically transformed chlorophyll‐a was used as the dependent variable. Single bands, band ratios and logarithmically transformed band ratios were the independent variables. R 2 values were computed and evaluated. Results from the study indicate that the ratio of ETM+1/ETM+3 was the most effective in estimating chlorophyll‐a. Using this model a chlorophyll‐a map was generated for Pensacola Bay.
Wolter, Peter T.; Johnston, Carol A.; Niemi, Gerald J.
doi: 10.1080/01431160500219208pmid: N/A
Submergent aquatic vegetation (SAV) is a powerful indicator of environmental conditions in both marine and fresh water ecosystems. Quickbird imagery was used to map SAV at three sites across the Great Lakes. Unsupervised classifications were performed at each site using summer Quickbird sensor data. At one site, a multi‐temporal classification approach was added, combining visible red difference (May–August) with August red and green visible band data. Multi‐temporal SAV classification was superior to single‐date results at this site. Muck bottom was not seriously confused with SAV, which was unexpected. Multi‐temporal classification results showed less confusion between deep water and SAV, although spectral variability due to sub‐surface sandbar structure was a source of error in both single‐ and multi‐date classifications. Nevertheless, some of the confounding effects of water column on SAV classification appear to have been mitigated using this multi‐temporal approach. Future efforts would be well served by incorporating detailed, continuous, bathymetry data in the classification process. Quickbird sensor data are very useful for classifying SAV under US Great Lakes conditions. However, regional classification efforts using these data may be impractical at this time, as high cost, rigid tasking parameters and unpredictable water conditions limit availability of suitable imagery.
doi: 10.1080/01431160500219224pmid: N/A
The degradation of world‐wide estuarine ecosystems as a result of accelerated human population growth accompanied by agricultural, industrial and urban development justifies a strong need to find efficient ways to manage and protect these sensitive environments. Starting from 2001, the authors have been involved in an interdisciplinary research project aiming to develop environmental indicators for integrated estuarine ecosystem assessment in the Gulf of Mexico. As part of this project, a study was conducted to characterize land‐use and land‐cover changes with the Pensacola estuarine drainage area as a case. The Pensacola bay was targeted because it is one of few exemplary large river‐driven estuarine systems across the northern Gulf of Mexico. The study had two major sections. The first part was dedicated to the development of an improved method for coastal land‐use and land‐cover mapping, which was built upon hierarchical classification and spatial reclassification. An image scene was separated into urban and rural regions early in the classification, with a ‘mask’ defined by road intersection density slices combined with road buffers. Each part was classified independently in its most effective context and, later, both were merged to form a complete map. In spatial reclassification, image interpretation procedures, auxiliary vector data and a variety of Geographical Information System (GIS) functions were synthesized to resolve spectral confusion and improve mapping accuracy. This method was used to map land use and land cover from Landsat Thematic Mapper/Enhanced Thematic Mapper Plus (TM/ETM+) imagery for 1989, 1996 and 2002, respectively. The accuracy assessment shows that the overall classification errors were less than 10%. The second part focused on the analysis of the spatio‐temporal dynamics of estuarine land‐use and land‐cover changes by using post‐classification comparison and GIS overlay techniques. The project has revealed that a substantial growth of low‐density urban land occurred in the lower drainage basin in connection with population and housing growth, as well as a significant increase of mixed forest land in the upper watershed as a result of active logging and harvesting operations. These growths were achieved at the cost of evergreen forest and wetlands, thus compromising safeguards for water quality, biodiversity of aquatic systems, habitat structure and watershed health in the Pensacola estuarine drainage area.
doi: 10.1080/01431160500219273pmid: N/A
Degradation of estuarine ecosystems caused by human‐induced stressors justifies finding efficient ways to manage and protect these environments. This study demonstrates the utilities of satellite remote sensing, landscape metrics and multivariate statistical analysis for quantifying landscape pattern and its change in a highly sensitive estuarine watershed. The objective of this study was to identify the appropriate method for landscape pattern characterization in the Pensacola estuarine drainage area (PEDA) as part of an interdisciplinary effort to develop environmental indicators for integrated estuarine ecosystem assessment in the Gulf of Mexico. The study has several components. First, two land‐use and land‐cover maps were produced from satellite imagery by using hierarchical classification and spatial reclassification techniques. Then, 56 metrics of landscape composition or configuration were computed from the two maps for different spatial observational units, including the PEDA, four sub‐watersheds, and three predefined buffer areas. Because some of the landscape metrics may be correlated with each other, landscape ecology principles, principal component analysis and Spearman's rank correlation analysis were used to eliminate redundant metrics. This resulted in a parsimonious set of core metrics which were not redundant but spanned the important dimensions of landscape structure and pattern. These core metrics were finally used to quantify landscape pattern for different spatial observational units at the two different years. Landscape structure has been found to be more fragmented in the Pensacola Bay watershed, around the city centres and along the coastlines, where urbanization and human economic activities are more concentrated. Over time, the landscape mosaics became more heterogeneous while the classes of patches tended to be more fragmented. Results of this study should help coastal managers in the PEDA target those areas in need of conservation and protection.
Showing 1 to 10 of 13 Articles