High Spatial Resolution Remote Sensing for Salt Marsh Mapping and Change Analysis at Fire Island National Seashore

Campbell, Andrew R.; Wang, Yeqiao

doi:10.3390/rs11091107

Cited by 36 publications

(22 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The objects were generated through the 'Mean-Shift' segmentation algorithm included in the Orfeo Toolbox (OTB 6.6.0) [51], and available in QGIS as an external plugin. This method has been used in other studies with satisfactory results [20,52,53]. It is worth mentioning that the polygon heterogeneity (i.e., size of the objects) is controlled by user-defined parameters.…”

Section: Image Segmentation With Obiamentioning

confidence: 99%

“…Fire Island is currently constrained by two main tidal inlets used as navigation channels: Fire Island Inlet to the west and Moriches Inlet to the east. The barrier island has been affected by extreme storm surges, such as Hurricane Sandy in 2012 which caused overwash on 41% of Fire Island and modified its shape due to large amount of sediment accumulation (approximately 50,800 m 3 ) [20]. From fall 2014 to spring 2016, an extensive RTK topographic survey over salt marsh regions was conducted on Fire Island as described in Section 2.2.…”

Section: Study Areamentioning

confidence: 99%

“…Over the last decade, the application of random forest (RF) and object-based image analysis (OBIA) to imagery datasets has proven to be a robust, stable and superior methodology for land cover mapping and land change analysis on a variety of landscapes (e.g., urban areas, wetlands, agricultural areas, natural forests) [19][20][21][22]. RF and OBIA present advantages over other classifiers (e.g., decision tree, maximum likelihood, k-nearest neighbor, support vector machine, artificial neural network) and over traditionally (sub)pixel-based classification techniques (e.g., supervised, unsupervised, contextual, hybrid, fuzzy, spectral mixture) [23][24][25].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Adjusting Emergent Herbaceous Wetland Elevation with Object-Based Image Analysis, Random Forest and the 2016 NLCD

2019

View full text Add to dashboard Cite

Emergent herbaceous wetlands are characterized by complex salt marsh ecosystems that play a key role in diverse coastal processes including carbon storage, nutrient cycling, flood attenuation and shoreline protection. Surface elevation characterization and spatiotemporal distribution of these ecosystems are commonly obtained from LiDAR measurements as this low-cost airborne technique has a wide range of applicability and usefulness in coastal environments. LiDAR techniques, despite significant advantages, show poor performance in generation of digital elevation models (DEMs) in tidal salt marshes due to large vertical errors. In this study, we present a methodology to (i) update emergent herbaceous wetlands (i.e., the ones delineated in the 2016 National Land Cover Database) to present-day conditions; and (ii) automate salt marsh elevation correction in estuarine systems. We integrate object-based image analysis and random forest technique with surface reflectance Landsat imagery to map three emergent U.S. wetlands in Weeks Bay, Alabama, Savannah Estuary, Georgia and Fire Island, New York. Conducting a hyperparameter tuning of random forest and following a hierarchical approach with three nomenclature levels for land cover classification, we are able to better map wetlands and improve overall accuracies in Weeks Bay (0.91), Savannah Estuary (0.97) and Fire Island (0.95). We then develop a tool in ArcGIS to automate salt marsh elevation correction. We use this ‘DEM-correction’ tool to modify an existing DEM (model input) with the calculated elevation correction over salt marsh regions. Our method and tool are validated with real-time kinematic elevation data and helps correct overestimated salt marsh elevation up to 0.50 m in the studied estuaries. The proposed tool can be easily adapted to different vegetation species in wetlands, and thus help provide accurate DEMs for flood inundation mapping in estuarine systems.

show abstract

Section: Image Segmentation With Obiamentioning

confidence: 99%

Section: Study Areamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Adjusting Emergent Herbaceous Wetland Elevation with Object-Based Image Analysis, Random Forest and the 2016 NLCD

2019

View full text Add to dashboard Cite

show abstract

“…The Coastal Mapping Program (CMP) (https://www.ngs.noaa.gov/ RSD/cmp.shtml) run by the National Oceanic and Atmospheric Administration's (NOAA) National Geodetic Survey (NGS) has well-developed procedures for mapping salt marsh shorelines [19][20][21]. Campbell and Wang [22] used imagery analysis to study the extent of marsh grass, and the evolution of pannes and pools. Kulei, Guneroglu and Dihkan [23] used Landsat imagery and an automated image analysis technique to extract shorelines in salt marshes in Turkey.…”

Section: Introductionmentioning

confidence: 99%

Identifying Salt Marsh Shorelines from Remotely Sensed Elevation Data and Imagery

2019

View full text Add to dashboard Cite

Salt marshes are valuable ecosystems that are vulnerable to lateral erosion, submergence, and internal disintegration due to sea level rise, storms, and sediment deficits. Because many salt marshes are losing area in response to these factors, it is important to monitor their lateral extent at high resolution over multiple timescales. In this study we describe two methods to calculate the location of the salt marsh shoreline. The marsh edge from elevation data (MEED) method uses remotely sensed elevation data to calculate an objective proxy for the shoreline of a salt marsh. This proxy is the abrupt change in elevation that usually characterizes the seaward edge of a salt marsh, designated the "marsh scarp." It is detected as the maximum slope along a cross-shore transect between mean high water and mean tide level. The method was tested using lidar topobathymetric and photogrammetric elevation data from Massachusetts, USA. The other method to calculate the salt marsh shoreline is the marsh edge by image processing (MEIP) method which finds the unvegetated/vegetated line. This method applies image classification techniques to multispectral imagery and elevation datasets for edge detection. The method was tested using aerial imagery and coastal elevation data from the Plum Island Estuary in Massachusetts, USA. Both methods calculate a line that closely follows the edge of vegetation seen in imagery. The two methods were compared to each other using high resolution unmanned aircraft systems (UAS) data, and to a heads-up digitized shoreline. The root-mean-square deviation was 0.6 meters between the two methods, and less than 0.43 meters from the digitized shoreline. The MEIP method was also applied to a lower resolution dataset to investigate the effect of horizontal resolution on the results. Both methods provide an accurate, efficient, and objective way to track salt marsh shorelines with spatially intensive data over large spatial scales, which is necessary to evaluate geomorphic change and wetland vulnerability.

show abstract

“…Remote sensing technologies [4] have proven to be effective tools in mapping the wetland distribution. One of the major imagery sources is multispectral data, including Landsat [5,6], Sentinel [7,8] and WorldView [9,10]. Unmanned aerial vehicles (UAV) [11] and hyperspectral sensors [12,13] are also frequently applied in wetland classification over the past few years.…”

Section: Introductionmentioning

confidence: 99%

Wetland Classification Based on a New Efficient Generative Adversarial Network and Jilin-1 Satellite Image

Mei

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Recent studies have shown that deep learning methods provide useful tools for wetland classification. However, it is difficult to perform species-level classification with limited labeled samples. In this paper, we propose a semi-supervised method for wetland species classification by using a new efficient generative adversarial network (GAN) and Jilin-1 satellite image. The main contributions of this paper are twofold. First, the proposed method, namely ShuffleGAN, requires only a small number of labeled samples. ShuffleGAN is composed of two neural networks (i.e., generator and discriminator), which perform an adversarial game in the training phase and ShuffleNet units are added in both generator and discriminator to obtain speed-accuracy tradeoff. Second, ShuffleGAN can perform species-level wetland classification. In addition to distinguishing the wetland areas from non-wetlands, different tree species located in the wetland are also identified, thus providing a more detailed distribution of the wetland land-covers. Experiments are conducted on the Haizhu Lake wetland data acquired by the Jilin-1 satellite. Compared with existing GAN, the improvement in overall accuracy (OA) of the proposed ShuffleGAN is more than 2%. This work can not only deepen the application of deep learning in wetland classification but also promote the study of fine classification of wetland land-covers.

show abstract

High Spatial Resolution Remote Sensing for Salt Marsh Mapping and Change Analysis at Fire Island National Seashore

Cited by 36 publications

References 54 publications

Adjusting Emergent Herbaceous Wetland Elevation with Object-Based Image Analysis, Random Forest and the 2016 NLCD

Adjusting Emergent Herbaceous Wetland Elevation with Object-Based Image Analysis, Random Forest and the 2016 NLCD

Identifying Salt Marsh Shorelines from Remotely Sensed Elevation Data and Imagery

Wetland Classification Based on a New Efficient Generative Adversarial Network and Jilin-1 Satellite Image

Contact Info

Product

Resources

About