Identification_Information: Citation: Citation_Information: Originator: YQ Wang Originator: Mike Traber Publication_Date: Unpublished Material Publication_Time: Unknown Title: sav_final.img Geospatial_Data_Presentation_Form: remote-sensing image Description: Abstract: Classification of Submerged Aquatic Vegetation (SAV) from Quickbird-2 Satellite imagery for the Fire Island National Seashore. An unsupervised classification was performed to classify the seagrass habitats only within the areas of the park boundary. Purpose: This data was created as part of National Park Service funded project to develop a protocol for mapping terrestrial and submerged aquatic habitats using satellite imagery. The SAV data presented here is to be used for monitoring of critical seagrass habitat. Time_Period_of_Content: Time_Period_Information: Range_of_Dates/Times: Beginning_Date: May 2004 Ending_Date: July 2004 Currentness_Reference: ground condition Status: Progress: Complete Maintenance_and_Update_Frequency: None planned Spatial_Domain: Bounding_Coordinates: West_Bounding_Coordinate: -73.309229 East_Bounding_Coordinate: -72.752818 North_Bounding_Coordinate: 40.789466 South_Bounding_Coordinate: 40.616398 Keywords: Theme: Theme_Keyword_Thesaurus: None Theme_Keyword: Digital Images Theme_Keyword: Vegetation Mapping Theme_Keyword: Points Theme_Keyword: Quickbird-2 Theme_Keyword: seagrass Theme_Keyword: Zostera Theme_Keyword: marina Theme_Keyword: eelgrass Theme_Keyword: ruppia Theme_Keyword: maritima Place: Place_Keyword_Thesaurus: None Place_Keyword: Fire Island Place_Keyword: FIIS Place_Keyword: Long Island Place_Keyword: Great South Bay Place_Keyword: New York Place_Keyword: USA Access_Constraints: none Use_Constraints: none Point_of_Contact: Contact_Information: Contact_Person_Primary: Contact_Person: YQ Wang Contact_Organization: University of Rhode Island, Dept of Natural Resources Science Contact_Position: Professor Contact_Address: Address_Type: mailing and physical address Address: University of Rhode Island Address: Coastal Institute Building Address: 1 Greenhouse Rd City: Kingston State_or_Province: RI Postal_Code: 02881 Country: USA Contact_Voice_Telephone: 401 874-4345 Contact_Electronic_Mail_Address: yqwang@gso.uri.edu Native_Data_Set_Environment: Microsoft Windows XP Version 5.1 (Build 2600) Service Pack 2; ESRI ArcCatalog 9.0.0.535 Data_Quality_Information: Attribute_Accuracy: Attribute_Accuracy_Report: Accuracy assessment was conducted using ERDAS Imagine. 50 points were randomly selected within each category. Each point was reviewed by the image analyst and categorized independently using image and reference data. The reference data included, the original Quickbird-2 image, 2003 1:12,000 color orthophotography, 2003 NY DOS manual delineation of seagrass coverage, underwater video transects, field photos, and personal knowledge of the study site. Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 74.7 % Attribute_Accuracy_Explanation: Overall Classification Accuracy Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 0.62 Attribute_Accuracy_Explanation: Overall Kappa Statistic Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 80 % Attribute_Accuracy_Explanation: High Density Seagrass, Producers Error Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 72 % Attribute_Accuracy_Explanation: High Density Seagrass, Users Error Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 76.74 % Attribute_Accuracy_Explanation: Low Density Seagrass, Producers Error Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 66.0 % Attribute_Accuracy_Explanation: Low Density Seagrass, Users Error Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 69.35 % Attribute_Accuracy_Explanation: Sand, Producers Error Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 86 % Attribute_Accuracy_Explanation: Sand, Users Error Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: 0.6 Attribute_Accuracy_Explanation: High Density Seagrass, Kappa Statistic Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: .52 Attribute_Accuracy_Explanation: Low Density Seagrass, Kappa Statistic Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: .76 Attribute_Accuracy_Explanation: Sand, Kappa Statistic Logical_Consistency_Report: This is a grid image with 2.5 meter pixels. No topology is present. Spatial and attribute accuracy was performed. Completeness_Report: All deep water habitats (> 2.0 meter MLW) were omitted from the classification due to conflicts of signature between deep water and seagrass. Areas with high surface sky glint due to surface roughness were also omitted for classification. Positional_Accuracy: Horizontal_Positional_Accuracy: Horizontal_Positional_Accuracy_Report: All image processing was conducted using ERDAS Imagine software version 8.7. Each scene of QuickBird images was independently reregistered using 2003 color aerial orthophotography as the base. The reported accuracy of the orthophotography is documented in the meta-data delivered along with the image was 5.5 meters. Between 10 and 20 control points were selected for each image scene. Each scene was registered to an accuracy of 0.5 pixels or 1.25 meters, determined from the root mean square error term provided by the software (ERDAS Imagine). The classified result was not altered spatially from the original image scene therefore the original positional accuracy is the same for the classification result as for the image. Quantitative_Horizontal_Positional_Accuracy_Assessment: Horizontal_Positional_Accuracy_Value: 6.25 Horizontal_Positional_Accuracy_Explanation: Meters, the spatial accuracy of the registered and mosaiced image scene. Vertical_Positional_Accuracy: Vertical_Positional_Accuracy_Report: none reported Lineage: Source_Information: Source_Scale_Denominator: 24000 Type_of_Source_Media: CD-ROM Source_Time_Period_of_Content: Time_Period_Information: Range_of_Dates/Times: Beginning_Date: May 2004 Ending_Date: July 2004 Source_Citation_Abbreviation: Digital Globe Source_Contribution: Imagery purchased from Digital Globe Process_Step: Process_Description: All image processing was conducted using ERDAS Imagine software version 8.7. Each scene of QuickBird images was independently reregistered using 2002 color aerial orthophotography as the base. The reported accuracy of the orthophotography is documented in the meta-data delivered along with the image was 5.5 meters. Between 10 and 20 control points were selected for each image scene. Each scene was registered to an accuracy of 0.5 pixels or 1.25 meters, determined from the root mean square error term provided by the software (ERDAS Imagine). This would yield a combined accuracy of 6.75 meters between our registration and the reported accuracy of the base image. Process_Date: Unknown Process_Contact: Contact_Information: Contact_Person_Primary: Contact_Person: YQ Wang Contact_Organization: University of Rhode Island Contact_Position: Professor Contact_Address: Address_Type: mailing and physical address Address: 1 Greenhouse Rd Address: University of Rhode Island City: Kingston State_or_Province: RI Postal_Code: 02881 Country: United States Contact_Voice_Telephone: 401 874-4035 Contact_Electronic_Mail_Address: yqwang@uri.edu Hours_of_Service: 8:30-4:30 EST Contact_Instructions: Phone or E-mail Process_Step: Process_Description: Image preprocessing. The registered image was clipped using a rough area of interest (AOI) to remove the majority of land pixels or water pixels depending on which was to be classified. Once registered the individual scenes were processed to mask the land pixels from the scene. The land mask was created by conducting a 2 class unsupervised classification using only the near infrared band. This technique accurately separates the wet areas from dry areas as water has a large influence on the return of near-IR radiation. The resultant classification was used to mask the land pixels from the image. The masking step was not performed for the terrestrial classification as it was important to include the in-land water component for the classification purposes. Process_Date: November 2005 to May 2005 Process_Step: Process_Description: Image Classification of seagrass habitats. Each image scene was classified individually. This level of individual classification is necessary to compensate for subtle environmental differences observed in each scene. These differences include changes in surface condition caused by high winds or areas of varying fetch. The high winds cause surface waves which scatter incoming light reflecting as sky glint back to the sensor washing out the subsurface signature. A second factor is changes in water depth due to different tidal heights. If the images were taken on different dates the water depth will vary from scene to scene causing changes in bottom signatures. Finally, varying water column clarity, from day to day can effect the albedo of the seagrass signatures. The processing software uses changes in the reflectance of each pixel to decide what category that pixel should be classified as. If the all the scenes were mosaiced together these environmental differences would cause error in the resultant classification. Each image scene was classified into 20 classes using an Isodata unsupervised classification. The convergence threshold was set to 0.95 and the maximum number of iteration was set to 20. All classifications performed reached the convergence threshold before the maximum number of iteration was reached. The resultant 20 classes are then reviewed by the interpreter. Ancillary information such as the 2002 SAV mapping data, the collected Virtual Video Field Reference Database, georeferenced point images, and the 2002 aerial orthophotography are used to guide the naming of each class into one of the three categories high density seagrass, low density seagrass, or sand. Process_Date: May to July 2005 Cloud_Cover: <10 percent Spatial_Data_Organization_Information: Direct_Spatial_Reference_Method: Raster Raster_Object_Information: Raster_Object_Type: Pixel Row_Count: 7574 Column_Count: 19427 Spatial_Reference_Information: Horizontal_Coordinate_System_Definition: Planar: Grid_Coordinate_System: Grid_Coordinate_System_Name: Universal Transverse Mercator Universal_Transverse_Mercator: UTM_Zone_Number: 18 Transverse_Mercator: Scale_Factor_at_Central_Meridian: 0.999600 Longitude_of_Central_Meridian: -75.000000 Latitude_of_Projection_Origin: 0.000000 False_Easting: 500000.000000 False_Northing: 0.000000 Planar_Coordinate_Information: Planar_Coordinate_Encoding_Method: row and column Coordinate_Representation: Abscissa_Resolution: 2.400000 Ordinate_Resolution: 2.400000 Planar_Distance_Units: meters Geodetic_Model: Horizontal_Datum_Name: North American Datum of 1983 Ellipsoid_Name: Geodetic Reference System 80 Semi-major_Axis: 6378137.000000 Denominator_of_Flattening_Ratio: 298.257222 Entity_and_Attribute_Information: Detailed_Description: Entity_Type: Entity_Type_Label: Layer_1 Attribute: Attribute_Label: ObjectID Attribute_Definition: Internal feature number. Attribute_Definition_Source: ESRI Attribute_Domain_Values: Unrepresentable_Domain: Sequential unique whole numbers that are automatically generated. Attribute: Attribute_Label: Value Attribute_Definition: Set of numbers representing the seagrass coverage for each pixel Attribute_Domain_Values: Enumerated_Domain: Enumerated_Domain_Value: 1 Enumerated_Domain_Value_Definition: number given for High Density Seagrass Enumerated_Domain: Enumerated_Domain_Value: 2 Enumerated_Domain_Value_Definition: Number given for low density Seagrass Enumerated_Domain: Enumerated_Domain_Value: 3 Enumerated_Domain_Value_Definition: Number given for Sand or Unvegetated bottom Attribute: Attribute_Label: Count Attribute_Definition: number of pixels within each value Attribute: Attribute_Label: Red Attribute_Definition: color map red value Attribute: Attribute_Label: Green Attribute_Definition: color map green value Attribute: Attribute_Label: Blue Attribute_Definition: color map blue value Attribute: Attribute_Label: Class_names Attribute_Definition: Names given to each class Attribute_Domain_Values: Enumerated_Domain: Enumerated_Domain_Value: HDS Enumerated_Domain_Value_Definition: High density Seagrass >50% cover Enumerated_Domain: Enumerated_Domain_Value: LDS Enumerated_Domain_Value_Definition: Low Density Seagrass < 50% cover to > 10% cover Enumerated_Domain: Enumerated_Domain_Value: SND Enumerated_Domain_Value_Definition: Sand < 10% cover Distribution_Information: Distributor: Contact_Information: Contact_Person_Primary: Contact_Person: Sara Stevens Contact_Organization: National Park Service Contact_Address: Address_Type: mailing and physical address Address: National Park Service Address: University of Rhode Island Address: Coastal Institute Building Address: Rm 105 City: Kingston State_or_Province: RI Postal_Code: 02881 Country: USA Contact_Voice_Telephone: 407 874-4603 Contact_Electronic_Mail_Address: stevens@nps.gov Resource_Description: Downloadable Data Standard_Order_Process: Digital_Form: Digital_Transfer_Information: Transfer_Size: 0.000 Fees: None Ordering_Instructions: Please call or e-mail Available_Time_Period: Time_Period_Information: Single_Date/Time: Calendar_Date: unknown Metadata_Reference_Information: Metadata_Date: 20060226 Metadata_Contact: Contact_Information: Contact_Person_Primary: Contact_Person: YQ Wang Contact_Organization: University of Rhode Island Department of Natural Resources Science Contact_Position: Professor Contact_Address: Address_Type: mailing and physical address Address: University of Rhode Island Address: Coastal Institute Building Address: Department of Natural Resources Science City: Kingston State_or_Province: RI Postal_Code: 02881 Country: USA Contact_Voice_Telephone: 401 874-4345 Contact_Electronic_Mail_Address: yqwang@uri.edu Metadata_Standard_Name: FGDC Content Standards for Digital Geospatial Metadata Metadata_Standard_Version: FGDC-STD-001-1998 Metadata_Time_Convention: local time Metadata_Access_Constraints: none Metadata_Use_Constraints: none Metadata_Extensions: Online_Linkage: http://www.esri.com/metadata/esriprof80.html Profile_Name: ESRI Metadata Profile