Spectral variability of oil slicks under different observing conditions derived from satellite and airborne optical remote sensing
No. of Downloads: 24
No. of Files: 56
File Size: 2.08 GB
File Format(s):
xlsx, hdr, nc, png, hdf, xml, txt
Funded By:
Gulf of Mexico Research Initiative
Funding Cycle:
RFP-IV
Research Group:
Center for the Integrated Modeling and Analysis of Gulf Ecosystems II (C-IMAGE II)
Chuanmin Hu
University of South Florida / College of Marine Science
huc@usf.edu
Oil spill, Optical remote sensing, Oil thickness, Oil emulsion, hyperspectral, multispectral, MODIS, MERIS, Landsat 7, MISR, AVIRIS, Rayleigh-corrected reflectance (Rrc), sun glint strength (LGN)
Abstract:
In this dataset, we present the spectral variability of oil slicks under different observing conditions using MODIS (Moderate Resolution Imaging Spectroradiometer), MERIS (Medium Resolution Imaging Spectrometer), MISR (Multi-angle Imaging SpectroRadiometer), Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and AVIRIS (Airborne Visible/ Infrared Imaging Spectrometer). Optical remote sensing is commonly used to detect oil in the surface ocean due to the spectral differences between oil and water, allowing to modulate oil–water spatial and spectral contrasts. However, understanding these contrasts is challenging because of variable results from laboratory and field experiments, as well as different observing conditions and spatial/spectral resolutions of remote sensing imagery. A multistep scheme is proposed to classify oil type (emulsion and non-emulsion) and to estimate relative oil thickness for each type based on the known optical properties of oil, with sample results from AVIRIS and MODIS imagery provided in the dataset. This dataset supports the publication: Sun, S., & Hu, C. (2018). The Challenges of Interpreting Oil-Water Spatial and Spectral Contrasts for the Estimation of Oil Thickness: Examples From Satellite and Airborne Measurements of the Deepwater Horizon Oil Spill. IEEE Transactions on Geoscience and Remote Sensing, 1–16. doi:10.1109/tgrs.2018.2876091
Suggested Citation:
Shaojie Sun; Chuanmin Hu. 2019. Spectral variability of oil slicks under different observing conditions derived from satellite and airborne optical remote sensing. Distributed by: GRIIDC, Harte Research Institute, Texas A&M University–Corpus Christi. doi:10.7266/n7-1nhg-ez10
Purpose:
The dataset was generated to demonstrate oil slick spectral variability under different observing conditions in multispectral and hyperspectral remote sensing imagery, and to figure out parameters other than oil thickness that contribute to the change of oil slick reflectance in remote sensing imagery in the real marine environment. The information and results derived in this study will contribute to the development of algorithms to estimate oil thickness by using multispectral satellite remote sensing imagery.
Data Parameters and Units:
MODIS: Four Moderate Resolution Imaging Spectroradiometer (MODIS; onboard Aqua and/or Terra) Rayleigh-corrected reflectance (Rrc, units=dimensionless) imagery in HDF format processed with the software SeaDAS (version 7.0) at a 250 m spatial resolution and mapped to a cylindrical equidistant projection [bounding box (North, South, West, East; decimal degrees) = 31, 26, -91, -84] for April 26, 2010 (Aqua), May 17, 2010 (Aqua and Terra) and May 28, 2010 (Terra). HDF files contain the Rrc data at the following wavelengths (nm): 412, 443, 469, 488, 531, 547, 555, 645, 667, 678, 748, 859, 869, 1240, 1640, 2130; the viewing geometry [solar zenith (solz), sensor zenith (senz), solar azimuth (sola), and sensor azimuth (sena)]; and glint coefficient (glint_coef). An RGB true-color image was created for each MODIS image. The Rrc data was extracted at selected locations and saved in excel files (latitude (decimal degrees), longitude (decimal degrees), wavelength (nm), Rrc, and Rrc difference). The Rrc difference is the difference between the selected location (with oil) and a nearby pixel (to represent the water reflectance and a non-oil pixel). MISR: Nine Multi-angle Imaging Spectroradiometer (MISR) Top-of-Atmosphere radiance (TOA) data HDF files from May 17, 2010 in the Mississippi Delta region. The filenames are MISR_AM1_GRP_TERRAIN_LM_Pmmm_Onnnnnn_cc_SITE_MISSDELTA_Fff_vvvv .hdf, where MISR_AM1_GRP_TERRAIN mean Level1B2 geo-rectified radiances, LM means the ellipsoid-projected local mode, Pmmm corresponds to the orbit path number, Onnnnnn is the absolute orbit number, cc is the camera identifier, ff is the file format version and vvvv is the version number (which relates to the reprocessing of a dataset with different software and/or ancillary inputs). The parameters defined to carry a Radiometric Data Quality Indicator (RDQI) associated with each measurement. The files contain geometric parameters which measure the sun and view angles at the reference ellipsoid; and Bidirectional Reflectance Factor (BRF) conversion factor for each band. Each file has 4 spectral bands: blue (446 nm), green (558 nm), red (672 nm) and near-infrared (NIR, 867 nm); and 9 viewing angles: 0, 26.1, 45.6, 60, 70.5, -26.1, -45.6, -60, -70.5 degrees. Each file has a corresponding XML file with metadata. The radiance data were extracted at selected locations (MISR_05_17_2010.xlsx) with emulsion, thin oil, thick oil and water for all angles and spectral bands. Radiance was also extracted at a transect for only the blue and NIR spectral bands and all angles. MERIS: Three Medium Resolution Imaging Spectrometer (MERIS) Rrc data files for April 25, 2010, April 26, 2010, and April 28, 2010 in HDF format processed with the software SeaDAS (version 7.0) at a 250 m spatial resolution and mapped to a cylindrical equidistant projection [bounding box (North, South, West, East; decimal degrees) = 31, 26, -91, -84]. The HDF files contain the Rrc data at the following wavelengths (nm): 413, 443, 490, 510, 560, 620, 665, 681, 709, 754, 762, 779, 865, and 885; the viewing geometry [solar zenith (solz), sensor zenith (senz), solar azimuth (sola), and sensor azimuth (sena)]; and glint coefficient (glint_coef), relative azimuth angle (phi), ozone, wind and pressure. An RGB true-color image was created for each MERIS file. The Rrc data was extracted at selected locations and saved in excel files (latitude (decimal degrees), longitude (decimal degrees), site name, wavelength (nm), Rrc, and Rrc difference). ETM+: One Landsat 7 Enhanced Thematic Mapper Plus (ETM+) Rrc data file in netCDF format processed with the software ACOLITE (V20170718.0) at a 30 m spatial resolution. The netCDF file contains chlorophyll-a derived with the OC3 algorithm (CHL-OC3), diffuse attenuation at 490 nm derived with the QAA algorithm (KD490), quality control flags, Scattering Line Height (SLH), latitude (decimal degrees), longitude (decimal degrees), the Rrc data at the following wavelengths (nm): 479, 561, 661, 835, 1650, and 2209. An RGB true-color image was created for the Landsat file. The Rrc data was extracted at selected locations and saved in excel files (latitude (decimal degrees), longitude (decimal degrees), site name, wavelength (nm), Rrc, and Rrc difference). AVIRIS_Classification_Result: The original AVIRIS file can be accessed openly via https://pubs.usgs.gov/of/2010/1167/downloads/figure16c-geotiff.tif. The reflectance spectra were extracted at selected locations and saved in AVIRIS_05_17_2010.xlsx (latitude (decimal degrees), longitude (decimal degrees), site name, wavelength (nm), Reflectance, and Reflectance difference). The emulsion_relative_thickness and non_emulsion_relative_thickness are the resulted relative thickness of oil emulsion and non-emulsions following the proposed classification scheme. The AVIRIS_scatter_plot.xlsx file contains the comparison of the relative thickness of oil emulsions from the multiband classification scheme with the USGS derived oil emulsion thickness using a hyperspectral algorithm. The AVIRIS_Classification_Result.xlsx includes data of mean and standard deviation spectra of classified pixels, using both classification approaches. MODIS_Classification_Result: This folder includes the results of MODIS relative thickness of oil emulsion following the same classification scheme and an AVIRIS derived emulsion thickness derived by USGS after binning to 250-m spatial resolution. LGN_vs_Rrc_Difference.xlsx: contains id, MODIS_filename, year, month, day, latitude (decimal degrees), longitude (decimal degrees), oil-water Rrc difference (i.e., Rrc difference between oil pixels and nearby oil-free pixels in the MODIS 859 nm band), sun glint strength (LGN) calculated from >300 natural slicks in the Gulf of Mexico using MODIS data. LGN was estimated using Cox and Munk (1954) model. The values represent the mean of the extracted oil slick (a polygon).
Methods:
AVIRIS and MODIS Classification: The classification scheme is a step-wise model to classify oil type (emulsion status) and classify relative thickness of both oil emulsions and non-emulsions using multiband satellite data; The USGS method using a hyperspectral approach to quantitatively map oil emulsions using hyperspectral AVIRIS data. Please refer to Sun and Hu, 2018 for further details of the classification scheme and its comparison with the USGS approach.
Provenance and Historical References:
Cox, C., & Munk, W. (1954). Measurement of the Roughness of the Sea Surface from Photographs of the Sun’s Glitter. Journal of the Optical Society of America, 44(11), 838. doi:10.1364/josa.44.000838