2021 Flowerfield Creek, MI Natural Color, Color Infrared and Hyperspectral Imagery

Create custom data files by choosing area, map projection, file format, etc.

Simple download of data files.

Date that the source FGDC record was last modified.

Converted from FGDC CSDGM (version FGDC-STD-001-1998) using 'fgdc_to_inport_xml.pl' script. Contact Tyler Christensen (NOS) for details.

Partial upload of Positional Accuracy fields only.

Partial upload to move data access links to Distribution Info.

Airborne Hyperspectral Data Collection - An AISA KESTREL 10 Visual through Near Infrared (VNIR) Hyperspectral Imaging Sensor System (400 - 1,000 nm spectral range) was utilized by Galileo Group Inc. (Galileo) for airborne collection operations. A total of 178 spectral bands for the VNIR (Visible Near Infrared) range were collected with a spectral resolution of around 3.5 nm and a Ground Sampling Distance (GSD) of 1.0m. The FOV (Field of View) was 40 degrees. An Oxford Solutions Survey+ 2nd Generation GPS/IMU simultaneously collected navigation and attitude data. TerraStar-C from Novatel Inc. was utilized for all airborne collection operations to further enhance spatial accuracy. TerraStar is a Real Time Kinetic (RTK) airborne correction service that improves horizontal (~4-5 cm RMS) and vertical accuracy (~6.5 cm RMS). The sensor system was installed in a Cessna 172 Skyhawk which was used as the airborne imaging platform for the duration of the project.

Ground Truthing Data Collection - Ground truthing was conducted primarily onboard a canoe and along the shoreline where accessible. An ASD FieldSpec Pro VNIR (400-1000nm) was utilized to collect spectrometry of wild rice when found. A Trimble Pro XRT DGPS system and two handheld digital cameras with integrated GPS were used to collect GPS points and geo-tagged imagery of target species and landmarks of interest. A specially modified iPhone 6 equipped with a Galileo custom created ground truthing application (ARMADA) and linked with a sub-meter accuracy Trimble Bluetooth antenna was utilized as a secondary GPS point logging device.

Processing Level 1a Radiometric Correction 1. Dark Noise Removal- Dark image data was acquired for every raw image at the end of every flight line by the sensor control software closing the shutter and recording 3 seconds of dark noise. To remove sensor noise the mean value of every line of the dark data was subtracted from the corresponding line of the raw data (dark noise removal or dark current removal). 2. Calibration from RAW data to radiance data using calibration file- After the dark noise removal, the raw data was calibrated to radiance units using a sensor specific calibration file. Every spatial and spectral pixel is multiplied with the corresponding value in the calibration file. The calibration values for each pixel on the CCD are calculated using an integrating sphere in the laboratory. The radiance units are equal to (mW/cm^2*str*um)*1000.00. 3. Smile Correction- Spectral smile is defined as changes in wavelength over the field of view (FOV). Smile correction was performed by proprietary algorithms. 4. Crosstrack Correction- A crosstrack correction was performed using proprietary algorithms in order to normalize illumination in the across-track direction of each flight line. 5. Quality Control- Quality control was accomplished using ASD measurements, established radiometric quality protocols and systematic manual evaluations. Processing Level 1b Atmospheric Correction 1. Model-based Atmospheric Correction- The radiance data were converted to reflectance using the Atmospheric & Topographic Correction (ATCOR-4) software package. ATCOR uses the Modtran 4, a complex radiative transfer model, to correct for atmospheric absorption (O2, O3, and CO2), scattering (Mie and Rayleigh) and path radiance effects. The calibrated radiance data was then converted to reflectance pixel count (reflectance ranges 0.0 to 1.0 multiplied by 10000) using modeled solar irradiance and solar angle data from the time of acquisition. The reflectance pixel counts were then converted to 16-bit reflectance data. The reflectance units are equal to Reflectance*10000. 2. Quality Control- The quality control was accomplished using established atmospheric correction quality protocols and systematic manual evaluations. Ground based spectrometer data at impervious and spectrally intransient surfaces such as asphalt were utilized for rigorous QA/QC of the reflectance data. Spectral curves from the ground spectrometer data were compared to the final reflectance results and found to be sufficiently consistent. Further parameter adjustment was then applied to the data to maximize spectral consistency between the flight lines.

Processing Level 2 Geometric Correction 1. Calculation of the sensor offset (Boresight correction) - During the aerial acquisition, four special Boresight flight lines were flown to geometrically calibrate the sensor and GPS/INS. The Boresight parameters were calculated using four overlapping flight lines flown in a crosshair pattern. 15 to 20 Ground Control Points per flight line pair (GCPs) were identified and used to calculate geometric correction values for Roll (0.180527), Pitch (-0.0533453) and Yaw (0.00885310). These values were then used as input parameters for the geometric correction process. 2. GPS/INS Data- The GPS/INS Data was encoded and processed for the use in the georectification process. 3. GLT files (Geographic Lookup Table)- A GLT file contains the geographic location of every pixel in the unrectified reflectance data. A GLT file was generated for every flight line using the GPS/INS Data, the Boresight Correction parameters and a Digital Elevation Model (DEM). Processing Level 3 Orthorectification 1. Georectification using GLT data- The unrectified reflectance data was then Georectified into North American Datum of 1983 (NAD83) and projected into the Universal Transverse Mercator (UTM) coordinate system using the corresponding GLT file. A DEM was used to correct for surface elevation variation across the scene. 2. Mosaicking- Individual flight lines for each AOI were combined into a single mosaic image which covered the entire AOI. This image was then masked to the AOI boundary and tiled according to the tiling scheme vectors included with this delivery. 3. Quality Control- Geo-accuracy was checked and confirmed to meet pre-established quality standards by systematic comparison of specific geo-locations acquired in the field with a GPS unit and high resolution RGB imagery of known geo-accuracy.