Abstract:

Collection of LIDAR and Hyperspectral data was conducted in August 2015 over two polygons located west of Port Sulphur, Louisiana. The north polygon covered the Bay Jimmy, encompassing 250 Square Km. The south polygon covered Cat Island and West Barataria, encompassing 350 Square Km. The flights took place on August 16, 17, 18 and 24, 2015 (DOY 228, 229, 230 and 236).

Suggested Citation:

Shrestha, Ramesh. 2017. LIDAR and Hyperspectral Data Collection Over Barataria Bay, Louisiana, August 2015. Distributed by: GRIIDC, Harte Research Institute, Texas A&M University–Corpus Christi. doi:10.7266/N76T0JP4

Purpose:

LIDAR and Hyperspectral data collection as part of the Coastal Waters Consortium funded by GoMRI

Data Parameters and Units:

Horizontal Datum: NAD83 (2011); Vertical Datum: NAVD88 (GEOID 12a); Projection: UTM Zone 15N; Units: meters; File Formats: 1. Point Cloud in LAS format (version 1.2), unclassified in 1 km square tiles. 2. ESRI float format 1.0-m DEM from unclassified points. 3. CASI Hyperspectral images: Each flight strip as an individual image in PCI Geomatica “.pix” format, 1.2 m pixel resolution, 48/72 bands with a corresponding metadata file in XML format File naming convention for LIDAR point clouds: 1 Km tiles follow a naming convention using the lower left coordinate (minimum X, Y) as the seed for the file name as follows: CXXXXXX_YYYYYYY. For example if the tile bounds coordinate values from easting equals 382000 through 383000, and northing equals 4130000 through 4131000 then the tile filename incorporates 382000_4130000. The ESRI DEMs are mosaic files created by combining together the 1 km tiles to get a single mosaic file for each: bare earth and first return surface models. Bounding box for North Polygon covering Bay of Jimmy: 29.42385928 -90.00280738 29.41794629 -89.74014485 29.50810747 -89.74332507 29.51435169 -89.99888432 Bounding box for South Polygon covering Cat island and West Barataria: 29.20177771 -90.09312047 29.26746138 -89.94583211 29.31890033 -89.95774282 29.32835682 -89.89120583 29.34149389 -89.79961490 29.39871991 -89.81293845 29.38715405 -89.90439310 29.37872826 -89.97128855 29.38524878 -89.97234533 29.35383926 -90.14156255 For additional technical specifications, see the word document "LIDAR and Hyperspectral Data Collection Over Barataria Bay_GOMRI" supplied by the data collector with the data package.

Methods:

a) GPS/IMU Data Processing Reference coordinates (NAD83 (2011) Epoch 2010.0000) for all the stations are derived from observation sessions taken over the project duration and submitted to the NGS on-line processor OPUS which processes static differential baselines tied to the international CORS network. For further information on OPUS see http://www.ngs.noaa.gov/OPUS/ and for more information on the CORS network see http://www.ngs.noaa.gov/CORS/ Airplane trajectories for this survey were processed using KARS (Kinematic and Rapid Static) software written by Dr. Gerald Mader of the NGS Research Laboratory. KARS kinematic GPS processing uses the dual-frequency phase history files of the reference and airborne receivers to determine a high-accuracy fixed integer ionosphere-free differential solution at 1 Hz. All final aircraft trajectories for this project are blended solutions from at least two stations. After GPS processing, the 1 Hz trajectory solution and the 200 Hz raw inertial measurement unit (IMU) data collected during the flights are combined in APPLANIX software POSPac MMS (Mobile Mapping Suite Version 7). POSPac MMS implements a Kalman Filter algorithm to produce a final, smoothed, and complete navigation solution including both aircraft position and orientation at 200 Hz. This final navigation solution is known as an SBET (Smoothed Best Estimated Trajectory). b) LiDAR Data Processing overview: Classification was done by automated means using TerraSolid software (TerraScan Version 14.017). http://www.terrasolid.com/products/terrascanpage.php System calibration of the 3 sensor bore sight angles (roll, pitch, and yaw) and scanner mirror scale factor is done by automated means using LMSPro software provided by Optech. Overlapping parallel project lines along with perpendicular cross lines and lines over developed neighborhoods with many sloping roof lines are used as input into automated optimization and calibration routines. It uses least-squares algorithms to compute and apply optimal bore sight offsets and scale values that minimize height mismatches in overlapping flight lines. These routines are run and calibration values are updated for each flight. Before the start of the project, 1719 ground check points were collected near the Ellington airfield to calculate for systematic bias in the LIDAR data. After the removal of the bias, the RMS for the elevation differences was 0.03 m. NCALM makes every effort to produce the highest quality LiDAR data possible but every LiDAR point cloud and derived DEM will have visible artifacts if it is examined at a sufficiently fine level. Examples of such artifacts include visible swath edges, corduroy (visible scan lines), and data gaps. A detailed discussion on the causes of data artifacts and how to recognize them can be found here: http://ncalm.berkeley.edu/reports/GEM_Rep_2005_01_002.pdf . A discussion of the procedures NCALM uses to ensure data quality can be found here: http://ncalm.berkeley.edu/reports/NCALM_WhitePaper_v1.2.pdf NCALM cannot devote the required time to remove all artifacts from data sets, but if researchers find areas with artifacts that impact their applications they should contact NCALM and we will assist them in removing the artifacts to the extent possible – but this may well involve the PIs devoting additional time and resources to this process. c) Hyperspectral Imagery Imagery processing took place in three steps: a) Radiometric Correction; b) Calibration; and c) Orthorectification. The calibration is usually done on a set of images taken over a calibration site with perpendicular and opposing headings directions. Calibration involves using bundle adjustment to solve for linear and angular misalignments using tie point and ground control points. For this purpose the calibration site is located over an area with distinct road markings such as parking lots or an airport. Finally the calculated misalignment values are fed back to give the orthorectified images. CASI is a pushbroom scanner so most of the processing is done using software utilities developed in house by ITRES. Radiometric calibration is done using ‘RCXSAV’ utility which does the following: 1) Corrects for the additive, non-target related components to the raw signal, 2) Calibrates the raw measured DN value for each pixel by multiplying by coefficients generated during the calibration of the instrument. In doing so, the data are converted to units of Spectral Radiance. These coefficients are contained within the instrument’s calibration (*.rad) files, which are unique to each CASI and calibration run, 3) Scales data: from 14-bit to 16-bit values, 4) Outputs raw CASI image files in *.pix format which is a 16-bit unsigned integer, BIP format The next two steps of bundle adjustment and orthorectification are carried out by ‘PBSbund’ and ‘Geocor’ software utilities respectively provided by ITRES.

Instruments:

This survey was performed with an Optech Titan multispectral airborne LiDAR sensor (serial number 14SEN340) and an ITRES CASI-1500H hyperspectral camera mounted in a twin-engine Piper PA-31-350 Navajo Chieftain (Tail Number N154WW).

Error Analysis:

Before the start of the project, 1719 ground check points were collected near the Ellington airfield to calculate for systematic bias in the LIDAR data. After the removal of the bias, the RMS for the elevation differences was 0.03 m.

Individual Files: