TOTAL COLUMN METHANE RETRIEVALS USING THE TROPOSPHERIC
INFRARED MAPPING SPECTROMETER OVER SUNGLINT
N. Larsen ^ *, J. Kumer?, R. Rairden ”, K. Jablonski °
? Lockheed Martin Advanced Technology Laboratories, 4301 N Fairfax Drive, Arlington, VA, USA -
north.larsen@lmco.com
? Lockheed Martin Advanced Technology Center, Palo Alto, CA, USA - (jack kumer, rick.rairden)@lmco.com
* Lockheed Martin ISGS, Valley Forge, PA, USA - kathryn.l.jablonski(gmco.com
Commission VIII, WG VIII/3
KEY WORDS: Exploration, Oceans, Aerial, Hyper Spectral, Detection, Sensor, Algorithms
ABSTRACT:
Because it is a greenhouse gas, the detection of methane concentrations is a global issue. Additionally, the presence of methane is
indicative of potential valuable petroleum and natural gas deposits. Therefore methane seep detection is useful for petroleum
exploration around the world. The detection of methane, and other absorbing gases, over water is an issue for passive systems
because one is seeking to detect an absorbing gas over an absorbing surface. The solution to this dilemma is to use the sun/sensor
geometry for sun glint off of water to measure the absorbing gas over a reflecting surface, and therefore significantly increase the
signal to noise of the measurement being taken. In September of 2010 Lockheed Martin performed a proof of concept by
demonstrating from an airship over San Francisco Bay the capability of the Tropospheric Infrared Mapping Spectrometer’s (TIMS)
hyper spectral sensor to passively measure methane, CO, and water vapor over sunglint water. The Lockheed Martin prototype TIMS
sensor system is a hyper spectral grating spectrometer instrument that operates in the 2.3 micron spectral region at 0.25 cm’
resolution. The Lockheed Martin retrieval algorithm developed applies the KCARTA (kCompressed Atmospheric Radiative Transfer
Algorithm) with Jacobians, with the HITRAN 2008 lineshape parameters, to retrieve the total column amount of atmospheric species
along with the calibrated TIMS sensors radiometric input. A cell with known amount of methane was placed into the input to the
TIMS to simulate atmospheric enhancements near the water surface. The amount in the cell was retrieved well within the uncertainty
of 1% of the amount in the cell. Multi frame retrievals on data in which the cell was not placed into the input beam demonstrated 1%
precision. In addition, in situ surface measurements were done over a landfill park, where measurements of methane were taken over
known hotspots. This research allows for the future development of a system to measure methane, and other gases, for climate and
petroleum exploration capabilities.
1. INTRODUCTION 2. METHODS
The Lockheed Martin (LM) Tropospheric Infrared Mapping 2.1 Prior Research
Spectrometers (TIMS) sensor, (Kumer et al., 2008) currently
under development will detect surface enhancements (over the
atmospheric background) of methane over land, in addition to
the very challenging problem of its detection over water. In
addition to being a greenhouse gas of increasing relevance,
methane gas seeps are indicative of both subterranean
petroleum sources and oil/gas pipeline leaks. LM researchers
In 2008-2009 LM scientists in the Atmospheric Sensing
Petroleum and Climate Research (ASPECT) team researched
the detection of methane over sunglint using data from the
NASA JPL AVIRIS instrument to show sensitivities to methane
seeps of ~200 ppm off of the coast of Santa Barbara, CA where
well-known hydrocarbon sources exist (Figure 1). Simulation
analyzed Airborne Visible Infrared Imaging Spectrometer
(AVIRIS) instrument sunglint imagery of California, and
detected positive methane hotspots. These results led LM
scientists to demonstrate an innovative proof-of-concept
technique for detecting methane over water when sunglint is
present. This technique was performed with the TIMS
instrument from an airship over the southern waters of San
Francisco Bay. The ability to passively image methane, and
other absorbing gases, over the water is challenging due
primarily to the low Signal-to-Noise (SNR) resulting from a low
surface reflectance. By imaging sunglint regions the background
reflectance is high, which allows for methane retrievals to be
performed. TIMS retrievals were radiometrically calibrated and
a LM Methane Retrieval Algorithm developed.
* Corresponding author. This is useful to know for communication with the appropriate person in cases with more than one author.
and Modeling had shown that the detection of methane over
sunglint was theoretically possible (Larsen and Stamnes, 2006).
The AVIRIS imagery used was collected under less than ideal
conditions (very low reflectivity of 3-10%— virtually no sun-
glint) and for a less than ideal instrument (AVIRIS moderate
spectral resolution, compared to the TIMS sensor). Using the
AVIRIS Reflectance imagery data at the methane absorption
features located near 2.3 microns, the LM ASPECT team
developed a Normalized Differential Methane Index (NDMI) to
identify the methane seeps over the water, given by Equation 1.
R(2.3) - RQ.1) (1)
R(2.3)+ R(2.1)
NDMI =
