Remote sensing of dissolved gases in resource-rich seawater

Japan is almost completely dependent on imports for mineral and energy resources. Some minerals—e.g., sulfides—are thought to be abundant around Japan in hydrothermal vents that surround the island nation and as methane hydrates in seafloor sediments. As Japan's maritime territories are 12 times larger than its land area, the ocean represents a significant source of important minerals and gases. In addition, the Japanese government has been testing carbon dioxide capture and storage (CCS) for practical use. This technology captures carbon dioxide (CO₂) released from large-scale emission sources, such as power stations and factories, and stores it in sediments. Before developments of the seafloor commence, however, the marine environment should be explored and monitored.

Lidar (light detection and ranging) technology can provide 3D imaging of a surface by measuring the round-trip time between the emission of a laser pulse and the detection of its reflected signal. Lidar is a promising technique for monitoring the seafloor. It offers a number of advantages over commonly used point sensors and is considered to be one of the best candidates for monitoring gases dissolved in water. Hydrothermal fluids emanating from seafloor vents contain gases such as hydrogen sulfide (H₂S) and hydrogen (H₂), whereas methane hydrate concentrates contain methane (CH₄) and CCS stores CO₂; thus, lidar is a suitable technique for monitoring undeveloped seafloor regions.

CO₂ gas is usually detected using IR absorption methods.^{1, 2} However, water is a strong light absorber and has relatively high transmission in the short-wavelength spectral region (∼10–510nm) only. Thus, conventional IR absorption spectroscopy is not suitable for sensing gases dissolved in water. Instead, we propose using a Raman (inelastic scattering) lidar instrument with a green laser that has relatively high transmission in water.³ This technique involves irradiating the water with a 532nm laser beam and subsequently detecting the elastic (Mie and Rayleigh) scatterings at 532nm, as well as the Raman signals from the water. Figure 1 shows that, when CO₂ gas is dissolved in water, its Raman signals are also detected.

Figure 1. Raman spectra of carbon dioxide (CO₂) dissolved in water (H₂O) at concentrations of 71.4 and 818mmol/kg. The broad peak at ∼583nm originates from the bending of the H-O-H bonds. The sharp peaks at 571 and 574nm are CO₂Raman signals.

We used a standard Q-switched Nd:YAG (neodymium-doped yttrium aluminum garnet) laser operating at 532nm, with a pulse energy of 100mJ, pulse width of 10ns, and pulse repetition rate of 10Hz. These specifications are similar to those of lasers used in conventional lidar applications. For the measurements shown in Figure 1, the samples of water with dissolved CO₂were prepared in a high-pressure chamber whereby the water temperature and pressure were monitored. We calculated the concentration of CO₂ in the water using Henry's law (i.e., at constant temperature the amount of CO₂ dissolved in the volume of water is proportional to the partial pressure of CO₂ in equilibrium with water). The strong CO₂ Raman signal at 574nm (see Figure 1) is useful for submarine lidar applications. We used the broad signal at 583nm from the water (see Figure 1) to calibrate the Raman intensity and calculate the CO₂ concentrations.

We conducted laboratory experiments to test our lidar detection method. Commercial carbonated water with a known CO₂ concentration (∼152mmol/kg) and deionized water in the same glass bottle were placed 20m from the signal detector. Backscattered light was collected by a telescope (20cm in diameter) on the receiver side. The Raman signals were dispersed using a monochromator grating and were detected by a photomultiplier. We tuned the monochromator with 0.16nm steps and 500 lidar shots were averaged at each wavelength. Figure 2 shows 3D representations of the resulting Raman signals from deionized water and carbonated water, with a range resolution of 1.5m. The Raman signals of the bottles have separate features in the region of the CO₂ spectrum, and can be clearly distinguished (see Figure 2). These experiments demonstrate our ability to identify the CO₂ dissolved in water from a distance of 20m.

Figure 2. (a) Schematic diagram of the Raman lidar measurement system. Raman lidar CO₂ signals from (b) deionized water and (c) sparkling carbonated water in 3D representation. Raman signals caused by the bottles used in the experiment have separate features in the region of the CO₂ and are circled in red. PMT: Photomultiplier tube.

We are now developing a Raman lidar for the remote identification of gases dissolved in water. This work is yielding interesting insights that will be useful for applications in ocean exploration and monitoring. Our system can also simultaneously explore hydrothermal vents and methane hydrate, as well as monitor CH₄ leakage during methane-hydrate drilling. We now plan to detect gases dissolved in water by direct monitoring from a patrol ship in shallow waters around Japan. We are also considering the use of optical-fiber monitoring from a ship or using a system mounted to a deep-sea vehicle for deep-water measurements.

This work was supported by Kansai Electric Power Co., Inc.