4:30 PM - 4:45 PM
[MIS20-10] An attempt to correct response time of methane sensor by deconvolution
Keywords:methane, gas hydrate, methane sensor, deconvolution
Gas hydrates are widely distributed in continental margin and permafrost regions. They are often composed mainly of methane gas and also called “methane hydrate“. Methane hydrate is expected to be an unconventional energy resource, while its formation and dissociation involving methane, a greenhouse gas, are also attracting attention. There are two types of ocean gas hydrates, the massive nodular type of hydrate in shallow sediments (shallow type) and pore-filling type hydrate in sandy horizon, and they have been identified around Japan in the Sea of Japan, Hokkaido, and the Pacific Ocean, respectively.
The concentration of dissolved methane in seawater in methane hydrate distribution areas is considered to be related to the methane flux from beneath the seafloor and the distribution of methane hydrate beneath the seafloor. Therefore, compared to the general marine environment, the concentration of dissolved methane may be observed to be higher in the seawater where shallow type methane hydrate is present. The distribution of dissolved methane concentrations is important information for understanding environmental effects and formation/dissociation processes in methane hydrate distribution areas. Although the concentration of dissolved methane has been measured in the past, there has been a problem of dissociation in response due to the characteristics of the sensor. In other words, when methane sensors are installed on ROV (Remotely Operated Vehicle) and AUV (Autonomous Underwater Vehicle), the response delay caused by movement is different from that of in-situ concentrations. Recently, a dissolved methane sensor with an extremely small response delay has been developed (Grilli et al., 2018). In this study, we assumed that this sensor has no response delay (extremely short response time) and tested whether the response delay of a conventional dissolved methane sensors can be corrected using a deconvolution technique. See Dølven et al. (2022) for details of the method.
This study was conducted offshore of Niigata Prefecture, where methane hydrate is known to be present and where gas plumes from the seafloor have been observed. Two types of methane sensors were used: LMS (Franatech GmbH) and Sub-Ocean (A2 Photonic Sensors), which is assumed to have no response delay. The ROV observed the seafloor surface along the planned survey line while maintaining an altitude of about 4-5 m from the seafloor and a speed of about 0.2-0.5 knots. Niskin water samplers were installed on the ROV to collect seawater samples as needed during the submergence to compare methane concentrations. After the submarine survey, seawater was transferred from the sampler to a glass vial, added to a benzalkonium chloride solution for sterilization, refrigerated, and used to measure dissolved methane concentrations.
The observed methane concentrations were commonly high regardless of the sensor type. The peak value of methane concentration differed among sensors, with Sub-Ocean observing the highest value. On the other hand, the LMS sensor tended to show a gradual decrease in value after the peak, suggesting that the response time of the detector was longer than that of the Sub-Ocean sensor. Although the response time could be corrected by deconvolution, its accuracy (agreement with Sub-Ocean) depended on the submergence of the ROV. It is difficult to correct by deconvolution when the change in dissolved methane concentration is small (< 300 ppm) or when the change in concentration is large (> 2000 ppm). This may be due to the fact that the difference between LMS and Sub-Ocean observations is small when the time series variations of concentration is small. In the future, we will aim to develop a method that can be applied to the above conditions, and also to establish a method that can reproduce concentration changes even in submarines that are not equipped with a sensor with a fast response time.
This study was conducted as a part of the methane hydrate research project funded by METI (the Ministry of Economy, Trade and Industry, Japan).
The concentration of dissolved methane in seawater in methane hydrate distribution areas is considered to be related to the methane flux from beneath the seafloor and the distribution of methane hydrate beneath the seafloor. Therefore, compared to the general marine environment, the concentration of dissolved methane may be observed to be higher in the seawater where shallow type methane hydrate is present. The distribution of dissolved methane concentrations is important information for understanding environmental effects and formation/dissociation processes in methane hydrate distribution areas. Although the concentration of dissolved methane has been measured in the past, there has been a problem of dissociation in response due to the characteristics of the sensor. In other words, when methane sensors are installed on ROV (Remotely Operated Vehicle) and AUV (Autonomous Underwater Vehicle), the response delay caused by movement is different from that of in-situ concentrations. Recently, a dissolved methane sensor with an extremely small response delay has been developed (Grilli et al., 2018). In this study, we assumed that this sensor has no response delay (extremely short response time) and tested whether the response delay of a conventional dissolved methane sensors can be corrected using a deconvolution technique. See Dølven et al. (2022) for details of the method.
This study was conducted offshore of Niigata Prefecture, where methane hydrate is known to be present and where gas plumes from the seafloor have been observed. Two types of methane sensors were used: LMS (Franatech GmbH) and Sub-Ocean (A2 Photonic Sensors), which is assumed to have no response delay. The ROV observed the seafloor surface along the planned survey line while maintaining an altitude of about 4-5 m from the seafloor and a speed of about 0.2-0.5 knots. Niskin water samplers were installed on the ROV to collect seawater samples as needed during the submergence to compare methane concentrations. After the submarine survey, seawater was transferred from the sampler to a glass vial, added to a benzalkonium chloride solution for sterilization, refrigerated, and used to measure dissolved methane concentrations.
The observed methane concentrations were commonly high regardless of the sensor type. The peak value of methane concentration differed among sensors, with Sub-Ocean observing the highest value. On the other hand, the LMS sensor tended to show a gradual decrease in value after the peak, suggesting that the response time of the detector was longer than that of the Sub-Ocean sensor. Although the response time could be corrected by deconvolution, its accuracy (agreement with Sub-Ocean) depended on the submergence of the ROV. It is difficult to correct by deconvolution when the change in dissolved methane concentration is small (< 300 ppm) or when the change in concentration is large (> 2000 ppm). This may be due to the fact that the difference between LMS and Sub-Ocean observations is small when the time series variations of concentration is small. In the future, we will aim to develop a method that can be applied to the above conditions, and also to establish a method that can reproduce concentration changes even in submarines that are not equipped with a sensor with a fast response time.
This study was conducted as a part of the methane hydrate research project funded by METI (the Ministry of Economy, Trade and Industry, Japan).