Response time correction of membrane equilibrium based methane sensor data

High resolution measurements with acceptable accuracy are crucial to increase our understanding of important environmental processes. The sharp spatial and temporal gradients which characterize seabed seepage environments often means that conventional measuring techniques fall short in representing seepage, content, and the environmental processes of interest. This is especially the case when it comes to measuring dissolved methane, which is still often done using discrete water sampling and subsequent laboratory analysis. This practice is time consuming, resulting in data with poor spatiotemporal resolution which is often unable to represent highly variable environmental processes. The overt solution to this problem is the employment of in situ sensors. Unfortunately, the common sensor design approach in off-the-shelf sensors, where measurements rely on a membrane equilibrium extraction technique and gas detection by some embedded device, are, while theoretically reliable, often plagued by high response times – especially in cold environments, making them unsuitable for applications where rapid changes are expected.

We present a new, easily applicable, lab and field-tested method for recovering fast response data from off-the-shelf methane sensors relying on the principle of membrane separation by using the theoretical framework of statistical inverse theory. This framework allows us to model the uncertainty of the measurements obtained by the internal detector, giving fast response data where measurement uncertainty is explicitly defined – something which has not been possible in the past. The solution is constrained by model sparsity, which in practice gives the user data with the resolution at which the sensor is able to give measurements with a reasonable uncertainty. Furthermore, our method requires no additional input from the user other than what is provided from the manufacturer, such as detector accuracy and response time. Getting reliable fast response data from relatively affordable, off-the-shelf in situ sensors, means that these can be used in new applications such as profiling, towing, or on autonomous platforms such as gliders. This can considerably improve our ability to quantify dissolved methane and resolve and understand related environmental processes in the ocean.