The IceWorm: an improved low-cost, low-power sensor for measuring dissolved CH4 in water bodies

Recent studies show emissions of dissolved methane (CH₄) in the meltwater from the Greenland Ice Sheet. To better understand the phenomenon and evaluate its potential significance for the Arctic CH₄ budget, continuous long-term measurements of dissolved CH₄ concentrations are needed. Commercially available dissolved CH₄ analyzers (DGEU-UGGA (LGR), CONTROS HydroC CH₄ (Kongsberg) and Mini CH₄ (pro-Oceanus)) generally have high power consumption and are very costly, limiting their operation in remote off-grid locations.

Here we present calibrations and field tests of a low-cost, low-power alternative – the "IceWorm" - for long-term monitoring of dissolved CH₄. The IceWorm uses a Figaro TGS2611-E00 metal oxide sensor (MOS). While MOS are cheap and power efficient, a known drawback is the sensitivity of the sensor's resistance to changes in humidity and temperature. In a previous prototype, we showed that by encasing the MOS in a hydrophobic and gas-permeable silicone membrane, a constant humidity in the headspace around the sensor can be achieved, yielding consistent results when deployed in glacial meltwater at constant temperature (0.0 – 0.1˚C)¹. In this updated version, the sensor was encased in a hydrophobic and gas-permeable Teflon membrane allowing for fast (~1 min) equilibrium between the water and headspace around the sensor and hence a rapid detection of changes in dissolved CH₄ concentrations.

The first calibration was performed by exposing the IceWorm to stepwise increasing Two field calibrations of the sensor performance in meltwater at 0.0˚C were done: Afterwards, the sensors remained in the field for several weeks in the subglacial meltwater stream and the sensors were recalibrated in lab air under the same conditions to check for long-term sensor drift. Initially, field calibrated to measure dissolved CH₄ in glacial meltwater at 0.0˚C, the IceWorm was also tested in a freshwater surface stream at temperatures between 1.6 – 15.7˚C. To account for the temperature difference, we compared the laboratory and field calibrations allowing us to correct the sensor output to temperature variations in the stream.

We will present time series of long-term measurements of dissolved CH₄ in two different types of water bodies and discuss the promising performance of the sensor at temperatures different to stable 0˚C as well as the usability of in-air calibrations compared to the field calibrations with discrete samples.

1. Sapper et al. (2022) DOI:10.5194/egusphere-egu22-9972