Quantifying Natural CO2 Emissions from Mofettes at the Starzach Site in Southwest Germany with a Low-Cost Sensor Network

The natural carbon dioxide (CO2) degassings at the Starzach site near the Black Forest in southwestern Germany are recovering from a century of industrial exploitation. CO2 of presumably non-volcanic magmatic origin is released into the atmosphere via a cluster of mofettes of different sizes or diffusively through the soil, in patterns following a geological fault line running from southwest to northeast. Since the end of the industrial mining period in 1995, no comprehensive estimation of the total degassing rate at the site is available. This work presents three independent quantification approaches for the core Starzach mofette area (20 m × 20 m) with a focus on cost-effectiveness, each addressing a specific spatial scale: Degassing quantification of the most active mofette via volumetric flow rate measurement (465 kg/d ±16%, roughly extrapolated to other mofettes in the core Starzach mofette area: ∼1500 kg/d), near-surface flux-gradient method validated by eddy covariance (EC) measurements (max. 25 mg/m²/s, areal extrapolation: ∼900 kg/d), and a custom divergence-theorem based method deriving CO2 transport vectors from cross- correlations between atmospheric point measurements across the site (3266 kg/d ±42%), which is considered the most representative of the three due to its spatial coverage. Based on this, it is suggested that including other visible mofettes and abandoned boreholes within around 100 m distance brings the total CO2 emission of the Starzach site to an order of magnitude of roughly 10 t/d. All techniques utilized low-cost equipment, with CO2 sensors priced well below 100 € per piece, and were embedded in a wireless sensor network that enabled real-time remote monitoring. Low-cost equipment proved to be sufficiently accurate for emission quantification at the Starzach site. The difference between the results of the three employed methods demonstrates the importance of a hierarchical measurement approach covering different spatial scales as suggested by Sauer et al. (2013) because cross-validation becomes possible.