Collar Properties and Measurement Time Confer Minimal Bias Overall on Annual Soil Respiration Estimates in a Global Database

Measuring the soil‐to‐atmosphere carbon dioxide (CO2) flux (soil respiration, RS) is important to understanding terrestrial carbon balance and to forecasting climate change. Such measurements are frequently made using measurement collars permanently inserted into the soil surface. However, differences in measurement duration and frequency, as well as collar properties, may lead to biases in the estimation of annual RS. Using a newly updated global RS database (SRDB‐V5), we investigated the annual RS bias associated with five methodological factors: collar height, collar coverage area, collar insertion depth, measurement duration, and measurement frequency. We found that annual RS was negatively correlated with collar insertion depth, consistent with the idea that collar insertion cuts roots and thus reduces RS. Annual RS was also negatively related with collar height and collar coverage area, perhaps because uniform head‐space mixing is difficult to achieve in larger volume chambers; however, these effects were quantitatively small (bias of ~2% to 10% of mean RS). We found no correlation of measurement duration or measurement frequency with annual RS. These findings suggest that variation in RS methodology generally introduces minimal bias overall. Therefore, compilations of minimally adjusted annual RS measurements provide a reliable resource for synthesis studies, global annual RS modeling, and investigation of how soil carbon responds to climate change. Plain Language Summary Soil‐to‐atmosphere carbon dioxide (CO2) is the second largest component in the terrestrial carbon cycle; thus, our ability to balance the terrestrial carbon budget and forecast climate change relies upon accurate measurements of this process. Collars permanently installed in the soil are commonly used to measure soil‐to‐atmosphere CO2. However, differences in collar properties, measurement duration, and measurement frequency may lead to biases in the estimation of annual soil‐to‐atmosphere CO2 amount. While many studies on the methodology for measuring soil‐to‐atmosphere CO2 have been conducted, a comprehensive evaluation of the influence of collar properties and measurement duration on annual soil‐to‐atmosphere CO2 variability has not been investigated before. In this study, we use a global data set to analyze soil‐to‐atmosphere CO2 measurement bias related to collar properties and measurement duration. We found that annual soil‐to‐atmosphere CO2 amount negatively correlated with