Correction to “Net Greenhouse Gas Balance in U.S. Croplands: How Can Soils Be Part of the Climate Solution?”

Abstract

You, Y., Tian, H., Pan, S., Shi, H., Lu, C., Batchelor, W. D., Cheng, B., Hui, D., Kicklighter, D., Liang, X.-Z., Li, X., Melillo, J., Pan, N., Prior, S. A., & Reilly, J. (2023). Net Greenhouse Gas Balance in U.S. Croplands: How Can Soils Be Part of the Climate Solution? Global Change Biology, 30, e17109. https://doi.org/10.1111/gcb.17109

In the article by You et al. (2023), we found an error in the unit of soil organic carbon (SOC) sequestration rate in Figure 2c. Specifically, the original unit was listed as g C/m²/year, but it should be g C/m²/experiment_duration. The data shown in Figure 2c represent the changes in SOC amount over the duration of the experimental period. In most of the papers we collected, the experimental duration for implementing specific agriculture practices (e.g., tillage) typically span multiple years (e.g., 5, 10 or 20 years). Consequently, these papers generally report changes in SOC amounts over the corresponding experimental period, rather than reporting an average annual SOC sequestration rate. Therefore, we compared our results (i.e., changes in SOC amount) with those papers over the same periods and experimental durations during model validation process.

To enhance comparability, we converted the original unit of observed SOC sequestration rates from g C/m²/experiment_duration to g C/m²/year and redrew Figure 2c. The main message of this article still upholds after correcting this mistake. The corrected version of Figure 2c is provided. Details of the experimental duration of these data points are also attached to this document (Table 1). Relevant changes and corrections related to Figure 2c made in the main text are also provided.

2.4 Model calibration, validation, and sensitivity analysis

Original text: Overall, DLEM can well-simulate emissions of N₂O and CH₄, SOC sequestration rates, and SOC stocks compared with field observations from the metadata collection, in which RMSE (NRMSE) values were 0.16 g N m⁻² (9.6%), 1.5 g C m⁻² (4.4%), 156.9 g C m⁻² (19.3%), and 1929.1 g C m⁻² (17.6%), respectively, and the R² values were 0.6, 0.91, 0.46, and 0.64, respectively (Figure 2).

Correct text: Overall, DLEM can well-simulate emissions of N₂O and CH₄, SOC sequestration rates, and SOC stocks compared with field observations from the metadata collection, in which RMSE (NRMSE) values were 0.16 g N m⁻² (9.6%), 1.5 g C m⁻² (4.4%), 17.04 g C m⁻² (9.17%), and 1929.1 g C m⁻² (17.6%), respectively, and the R² values were 0.6, 0.91, 0.44, and 0.64, respectively (Figure 2).

We apologize for this error.