Abstract Accurately scaling soil respiration (SR, the soil-to-atmosphere flow of CO2) across time and space is important to constrain and understand ecosystem to global scale SR, the largest terrestrial carbon flux to the atmosphere. Year-round SR measurements are however expensive and sometimes impossible to perform. Bahn et al. (2010) developed an approach to estimate annual SR (SRannual) from the flux measured at mean soil temperature (SRMAST), but the robustness of this approach needs to be evaluated in diverse ecosystem types and climatic conditions globally. We used a global soil respiration database (SRDB-V4, with 823 SR observations worldwide) to test the capability of SRMAST to predict SRannual. SRMAST estimated using a variety of methods all showed clear relationships with annual SRannual. Two single-rate methods (i.e., using the single SR rate closest to mean annual soil temperature, or the single SR rate closest to mean annual air temperature) showed the most pronounced divergence from the true SRannual, but errors significantly decreased when using multiple SR rates within 1 ℃ of the mean annual soil temperature to estimate SRMAST. SRannual was most closely correlated with SRMAST estimated via a Q10 relationship, but this method has a potential autocorrelation issue that we explore and discuss. Air temperature data are much more widely available than is soil temperature, and we found that SR at mean annual air temperature (SRMAAT) can be used to predict SRannual as well. This study builds on Bahn et al. (2010) to demonstrate that SR measured at both annual mean soil and air temperature can be used to predict annual SR, with well-quantified errors. This capability could be used to reduce SR measurement frequency required for estimating SRannual and greatly decrease cost, factors that are generally important but especially in lower-income countries and cold, inaccessible regions.

中文翻译：

---

年平均温度下土壤年呼吸通量的预测

摘要 跨时间和空间准确缩放土壤呼吸（SR，CO2 的土壤到大气的流动）对于将生态系统限制和理解为全球尺度的 SR 是重要的，SR 是最大的陆地碳通量。然而，全年的 SR 测量非常昂贵，有时甚至无法执行。班等人。(2010) 开发了一种从平均土壤温度 (SRMAST) 下测量的通量估算年度 SR (SRannual) 的方法，但需要在全球不同的生态系统类型和气候条件下评估该方法的稳健性。我们使用全球土壤呼吸数据库（SRDB-V4，在全球有 823 个 SR 观测值）来测试 SRMAST 预测 SRannual 的能力。使用各种方法估计的 SRMAST 都显示出与年度 SRannual 的明确关系。两种单速率方法（即 使用最接近年平均土壤温度的单一 SR 率，或最接近年平均气温的单一 SR 率）显示与真实 SRannual 的最显着差异，但在年平均 1 ℃ 内使用多个 SR 率时误差显着降低土壤温度来估计 SRMAST。SRannual 与通过 Q10 关系估计的 SRMAST 最密切相关，但这种方法具有我们探索和讨论的潜在自相关问题。气温数据比土壤温度更容易获得，我们发现年平均气温 (SRMAAT) 的 SR 也可用于预测 SRannual。这项研究建立在 Bahn 等人的基础上。(2010) 证明在年平均土壤和空气温度下测量的 SR 可用于预测年度 SR，并具有良好量化的误差。