Simulating measurable ecosystem carbon and nitrogen dynamics with the mechanistically defined MEMS 2.0 model
Simulating measurable ecosystem carbon and nitrogen dynamics with the mechanistically defined MEMS 2.0 model
复制标题
DOI：
10.5194/bg-18-3147-2021
复制
发表时间：
2021-05
期刊：
影响因子：
4.9
通讯作者：
Yao Zhang;J. Lavallee;A. Robertson;Rebecca Even;S. Ogle;K. Paustian;M. Cotrufo
中科院分区：
文献类型：
--
作者：
Yao Zhang;J. Lavallee;A. Robertson;Rebecca Even;S. Ogle;K. Paustian;M. Cotrufo
摘要。几十年来，主流的土壤生物地球化学模型使用概念土壤有机质（SOM）库，并且只在土壤中较浅的深度模拟它们。克服这些限制的努力推动了新一代SOM模型的发展，包括MEMS 1.0，它代表了整个根区可测量的生物物理SOM分数，并体现了对控制SOM动力学过程的最新理解。在这里，我们展示了MEMS模型继续发展的结果，2.0版本。MEMS 2.0是一个完整的生态系统模型，其模块模拟植物在地上和地下输入、土壤水分和温度的分层、植物输入和SOM的分解以及氮的矿化和固定化。该模型模拟了两个通常测量的有机质池-颗粒和矿物相关有机质（分别为POM和MAOM）。我们给出了在美国几个草地站点对模型进行校准和验证的结果。MEMS 2.0总体上捕获了整个土壤剖面的土壤碳(C)储量（R2分别为0.89和0.6）及其在POM和MAOM之间的分布。模拟土壤N与实测值相匹配，但精度低于土壤c （SOM中全氮的校正和验证R2分别为0.73和0.31）。模拟土壤水和温度与实测值进行了比较，精度与其他常用模型相当。该模型较好地捕捉了总初级生产量（GPP, R2 = 0.83）、生态系统呼吸（ER, R2 = 0.89）、净生态系统交换（NEE, R2 = 0.67）和蒸散量（ET, R2 = 0.71）的季节变化。我们将进一步发展该模型来代表森林和农业系统，并对其进行改进，以纳入对SOM分解的新认识。
Abstract. For decades, predominant soil biogeochemical models have used conceptual soil organic matter (SOM) pools and only simulated them to a shallow depth in soil. Efforts to overcome these limitations have prompted the development of the new generation SOM models, including MEMS 1.0, which represents measurable biophysical SOM fractions, over the entire root zone, and embodies recent understanding of the processes that govern SOM dynamics. Here we present the result of continued development of the MEMS model, version 2.0. MEMS 2.0 is a full ecosystem model with modules simulating plant growth with above- and belowground inputs, soil water and temperature by layer, decomposition of plant inputs and SOM, and mineralization and immobilization of nitrogen (N). The model simulates two commonly measured SOM pools – particulate and mineral-associated organic matter (POM and MAOM, respectively). We present results of calibration and validation of the model with several grassland sites in the US. MEMS 2.0 generally captured the soil carbon (C) stocks (R2 of 0.89 and 0.6 for calibration and validation, respectively) and their distributions between POM and MAOM throughout the entire soil profile. The simulated soil N matches measurements but with lower accuracy (R2 of 0.73 and 0.31 for calibration and validation of total N in SOM, respectively) than for soil C. Simulated soil water and temperature were compared with measurements, and the accuracy is comparable to the other commonly used models. The seasonal variation in gross primary production (GPP; R2 = 0.83), ecosystem respiration (ER; R2 = 0.89), net ecosystem exchange (NEE; R2 = 0.67), and evapotranspiration (ET; R2 = 0.71) was well captured by the model. We will further develop the model to represent forest and agricultural systems and improve it to incorporate new understanding of SOM decomposition.