Soil carbon (C) feedbacks to the atmosphere remain one of the greatest areas of uncertainty in future climate predictions, largely due to an inadequate representation of three key factors in ecosystem models -- soil organic matter (SOM) chemistry, microbial composition and function, and ecosystem and soil heterogeneity. Soil biogeochemical models fail to accurately represent the chemical and physical composition of SOM fractions and instead rely on black box characterizations of the complex soil matrix. To gain a better understanding of how SOM chemistry and ecosystem and soil heterogeneity impact carbon feedbacks under warming, we propose to use a global scale experiment established in several montane ecosystems to break open the simple C pools represented in two models. Specifically, we have two objectives: (1) We will characterize soil C biochemical classes in SOM across experimental treatments using powerful, high-resolution spectrometry and spectroscopy techniques. Soil C data will provide a detailed understanding of C pools within soils across a globally distributed network of experiments in montane regions undergoing experimental warming manipulations. Mountain gradients will enable us to distinguish how long-term warming (high vs. low elevations), short-term warming (experimental manipulation, ~2 degrees C warming), impact soils in regions with varying edaphic factors. (2) We will attempt to incorporate the experimental results with two mechanistic SOC cycling models (MIMICS & CORPSE) that explicitly represent microbial growth and decomposition activity and have previously been applied at ecosystem and global scales. We aim to break apart the existing model organic matter pools into mechanistically-defined, measurable chemical functional groups that correspond to measurable C fractions. We will test if this level of detail is needed in modeling efforts (and if so, this will be a nice proof of concept for a larger project). In sum, using gradients of climate and SOC quality from experiments distributed in several mountain regions coupled with in depth soil C characterization, we can begin to test if incorporating more soil C measurements in as well as information into experiments and models is a useful way forward in increasing our understanding of soil C responses to warming.