Large scale changes in the state of the land surface affect the circulation of the atmosphere, the structure and function of ecosystems, as well as health and economy of mankind. As global temperatures increase and regional climates change, the timing of plant phenological events will shift as well. Understanding and anticipating those changes require both observations of large-scale interannual phenological variability and global climate model simulations with realistic land surface phenology routines. Therefore, in my dissertation, I combined thermal-based indices, satellite remote sensing, and the Community Land Model (CLM) to characterize spring phenology variability and its linkage to the climate system and evaluate the skill of the CLM to represent spring onset and seasonal variations of plants. I developed a new suite of thermal-based indicators to characterize the seasonal window of spring onset. Results showed that temperature dominates spring onset timing over the Northern Hemisphere and shifts the seasonal window of spring phenological changes as a whole. In addition, spring phenology has large interannual to decadal variation and its trends depend dramatically on the examined historical periods. Because of their long temporal depth and good spatial coverage, the newly-developed thermal-based indices can provide useful information in isolating the role of the climate system in altering spring onset. Evaluating how well state-of-the-art climate models can represent the above variabilities and trends are important for understanding and improving model performance. Therefore, I also evaluate a new suite of phenometrics designed to facilitate an “apples to apples” comparison between remote sensing products and climate model output. This systematic approach to comparing phenologically-relevant variables reveals broad consistency between the model and observations in large-scale spatial gradients of LAI amplitudes and mean spring onset dates. However, it exhibits fundamental difference between CLM and MODIS LAI seasonal cycle and spring onset timing. Therefore, any coupling between the land surface and the atmosphere that depends on vegetation state might not be fully captured by the existing generation of models. As a result, any future feedback of carbon, moisture, and energy that affect this coupling would be subject to sources of uncertainty originating in model phenology.