Land surface models are an important piece of any numerical weather prediction or climate model. As we gain computing power and scientific knowledge it is important to update these models to more accurately represent the physics of the land surface. One of the current challenges in land surface modeling is to accurately represent vegetation, and in particular roots. A pressing issue is the inclusion of a realistic root distribution profile in models. Currently the Noah land surface model distributes roots evenly throughout each layer in the user defined root zone. Ideally a more realistic root distribution function would be used. We test the sensitivity of the model to differences in root mass and analyze the resulting energy fluxes from the surface, as well as the soil moisture profile. We then add a more realistic root distribution to the model, and validate our results against datasets from Boissy-le-Châtel, France, Illinois, and Oklahoma. The validation data include soil moisture profiles, and energy fluxes.