Snow water equivalent (SWE) is a critical measurement in hydrology and water resources management. Microwave remote sensing can estimate snow water equivalent (SWE). However, the algorithms used to estimate SWE require snow grain size information. Thus, determining snow grain size is pertinent to estimate SWE. Currently, there are several models that can estimate snow grain size in open environments. However, little consideration has been given to determine if there are differences in snow grain evolution between open and forested regions. In addition, there are few observational studies that have sought to understand these variations. Snow grain size is characterized using specific surface area (SSA). This study 1) compares SSA observations in forested and open settings; 2) evaluates four different snow grain evolution algorithms, incorporating each of them into a snowpack mass and energy balance. The study area is Grand Mesa, Colorado where snow grain size data were collected as part of the 2017 NASA SnowEx field campaign. Analysis of the observational data shows that forested regions tend to have less variation in SSA throughout the snow profile, due to the consistent and stable energy balance in the forest compared to the higher energy shifts that often occur in the open (i.e. diurnal fluxes, cloudy/sunny days, wind, etc.). Four different snow grain algorithms were incorporated into SnowModel. Two of the algorithms were empirically-based with fitted parameters and the other two algorithms were process-based with limited parameterizations. Additionally, a temperature adjustment was incorporated and offered a more refined estimate of grain size in forested areas. The four snow grain evolution algorithms were assessed by comparison between the observed SSA data, the outputs from the forest energy adjusted model, and outputs from the non-adjusted model. Performance was evaluated using goodness-of-fit metrics. In most cases, the temperature-adjusted forest energy balance model produced SSA results that better fit observed SSA values. This study enhances our understanding and the differences between snow grain size evolution in open and forested areas, as sample variance of SSA with canopy cover less than 50% is greater than it is when canopy cover is greater than 50%, demonstrating the value of physically-based snow modeling. Acknowledging these differences will advance estimation of snow grain size in the forest, an input into SWE algorithms inputs, thereby improving SWE retrieval over varied domains.