Earth's plasmasphere persists as an extension of the ionosphere into space. The toroidal region of plasma is shaped by electric and magnetic forces in the terrestrial magnetosphere. As a dense population of cold plasma, the plasmasphere interacts with particles in the hot ring current and energetic radiation belts. Evolution of plasmaspheric density under the driving influence of the solar wind crosses many physical scales. Convective erosion during geomagnetic storms occurs on the order of hours, reducing the size of the plasmasphere by forming an abrupt plasmapause density gradient that varies in radial and diurnal location. The history of geomagntic activity determines the presence of morphological structures as small as notches and as large as plumes. Plasma of atmospheric origin is carried sunward by convection through drainage plumes towards the magnetopause where it can diminish the effectiveness of magnetic reconnection. Long-lived plumes are sustained by a higher rate of refilling than typically observed during plasmasphere recovery from geomagnetic disturbances. The response of the plasmasphere, then, is an integral part of the feedback cycle between the magnetosphere and ionosphere in the exchange of energy and particles. This thesis aims to address three questions concerning the nature of the plasmasphere through the development of empirical and physics-based models under recent observations provided by the Van Allen Probes (RBSP-A & -B). First, what is the distribution of density content in the plasmasphere? For a two year period with full MLT coverage by RBSP, the upper-hybrid resonance frequency in plasma wave spectra is used to identify sudden changes consistent with the plasmapause feature and to calculate the magnetic equatorial electron density. Plasmapause encounter radial locations for both spacecraft are correlated with a geomagnetic activity index showing significant scatter around a linear fit. On average, the predicted plasmapause location does account for the separation between the saturated plasmasphere and the depleted plasmatrough. A density threshold corresponding to the plasmapause boundary is used to sort RBSP measurements into these two classified plasma regions. Model profiles are developed for each region and compared to the results from previous missions. The importance of solar wind properties in regulating the severity of plasmasphere erosion is demonstrated. Second, how does the plasmapause form and vary with geomagnetic activity? The two-dimensional plasmasphere density model, RAM-CPL, is employed to simulate two geomagnetic storms observed by the RBSP spacecraft. Inner-magnetospheric convection is parameterized by the Kp-index and solar wind properties. The performance of RAM-CPL is evaluated by the correspondence between virtual and actual plasmapause encounters. Overall, RAM-CPL achieved good agreement with RBSP observations of the plasmapause to within 0.5 L and measurements of electron density to within one order of magnetude inside the plasmasphere. An empirical model of ring current-ionosphere feedback was included to account for asymmetric erosion, but did not contribute significantly in the MLT sectors of interest when compared to electric field measurements. The difference in background activity level during quiet conditions between the two convection parameterizations was found to lead to 1 L difference in plasmapause location for each simulation trial. Solar wind driven simulations produce sharper and deeper erosion of the plasmapause at the onset of a geomagnetic storm, but also allow for larger recovery of the plasmasphere when compared to Kp-index driven simulations. Third, what is the role of the ionosphere in sustaining the plasmasphere? Four geomagnetic events are observed by RBSP in opposing MLT sectors to exhibit undisturbed plasmasphere refilling following significant erosion of the plasmapause. RAM-CPL simulations of the strongest storm parameterized by solar wind properties shows the full evolution of plasmasphere density from the narrowing of a sunward plume at the onset of erosion, that begins to corotate into a duskside bulge as activity diminishes, to the outward recovery of the plasmapause over several days. A piecewise empirical model of plasmasphere refilling is composed from profiles of equatorial electron density and the observed correlation between the Kp-index and plasmapause location. The RAM-CPL timescale of refilling mediates the increase in density from plasmatrough to plasmasphere levels matching RBSP measurements during the quiet period after the storm. Density observations of the other geomagnetic events are consistent with reports of a two-stage refilling process.