The Earth and its atmosphere are embedded in the magnetosphere, a region in space dominated by the geomagnetic field, shielding our planet as it acts to deflect the energetic solar wind. Even though the atmosphere is protected from direct interaction with the solar wind it is indirectly affected by significant magnetosphere-solar wind interaction processes, causing constituents of the upper atmosphere to flow up into the magnetosphere. The fate of the atmospheric originating ions is interesting from a planetary evolution point of view. If the upflowing ions in the magnetosphere are to escape into the solar wind they need to not only overcome gravity, but also the magnetic forces, and therefore need to be energized and accelerated significantly. The subject of this thesis is analysis of oxygen ions (O+) and wave field observations in the high altitude cusp/mantle and in the high latitude dayside magnetosheath of Earth, investigating magnetospheric processes behind ion heating, outflow and escape. Most data analysis is based on observational data from the Cluster satellites, orbiting the Earth and altitudes corresponding to different key regions of the magnetosphere and the immediate solar wind environment. The mechanism behind O+ heating mainly discussed in this thesis is energization through interactions between the ions and low-frequency waves. The average electric spectral densities in the altitude range of 8-15 Earth radii are able to explain the average perpendicular temperatures, using a gyroresonance model and 50% of the observed spectral density at the O+ gyrofrequency. Strong heating is sporadic and spatially limited. The regions of enhanced wave activity are at least one order of magnitude larger than the local gyroradius of the ions, which is a necessary condition for the gyroresonance model to be valid. An analysis indicates that enhanced perpendicular temperatures can be observed over several Earth radii after heating has ceased, suggesting that high perpendicular-to-parallel temperature ratio is not necessarily a sign of local heating. This also explains why we sometimes observe enhanced temperatures and low spectral densities. We also show that the phase velocities derived from the observed low frequency electric and magnetic fields are consistent with Alfvén waves. Outflowing ions flow along magnetic field lines leading downstream in the magnetotail, where the ions may convect into the plasma sheet and be brought back toward Earth. However, the effective heating in the cusp and mantle provides a majority of the O+ enough acceleration to escape into the solar wind and be lost, rather than entering the plasma sheet. The heating can actually be effective enough to allow outflowing cusp O+ to escape immediately from the high altitude cusp and mantle along recently opened magnetic field lines, facilitating a direct coupling between the magnetospheric plasma and interplanetary space. Observations in the shocked and turbulent solar wind (the magnetosheath) reveals hot O+ flowing downstream and approximately tangentially to the magnetopause and often close to it. An estimated total flux of O+ in the high-latitude magnetosheath of 0.7 ·1025 s-1 is significant in relation to the observed cusp outflows at lower altitudes, pointing to that escape of hot O+ from the cusp and mantle into the dayside magnetosheath being an important loss route.