This thesis aims to disentangle internal gravity waves from balanced flow to investigatetheir role in the downscale transfer of energy in the ocean by using two conceptuallydifferent diagnostic procedures—a modal decomposition based on Machenhauer (1977)and a novel balance procedure called the quasi-geostrophic filter. Furthermore, using anidealized numerical model for baroclinically unstable flow in various dynamical scenarios,this thesis investigates both the strength of internal gravity wave generation and its rolein the energy dissipation at small scales.Internal gravity waves (IGWs) are first diagnosed using the non-linear initializationtechnique proposed by Machenhauer (1977) and are assessed in different dynamical regimesranging from ageostrophic to quasi-geostrophic. The complex coupling between IGWs andbalanced flow, in particular for an ageostrophic regime, renders frequency-wavenumberspectra inefficient for a precise detection of IGW signals. To tackle this complexity thefull flow field is decomposed into its balanced and unbalanced components using linearand non-linear modal decompositions. The linear decomposition performs reasonably wellbut does not include the non-linearity of the flow leading to discrepancies. To accountfor this, a non-linear decomposition based on Machenhauer’s technique is applied whichturns out to be more efficient and a robust diagnostic tool to obtain IGWs. An assessmentof the diagnosed IGWs in different dynamical regimes reveals that IGW activity is muchhigher in the ageostrophic than the quasi-geostrophic regime. Furthermore, IGWs dissipatepredominantly through small-scale dissipation, and therefore play a significant role in thedownscale transfer of energy in the ocean.A new balance procedure called the quasi-geostrophic (QG) filter is introduced to diag-nose IGWs, which additionally provides the ageostrophic horizontal velocities unavailablefrom previous potential vorticity inversion methods. The QG filter is implemented in aversion consistent at the discrete grid level and can be used as a balance procedure or asa diagnostic tool. Results show that the QG filter is an efficient tool to obtain balance,however a reformulated discrete version of the Machenhauer’s technique performs evenbetter. Further, IGWs diagnosed from the QG filter are used to examine IGW emissionfrom balanced flow in the ocean by different mechanisms: spontaneous emission, convectiveinstability by frontogenesis, and lateral boundary instability, where convective instabilityshows more IGW activity than the other two processes. The results establish IGWs as apotential sink of balanced flow energy which has key implications for the ocean’s energybudget.iii