With　the　advent　of　the　Internet　of　vehicles　(IoV),　the　explosive　growth　of　data　from　vehicular　sensors　places　a　heavy　computing　burden　on　green-enabled　intelligent　transportation　systems.　In　this　study,　a　paradigm　of　vehicular　fog　computing　(VFC)　is　introduced,　which　is　able　to　offload　computation　tasks　to　the　nearby　fog　nodes.　In　addition,　considering　the　advantage　of　non-orthogonal　multiple　access　(NOMA),　a　NOMA-enabled　VFC　is　proposed,　in　which　a　task　vehicle,　a　main　fog　access　point　(F-AP),　an　idle　vehicle,　and　other　cooperative　F-APs　are　involved　to　process　the　task.　By　jointly　optimizing　NOMA　power　allocation,　task　allocation　ratio,　bandwidth　allocation　and　time　slot　allocation,　the　offloading　data　maximization　problem　is　investigated.　After　simplifying　the　problem　through　mathematical　analysis,　the　offloading　data　maximization　problem　is　solved　by　the　proposed　interior-point　method　based　on　successive　convex　approximation　(SCA).　Simulation　results　show　that　the　proposed　offloading　scheme　performs　better　than　other　existing　schemes　in　terms　of　offloading　data.