@article{PAR00025687, title = {{U}nderstanding energy pathways in the {G}ulf {S}tream}, author = {{C}ontreras, {M}. and {R}enault, {L}ionel and {M}archesiello, {P}atrick}, editor = {}, language = {{ENG}}, abstract = {{T}he {G}ulf {S}tream ({GS}) is one of the strongest ocean currents on the planet. {E}ddy-rich resolution models are needed to properly represent the dynamics of the {GS}; however, kinetic energy ({KE}) can be in excess in these models if not dissipated efficiently. {T}he question of how and how much energy is dissipated and in particular how it flows through ocean scales thus remains an important and largely unanswered question. {U}sing a high-resolution (;2 km) ocean model [{C}oastal and {R}egional {O}cean {C}ommunity ({CROCO})], we characterize the spatial and temporal distribution of turbulent cascades in the {GS} based on a coarse-grained method. {W}e show that the balanced flow is associated with an inverse cascade while the forward cascade is explained by ageostrophic advection associated with frontogenesis. {D}ownscale fluxes are dominant at scales smaller than about 20 km near the surface and most intense at the {GS} {N}orth {W}all. {T}here is also strong seasonal variability in {KE} flux, with the forward cascade intensifying in winter and the inverse cascade later in spring. {T}he forward cascade, which represents an interior route to dissipation, is compared with both numerical and boundary dissipation pro-cesses. {T}he contribution of interior dissipation is an order of magnitude smaller than that of the other energy sinks. {W}e thus evaluate the sensitivity of horizontal momentum advection schemes on energy dissipation and show that the decrease in numerical dissipation in a high-order scheme leads to an increase in dissipation at the boundaries, not in the downscale flux.}, keywords = {{N}onlinear dynamics ; {S}mall-scale processes ; {O}cean models ; {ATLANTIQUE} ; {ATLANTIQUE} {NORD}}, booktitle = {}, journal = {{J}ournal of {P}hysical {O}ceanography}, volume = {53}, numero = {3}, pages = {719--736}, ISSN = {0022-3670}, year = {2023}, DOI = {10.1175/jpo-d-22-0146.1}, URL = {https://www.documentation.ird.fr/hor/{PAR}00025687}, }