As a fundamental energy converting process in space, magnetic reconnection is the key to understanding the coupling between the microscale kinetic process and the large-scale MHD process in the magnetosphere. When magnetic reconnection occurs, the magnetic field undergoes a topological change, and the free energy is released to plasmas (Birn & Priest, 2007;Yamada et al., 2010). Due to its ubiquitous presence, magnetic reconnection is often invoked as a promising source of explosive activities such as solar flares (Masuda et al., 1994), coronal mass ejection (Lin & Forbes, 2000), geomagnetic storms and substorms (Baker et al., 1996;Kepko et al., 2015). Furthermore, reconnection also plays a vital role in laboratory plasma facilities (Yamada et al., 1994).Significant progress has been made in investigating the energy conversion at the reconnection site during magnetic reconnection. Birn and Hesse (2005) have shown that although magnetic reconnection is triggered in the diffusion region, which is merely several ion inertial lengths ( 𝐴𝐴 𝐴𝐴𝑖𝑖 ≡ 𝑐𝑐∕𝜔𝜔𝑝𝑝𝑖𝑖 ) wide, the energy conversion is not localized in the diffusion region. They also found that in contrast to the case in the classical resistive Sweet-Parker model, Joule and ohmic dissipation can be neglected in the overall energy transfer in the collisionless magnetic reconnection. The Poynting flux driven into the reconnection site from the inflow region is partially converted to the bulk kinetic flow and the enthalpy flux of plasmas; the rest part is diverted to the exhaust region, forming the pileup front. Satellite observations and numerical simulations have shown that outflow mainly consists of the Poynting flux and the enthalpy flux; the bulk kinetic flux commonly plays a minor role (Birn & Hesse, 2010;Eastwood et al., 2013). In addition, ion enthalpy flux is larger than the electron enthalpy flux in most cases, indicating that ions intend to gain more energy than electrons during the reconnection (