The upper limits of the ion pickup and cold ion outflow loss rates from the early martian atmosphere shortly after the Sun arrived at the Zero-Age-Main-Sequence (ZAMS) were investigated. We applied a comprehensive 3-D multi-species magnetohydrodynamic (MHD) model to an early martian CO(2)-rich atmosphere, which was assumed to have been exposed to a solar XUV [X-ray and extreme ultraviolet (EUV)] flux that was 100 times higher than today and a solar wind that was about 300 times denser. We also assumed the late onset of a planetary magnetic dynamo, so that Mars had no strong intrinsic magnetic field at that early period. We found that, due to such extreme solar wind-atmosphere interaction, a strong magnetic field of about approximately 4000 nT was induced in the entire dayside ionosphere, which could efficiently protect the upper atmosphere from sputtering loss. A planetary obstacle ( approximately ionopause) was formed at an altitude of about 1000 km above the surface due to the drag force and the mass loading by newly created ions in the highly extended upper atmosphere. We obtained an O(+) loss rate by the ion pickup process, which takes place above the ionopause, of about 1.5 x 10(28) ions/s during the first < or =150 million years, which is about 10(4) times greater than today and corresponds to a water loss equivalent to a global martian ocean with a depth of approximately 8 m. Consequently, even if the magnetic protection due to the expected early martian magnetic dynamo is neglected, ion pickup and sputtering were most likely not the dominant loss processes for the planet's initial atmosphere and water inventory. However, it appears that the cold ion outflow into the martian tail, due to the transfer of momentum from the solar wind to the ionospheric plasma, could have removed a global ocean with a depth of 10-70 m during the first < or =150 million years after the Sun arrived at the ZAMS.

[1] The energy transport of bursty bulk flows (BBFs) is very important to the understanding of substorm energy transport. Previous studies all use the MHD bulk parameters to calculate the energy flux density of BBFs. In this paper, we use the kinetic approach, i.e., ion velocity distribution function, to study the energy transport of an earthward bursty bulk flow observed by Cluster C1 on 30 July 2002. The earthward energy flux density calculated using kinetic approach Q Kx is obviously larger than that calculated using MHD bulk parameters Q MHDx . The mean ratio Q Kx /Q MHDx in the flow velocity range 200-800 km/s is 2.7, implying that the previous energy transport of BBF estimated using MHD approach is much underestimated. The underestimation results from the deviation of ion velocity distribution from ideal Maxwellian distribution. The energy transport of BBF is mainly provided by ions above 10 keV although their number density N f is much smaller than the total ion number density N. The ratio Q Kx /Q MHDx is basically proportional to the ratio N/N f . The flow velocity v(E) increases with increasing energy. The ratio N f /N is perfectly proportional to flow velocity V x . A double ion component model is proposed to explain the above results. The increase of energy transport capability of BBF is important to understanding substorm energy transport. It is inferred that for a typical substorm, the ratio of the energy transport of BBF to the substorm energy consumption may increase from the previously estimated 5% to 34% or more.