Heliospheric interstellar H temperature from SOHO/SWAN H cell data

International audienceWe show a first comparison between selected SOHO/SWAN H cell data recorded in 1996-1997 and a simple classical "hot model" of the interstellar (IS) H flow in the inner heliosphere. Our goal is to obtain some constraints on the interplanetary background Ly-[FORMULA] profiles, for the first time without any assumption on the H cell characteristics. For this purpose the H cell optical thickness and its temperature are free parameters of the study, but we assume that the direction of the flow and the allowed range for the upwind line-of-sight apparent Doppler shift are known from previous studies.We derive apparent temperatures (or line-of-sight (LOS) temperatures) between 11,000 and 20,000 K according to the direction. This implies a significant broadening with respect to the linewidths expected for a flow at the same temperature as the interstellar helium flow (6,000 [FORMULA] 1000 K) in the optically thin approximation. Radiative transfer is probably responsible for a fraction of this effect, and heating at the heliospheric interface for the remaining. The best solutions are found for an upwind velocity of 26 km s-1, in excellent agreement with an independent study by Quémerais et al. (1999), and for very similar H cell absorption width and temporal decrease. The deceleration of interstellar H at heliopause crossing is found to be between 2.5 and 4.5 km s-1.We also use one particular H cell absorption map to derive directly from the data how the LOS temperature (or linewidth) varies with the angle with the wind direction. Interestingly, we measure a temperature minimum between the upwind and crosswind directions, while classical models predict a monotonic increase of the LOS temperature from upwind to downwind. We believe that this behavior is the first evidence for the existence of two distinct populations at different velocities (primary and secondary IS atoms), as predicted by heliosphere-interstellar gas interface models. If confirmed, this should be an extremely good diagnostic of the interface