First continuous temperature measurements within polar mesosphere summer echoes

Simultaneous measurements of polar mesosphere summer echoes (PMSE) by a VHF radar and temperatures in the mesopause region by a potassium lidar were performed on Spitsbergen (78∘N) in the years 2001 and 2003. In total 254 h of simultaneous observations are available. The lidar also detected 226 h of noctilucent clouds. In the presence of PMSE temperatures are practically always lower than the frost point which confirms the importance of ice particles for PMSE. However, we also observe extremely low temperatures but no PMSE which we explain by too weak or missing turbulence. The potential role of missing water vapor is discussed in the manuscript. Around the mesopause approximately 80% of the observations show very low temperatures but not PMSE. The importance of sufficiently strong turbulence increases with height since increasing viscosity destroys small scale fluctuations. This explains why PMSE intensity rapidly decreases above the maximum signal at ∼83-89 km, and why the radar very seldom detected PMSE above ∼92km, although it is persistently cold there and ice particles presumably exist. These observations demonstrate that the non-existence of PMSE cannot be taken as a proof for the absence of ice particles. Furthermore, our measurements indicate that the freeze-drying effect may not be as important as previously anticipated. From our temperature measurements we derive huge degrees of saturation (up to 106-108) which allows for homogeneous nucleation. In the altitude range of best data coverage (∼88-93 km) the contribution of S>100(S>10,000) to all cases of supersaturation is approximately 40-50% (10-20%). The very large values may explain why models which ignore homogeneous nucleation notoriously create too few ice particles.