Checking the LuSci profile restoration

Lunar scintillometer, LuSci, serves for measurement of surface-layer turbulence at a number of established and new sites. An original 4-channel prototype worked at CTIO and at LCO, while 6-channel instruments are being used by ESO. Methods of extracting turbulence profile (TP) C n(h) from scintillation covariances are still a subject of research. Of interest are the accuracy and robustness of restored TPs. The linear method of “layers” [2] was replaced in 2008 by a more elaborate model-fitting, representing the TP by linear (in log-log coordinates) segments between selected pivot points [4]. This technique is inspired by data analysis of SHABAR [1]. Meanwhile, the scintillometer array developed by the University of Vancouver fits data with double-exponential model [5]. It was demonstrated that the pivot-point method (PPM) produces results not very different from the previous layers method when applied to the 4-channel prototype [4]. However, the TPs derived from the 6-element LuScis systematically show low C n values at the 16-m point, which is un-realistic. Limited comparison of LuSci with SL-SLODAR at Paranal in October 2008 also demonstrated this effect. The reliability of the PPM is thus put in question, warranting further study.