erratum : the correct HR surface absorption should be about 34 ppm (eg peak at z = 32 mm) while the previous estimated value is due to interferences.