Small geophysical signals detected in differences of altimetric tides and bottom-pressure tides

Very precise tidal estimates derived from seafloor pressure measurements and from two decades of satellite altimetry are capable of revealing tiny, subtle signals in their differences. In particular, these differences can place constraints on models of atmospheric tides and solid tides. Such signals arise because altimetry is sensitive to the solid tide while bottom pressure (BP) is not, while BP is sensitive to the air tide while altimetry is not (not directly anyway). I have constructed a database of precise BP tide estimates from 151 deep-ocean stations, most based on multiple years of hourly data. The relationship between bottom pressure and equivalent sea-surface elevation must be carefully handled, since it depends on ocean stratification and the compressibility of seawater. The altimeter-BP rms differences are 5 mm or better. In these differences, the air tide is easily detected at the S1 and S2 periods and more marginally detected at the T2 and K1 periods (the latter represent seasonal sidelines of either S1 or S2); a comparison of various air-tide models favors one derived from recent 3-hourly ECMWF operational analyses. Results also show that altimetry can benefit from correcting for crustal loading by the S2 air tide; the signal is only 1.4 mm or less and has previously been ignored, but it is detectable in the altimeter-BP differences. Similarly altimetry must be corrected for the presence of the air tide in the dry troposphere correction. Finally differences also allow Love numbers to be estimated at the M2, O1, and K1 periods, with the latter clearly showing suppression from the free core-nutation resonance. However, recent estimates just published by Krasna and colleagues, using VLBI data, are far more accurate, in part because VLBI need not contend with large, possibly inconsistent ocean signals in their data. Our results could be improved once baroclininc tides are properly accounted for, but these tides, with wavelengths of order 100 km, are currently insufficiently known and may require wide-swath altimetry before they are accurately mapped.