Sea Level, Land Motion, and the Anomalous Tide at Churchill, Hudson Bay

Abstract:

The importance of the tide gauge at Churchill, Manitoba, cannot be overstated. It is the only permanently operating tide gauge in the central Canadian Arctic, and it sits on a prime spot for monitoring the mantle’s rebound from the Laurentide ice loss. Yet interpretation of the sea-level time series at Churchill has long been problematic, going back even to early work by Gutenberg in the 1940s. The long-term relative sea-level rates are inconsistent: approximately -4, -19, -5 ± 1 mm/y for the periods 1940-1970, 1970-1990, 1990-2014 respectively. Annual mean high water (MHW) and mean low water (MLW) reflect these trends until around 1990, after which MLW leveled off and is now nearly unchanging. Slightly later, around 2000, the semidiurnal tides became very anomalous, with falling amplitudes and slightly increasing phase lags. The amplitude of M2 was approximately 154 cm before 2000; it dropped to about 146 cm by 2010 and reached an all-time low of 142 cm in 2014. Satellite altimeter estimates of the tide in this region, although challenging because of seasonal ice cover, show no comparable M2 changes, so the tidal changes must be localized to the near vicinity of the gauge (or to the gauge itself if caused by a malfunction). On the other hand, altimetry confirms the post-1992 Churchill measurements of mean sea level, thanks to the long time series of land motion measurements obtained at GPS station CHUR, which gives a vertical uplift of 10.1 mm/y. Combining satellite altimeter data with the Churchill tide-gauge data gives an implied vertical crustal rate of about 9.0 ± 0.8 mm/y, in reasonable agreement with the GPS. In summary, we have still anomalous MSL measurements at the Churchill gauge for the intermediate 1970-1990 era, and very anomalous tidal measurements since 2000, but we have apparently quite reliable MSL rates since 1990.