Summary: | Icebergs account for approximately half the freshwater flux into the ocean from the polar ice sheets, and play a major role in the distribution of melt water in the ocean, but their melting is currently parameterized in a rudimentary way in global climate models. In this thesis, a combination of observations, laboratory experiments, and numerical modeling is used to suggest improved iceberg melt rate parameterizations. This work is further introduced and motivated in Chapter 1. In Chapter 2, an observational record of icebergs in a Greenland fjord is used to show that icebergs subject to strongly sheared flows predominantly move with the vertical average of the ocean currents. If, as typical in iceberg parameterizations, only the surface ocean velocity is taken into account, iceberg speed and basal melt may have errors in excess of 60%. These results emphasize the need for parameterizations to consider ocean properties over the entire iceberg draft. A series of laboratory experiments conducted to determine the dependence of submarine iceberg side melting on a background flow is described in Chapter 3. It is shown that two distinct regimes of melting exist depending on the flow magnitude and consequent behavior of melt plumes (side-attached or side-detached), with correspondingly different meltwater spreading characteristics. These results are subsequently extended to ice blocks melting in a two-layer vertically sheared flow. Finally, an idealized numerical configuration is used in Chapter 4 to compare the three-equation parameterization of melt, developed for application under ice shelves, and the standardly used bulk parameterization of iceberg basal melt. The results suggest an updated velocity-independent version of this parameterization for use in calculating the basal melt rate of tabular icebergs, to account for the changes in ocean properties caused by the physical presence of a large iceberg in the ocean. The results of this thesis and their implications are discussed in Chapter 5.
|