# Seminars

# Atmosphere Ocean Science Friday Seminar

#### Two talks from our new students

**Speaker:**
Matt Pudig and Nick DeFilippis, CAOS

**Location:**
Warren Weaver Hall 1314

**Date:**
Friday, September 23, 2022, 4 p.m.

**Synopsis:**

Matt: Rectified ocean heat uptake from oscillatory surface forcing

It is known that surface cooling anomalies penetrate into the ocean faster than surface warming anomalies. Because of this asymmetry, the time-variable component of radiative forcing can induce a long-term, rectified cooling trend in ocean heat content. We explore this asymmetry and rectification on global and interannual scales, its implications and its possible dependence on model parameters. We do so using a full complexity global ocean--sea ice general circulation model and an idealized one-dimensional vertical mixing model. In both models, the ocean heat uptake response to an abrupt cooling perturbation is larger than an equal magnitude warming perturbation. This asymmetry appears to be larger when the background vertical diffusivity is smaller, and is therefore likely model-dependent. Sinusoidal oscillatory forcing with zero-time-mean induces a rectified cooling trend in both models, whose magnitude depends on both the diffusivity and the frequency of the interannual oscillatory forcing. We discuss this rectification effect in the context of volcanic forcing in climate models, whose time-variable component may cause artefactual and model-dependent ocean cooling in CMIP6 historical simulations.

Nick: Frugality of uncapacitated facility location

The Uncapacitated Facility Location Problem focuses on minimizing facility opening costs and client connection costs over a set of given facilities. We examine the private information setting of this problem where facility opening costs are only known to the facilities themselves and we must determine their value through an auction. Our main focus is establishing a bound on the frugality ratio of this problem, where frugality is defined as total overpayment compared to the second-cheapest disjoint solution, which we resolve by providing a tight bound of 2.