Search

Over the south pole of Enceladus, an icy moon of Saturn, geysers eject water into space in a striped pattern, making Enceladus one of the most attractive destinations in the search for extraterrestrial life. We explore the ocean dynamics and tracer/heat transport associated with geysers as a function of the assumed salinity of the ocean and various core-shell heat partitions and bottom heating patterns. We find that, even if heating is concentrated into a narrow band on the seafloor directly beneath the south pole, the warm fluid becomes quickly mixed with its surroundings due to baroclinic instability. The warming signal beneath the ice is diffuse and insufficient to prevent the geyser from freezing over. Instead, if heating is assumed to be local to the geyser, emanating from tidal dissipation in the ice itself, the geyser can be sustained. In this case, the upper ocean beneath the ice becomes stably stratified and thus a barrier to vertical communication, leading to transit timescales from the core to the ice shell of hundreds of years.

The ice shell of Enceladus exhibits strong asymmetry between its hemispheres, with all known geysers concentrated over the south pole, even though its orbital configuration is almost perfectly symmetric. By exploring ocean circulation across a range of ocean salinities and core/shell heating partitions, we study the role of ice-ocean interaction in hemispheric symmetry breaking. We find that: (i) asymmetry is enhanced by cross-equatorial ocean heat transport when the ice shell is the major heat source and vice versa, (ii) the magnitude of ocean heat transport is comparable to the global heat production, significantly affecting the ice shell evolution and equilibrium state and (iii) more than one equilibrium state can exist due to a positive feedback between melting and ocean circulation.

Of profound astrobiological interest, Enceladus appears to have a global subsurface ocean that is salty, indicating water-rock reaction at present or in the past, important for its habitability. Here, we investigate how salinity and the partition of heat production between the silicate core and the ice shell affect ocean dynamics and the associated heat transport -- a key factor that determines the equilibrium ice shell geometry. Assuming steady state conditions, we show that the meridional overturning circulation of the ocean, driven by heat and salt exchange with the ice, has opposing signs at very low and very high salinities. Regardless of these differing circulations, heat and freshwater converge towards the equator, where the ice is thick, acting to homogenize thickness variations. In order to maintain the observed ice thickness variation, the polar-amplified ice dissipation needs to be strong enough and ocean heat convergence cannot overwhelm well-constrained heat loss rates through the thick equatorial ice sheet. This requirement is found violated if the main heat source is in the core rather than the ice shell, or if the ocean is very fresh or very salty. Instead, with a salinity of intermediate range, the temperature- and salinity-induced density gradient largely cancel one another, leading to much reduced overturning and equatorial heat convergence rates and consistent budgets in appearance of a significant ice dissipation.

Deep learning-based models have recently outperformed state-of-the-art seasonal forecasting models, such as for predicting El Ni\~no-Southern Oscillation (ENSO). However, current deep learning models are based on convolutional neural networks which are difficult to interpret and can fail to model large-scale atmospheric patterns. In comparison, graph neural networks (GNNs) are capable of modeling large-scale spatial dependencies and are more interpretable due to the explicit modeling of information flow through edge connections. We propose the first application of graph neural networks to seasonal forecasting. We design a novel graph connectivity learning module that enables our GNN model to learn large-scale spatial interactions jointly with the actual ENSO forecasting task. Our model, \graphino, outperforms state-of-the-art deep learning-based models for forecasts up to six months ahead. Additionally, we show that our model is more interpretable as it learns sensible connectivity structures that correlate with the ENSO anomaly pattern.

Beneath the icy shell encasing Enceladus, a small icy moon of Saturn, a global ocean of liquid water ejects geyser-like plumes into space through fissures in the ice, making it an attractive place to investigate habitability and to search for extraterrestrial life. The existence of an ocean on Enceladus has been attributed to the heat generated in dissipative processes associated with deformation by tidal forcing. However, it remains unclear whether that heat is mostly generated in its ice shell or silicate core. Answering this question is crucial if we are to unravel patterns of ocean circulation and tracer transport that will impact both the habitability of Enceladus and our ability to interpret putative evidence of any habitability and/or life. Using a nonhydrostatic ocean circulation model, we describe and contrast the differing circulation patterns and implied ice shell geometries to be expected as a result of heating in the ice shell above and heating in the core below Enceladus' ocean layer. If heat is generated primarily in the silicate core we would predict enhanced melting rates at the equator. In contrast, if heat is primarily generated in the ice shell we would infer a poleward-thinning ice geometry consistent with Cassini Mission observations.