Laure Zanna

Abstract:

© 2019, The Author(s). Interannual sea surface height (SSH) forecasts are subject to several sources of uncertainty. Methods relying on statistical forecasts have proven useful in assessing predictability and associated uncertainty due to both initial conditions and boundary conditions. In this study, the interannual predictability of SSH dynamics in the North Atlantic is investigated using the output from a 150 year long control simulation based on HadGEM3, a coupled climate model at eddy-permitting resolution. Linear inverse modeling (LIM) is used to create a statistical model for the evolution of monthly-mean SSH anomalies. The forecasts based on the LIM model demonstrate skill on interannanual timescales O(1–2 years). Forecast skill is found to be largest in both the subtropical and subpolar gyres, with decreased skill in the Gulf Stream extension region. The SSH initial conditions involving a tripolar anomaly off Cape Hatteras lead to a maximum growth in SSH about 20 months later. At this time, there is a meridional shift in the 0 m-SSH contour on the order of 0.5 ∘–1.5 ∘-latitude, coupled with a change in SSH along the US East Coast. To complement the LIM-based study, interannual SSH predictability is also quantified using the system’s average predictability time (APT). The APT analysis extracted large-scale SSH patterns which displayed predictability on timescales longer than 2 years. These patterns are responsible for changes in SSH on the order of 10 cm along the US East Coast, driven by variations in Ekman velocity. Our results shed light on the timescales of SSH predictability in the North Atlantic. In addition, the diagnosed optimal initial conditions and predictable patterns could improve interannual forecasts of the Gulf Stream’s characteristics and coastal SSH.

Abstract:

Tropical sea surface temperature (SST) and winds vary on a wide range of timescales and have a substantial impact on weather and climate across the globe. Here we study the variability of SST and zonal wind during El Niño‐Southern Oscillation (ENSO) between 1982 and 2014. We focus on changes in extreme statistics using higher‐order moments of SST and zonal winds. We find that ENSO characteristics exhibit bimodal distributions and fat tails with extreme warm and cold temperatures in 1982–1999, but not during 2000–2014. The changes in the distributions coincide with changes in the intensity of ENSO events and the phase of the Interdecadal Pacific Oscillation. We also find that the strongest Easterly Wind Bursts occur during extreme El Niños and not during La Niñas. Maps of SST kurtosis can serve as a diagnostic for the thermocline feedback mechanism responsible for the differences in ENSO diversity between the two periods.

Abstract:

Most of the excess energy stored in the climate system due to anthropogenic greenhouse gas emissions has been taken up by the oceans, leading to thermal expansion and sea-level rise. The oceans thus have an important role in the Earth’s energy imbalance. Observational constraints on future anthropogenic warming critically depend on accurate estimates of past ocean heat content (OHC) change. We present a reconstruction of OHC since 1871, with global coverage of the full ocean depth. Our estimates combine timeseries of observed sea surface temperatures with much longer historical coverage than those in the ocean interior together with a representation (a Green’s function) of time-independent ocean transport processes. For 1955–2017, our estimates are comparable with direct estimates made by infilling the available 3D time-dependent ocean temperature observations. We find that the global ocean absorbed heat during this period at a rate of 0.30 ± 0.06 W/m2 in the upper 2,000 m and 0.028 ± 0.026 W/m2 below 2,000 m, with large decadal fluctuations. The total OHC change since 1871 is estimated at 436 ± 91 ×1021 J, with an increase during 1921–1946 (145 ± 62 ×1021 J) that is as large as during 1990–2015. By comparing with direct estimates, we also infer that, during 1955–2017, up to one-half of the Atlantic Ocean warming and thermosteric sea-level rise at low latitudes to midlatitudes emerged due to heat convergence from changes in ocean transport.

Abstract:

©2019. The Authors. Oceanographic observations are limited by sampling rates, while ocean models are limited by finite resolution and high viscosity and diffusion coefficients. Therefore, both data from observations and ocean models lack information at small and fast scales. Methods are needed to either extract information, extrapolate, or upscale existing oceanographic data sets, to account for or represent unresolved physical processes. Here we use machine learning to leverage observations and model data by predicting unresolved turbulent processes and subsurface flow fields. As a proof of concept, we train convolutional neural networks on degraded data from a high-resolution quasi-geostrophic ocean model. We demonstrate that convolutional neural networks successfully replicate the spatiotemporal variability of the subgrid eddy momentum forcing, are capable of generalizing to a range of dynamical behaviors, and can be forced to respect global momentum conservation. The training data of our convolutional neural networks can be subsampled to 10–20% of the original size without a significant decrease in accuracy. We also show that the subsurface flow field can be predicted using only information at the surface (e.g., using only satellite altimetry data). Our results indicate that data-driven approaches can be exploited to predict both subgrid and large-scale processes, while respecting physical principles, even when data are limited to a particular region or external forcing. Our in-depth study presents evidence for the successful design of ocean eddy parameterizations for implementation in coarse-resolution climate models.

Abstract:

The ocean plays an important role in the climate system on timescales of weeks to centuries. Despite improvements in ocean models, dynamical processes involving multi‐scale interactions remain poorly represented, leading to errors in forecasts. We present recent advances in understanding, quantifying and representing of physical and numerical sources of uncertainty in novel regional and global ocean ensembles at different horizontal resolutions. At coarse‐resolution, uncertainty in 21st‐century projections of the upper overturning cell in the Atlantic is mostly a result of buoyancy fluxes, while the uncertainty in projections of the bottom cell is driven equally by both wind and buoyancy flux uncertainty. In addition, freshwater and heat fluxes are the largest contributors to Atlantic Ocean Heat Content regional projections and to its uncertainty, mostly as a result of uncertain ocean circulation projections. At both coarse‐ and eddy‐permitting resolution, the unresolved stochastic temperature and salinity fluctuations can lead to significant changes in large‐scale density across the Gulf Stream front, therefore leading to major changes in large‐scale transport. These perturbations can have an impact on the ensemble spread on monthly time‐scales and subsequently interact non‐linearly with the dynamics of the flow generating chaotic variability on multi‐annual timescales. In the Gulf Stream region, the ratio of chaotic variability to atmospheric‐forced variability in meridional heat transport is larger than 50% on timescales shorter than 2 years; while between 40 and 48 ∘S the ratio exceeds 50% on on time scales up to 28 years. Based on these simulations, we show that air‐sea interaction and ocean sub‐grid eddies remain an important source of error for simulating and predicting ocean circulation, sea level, and heat uptake on a range of spatial and temporal scales. We discuss how further refinement of these ensembles can help us assess the relative importance of oceanic versus atmospheric uncertainty in weather and climate.