More Like this

1. Consistent Predictability of the Ocean State Ocean Model (OSOM) using Information Theory and Flushing Timescales. Earth and Space Science Open Archive 20200. doi: 10.1002/essoar.10504826.1.

   https://doi.org/10.1002/essoar.10504826.1  

   Sane, A ; Fox-Kemper, B ; Ullman, D ; Kincaid, C ; Rothstein, L. ( November 2020 , Journal of geophysical research)

   null (Ed.)

   The Ocean State Ocean Model OSOM is an application of the Regional Ocean Modeling System spanning the Rhode Island waterways, including Narragansett Bay, Mt. Hope Bay, larger rivers, and the Block Island Shelf circulation from Long Island to Nantucket. This paper discusses the physical aspects of the estuary (Narragansett and Mount Hope Bays and larger rivers) to evaluate physical circulation predictability. This estimate is intended to help decide if a forecast and prediction system is warranted, to prepare for coupling with biogeochemistry and fisheries models with widely disparate timescales, and to find the spin-up time needed to establish the climatological circulation of the region. Perturbed initial condition ensemble simulations are combined with metrics from information theory to quantify the predictability of the OSOM forecast system--i.e., how long anomalies from different initial conditions persist. The predictability timescale in this model agrees with readily estimable timescales such as the freshwater flushing timescale evaluated using the total exchange flow (TEF) framework, indicating that the estuarine dynamics rather than chaotic transport is the dominant model behavior limiting predictions. The predictability of the OSOM is ~ 7 to 40 days, varying with parameters, region, and season. 

   more » « less
2. A Comparison of Data‐Driven Approaches to Build Low‐Dimensional Ocean Models

   https://doi.org/10.1029/2021MS002537  

   Agarwal, Niraj ; Kondrashov, D. ; Dueben, P. ; Ryzhov, E. ; Berloff, P. ( September 2021 , Journal of Advances in Modeling Earth Systems)

   We present a comprehensive inter‐comparison of linear regression (LR), stochastic, and deep‐learning approaches for reduced‐order statistical emulation of ocean circulation. The reference data set is provided by an idealized, eddy‐resolving, double‐gyre ocean circulation model. Our goal is to conduct a systematic and comprehensive assessment and comparison of skill, cost, and complexity of statistical models from the three methodological classes. The model based on LR is considered as a baseline. Additionally, we investigate its additive white noise augmentation and a multi‐level stochastic approach, deep‐learning methods, hybrid frameworks (LR plus deep‐learning), and simple stochastic extensions of deep‐learning and hybrid methods. The assessment metrics considered are: root mean squared error, anomaly cross‐correlation, climatology, variance, frequency map, forecast horizon, and computational cost. We found that the multi‐level linear stochastic approach performs the best for both short‐ and long‐timescale forecasts. The deep‐learning hybrid models augmented by additive state‐dependent white noise came second, while their deterministic counterparts failed to reproduce the characteristic frequencies in climate‐range forecasts. Pure deep learning implementations performed worse than LR and its simple white noise augmentation. Skills of LR and its white noise extension were similar on short timescales, but the latter performed better on long timescales, while LR‐only outputs decay to zero for long simulations. Overall, our analysis promotes multi‐level LR stochastic models with memory effects, and hybrid models with linear dynamical core augmented by additive stochastic terms learned via deep learning, as a more practical, accurate, and cost‐effective option for ocean emulation than pure deep‐learning solutions.

    

   more » « less
3. Modeling the Impact of Ocean Circulation on Chlorophyll Blooms Around South Georgia, Southern Ocean

   https://doi.org/10.1029/2020JC016391  

   Matano, R. P. ; Combes, V. ; Young, E. F. ; Meredith, M. P. ( September 2020 , Journal of Geophysical Research: Oceans)

   The northeast periphery of the Scotia Sea hosts one of the largest chlorophyll‐a blooms of the Southern Ocean. This bloom peaks to the northwest of the island of South Georgia, extending eastward for hundreds of kilometers. Although the Southern Ocean has many islands of similar size, South Georgia is ecologically one of the most significant: It not only sustains one of the Southern Ocean's largest and most diverse ecosystems but also constitutes its single most important region for biological carbon sequestration. While the exceptional nature of South Georgia's blooms has been recognized widely, both the physical processes that contribute to their fertilization and the reasons why these blooms are larger than those of other similar regions (e.g., Kerguelen or Crozet Islands) are poorly understood. We use the results of a high‐resolution ocean model to investigate the physical processes that mediate the entrainment of deep, iron‐rich waters into the surface layers of the South Georgia region. We show that the Southern Antarctic Circumpolar Current Front, the southernmost jet of the Antarctic Circumpolar Current (ACC), pumps iron‐enriched waters from the deep ocean onto the bottom layers of South Georgia's shelf. These waters are upwelled along the northern coast of the island and are then exported into the Georgia Basin, where topographically steered circulation shields them from the dispersive effects of local currents and eddies, thus allowing the bloom development.

    

   more » « less
4. Reviews and syntheses: Physical and biogeochemical processes associated with upwelling in the Indian Ocean

   https://doi.org/10.5194/bg-18-5967-2021  

   Vinayachandran, Puthenveettil Narayana ; Masumoto, Yukio ; Roberts, Michael J. ; Huggett, Jenny A. ; Halo, Issufo ; Chatterjee, Abhisek ; Amol, Prakash ; Gupta, Garuda V. ; Singh, Arvind ; Mukherjee, Arnab ; et al ( January 2021 , Biogeosciences)

   Abstract. The Indian Ocean presents two distinct climate regimes. The north Indian Ocean is dominated by the monsoons, whereas the seasonal reversal is less pronounced in the south. The prevailing wind pattern produces upwelling along different parts of the coast in both hemispheres during different times of the year. Additionally, dynamical processes and eddies either cause or enhance upwelling. This paper reviews the phenomena of upwelling along the coast of the Indian Ocean extending from the tip of South Africa to the southern tip of the west coast of Australia. Observed features, underlying mechanisms, and the impact of upwelling on the ecosystem are presented. In the Agulhas Current region, cyclonic eddies associated with Natal pulses drive slope upwelling and enhance chlorophyll concentrations along the continental margin. The Durban break-away eddy spun up by the Agulhas upwells cold nutrient-rich water. Additionally, topographically induced upwelling occurs along the inshore edges of the Agulhas Current. Wind-driven coastal upwelling occurs along the south coast of Africa and augments the dynamical upwelling in the Agulhas Current. Upwelling hotspots along the Mozambique coast are present in the northern and southern sectors of the channel and are ascribed to dynamical effects of ocean circulation in addition to wind forcing. Interaction of mesoscale eddies with the western boundary, dipole eddy pair interactions, and passage of cyclonic eddies cause upwelling. Upwelling along the southern coast of Madagascar is caused by the Ekman wind-driven mechanism and by eddy generation and is inhibited by the Southwest Madagascar Coastal Current. Seasonal upwelling along the East African coast is primarily driven by the northeast monsoon winds and enhanced by topographically induced shelf breaking and shear instability between the East African Coastal Current and the island chains. The Somali coast presents a strong case for the classical Ekman type of upwelling; such upwelling can be inhibited by the arrival of deeper thermocline signals generated in the offshore region by wind stress curl. Upwelling is nearly uniform along the coast of Arabia, caused by the alongshore component of the summer monsoon winds and modulated by the arrival of Rossby waves generated in the offshore region by cyclonic wind stress curl. Along the west coast of India, upwelling is driven by coastally trapped waves together with the alongshore component of the monsoon winds. Along the southern tip of India and Sri Lanka, the strong Ekman transport drives upwelling. Upwelling along the east coast of India is weak and occurs during summer, caused by alongshore winds. In addition, mesoscale eddies lead to upwelling, but the arrival of river water plumes inhibits upwelling along this coast. Southeasterly winds drive upwelling along the coast of Sumatra and Java during summer, with Kelvin wave propagation originating from the equatorial Indian Ocean affecting the magnitude and extent of the upwelling. Both El Niño–Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) events cause large variability in upwelling here. Along the west coast of Australia, which is characterized by the anomalous Leeuwin Current, southerly winds can cause sporadic upwelling, which is prominent along the southwest, central, and Gascoyne coasts during summer. Open-ocean upwelling in the southern tropical Indian Ocean and within the Sri Lanka Dome is driven primarily by the wind stress curl but is also impacted by Rossby wave propagations. Upwelling is a key driver enhancing biological productivity in all sectors of the coast, as indicated by enhanced sea surface chlorophyll concentrations. Additional knowledge at varying levels has been gained through in situ observations and model simulations. In the Mozambique Channel, upwelling simulates new production and circulation redistributes the production generated by upwelling and mesoscale eddies, leading to observations of higher ecosystem impacts along the edges of eddies. Similarly, along the southern Madagascar coast, biological connectivity is influenced by the transport of phytoplankton from upwelling zones. Along the coast of Kenya, both productivity rates and zooplankton biomass are higher during the upwelling season. Along the Somali coast, accumulation of upwelled nutrients in the northern part of the coast leads to spatial heterogeneity in productivity. In contrast, productivity is more uniform along the coasts of Yemen and Oman. Upwelling along the west coast of India has several biogeochemical implications, including oxygen depletion, denitrification, and high production of CH4 and dimethyl sulfide. Although weak, wind-driven upwelling leads to significant enhancement of phytoplankton in the northwest Bay of Bengal during the summer monsoon. Along the Sumatra and Java coasts, upwelling affects the phytoplankton composition and assemblages. Dissimilarities in copepod assemblages occur during the upwelling periods along the west coast of Australia. Phytoplankton abundance characterizes inshore edges of the slope during upwelling season, and upwelling eddies are associated with krill abundance. The review identifies the northern coast of the Arabian Sea and eastern coasts of the Bay of Bengal as the least observed sectors. Additionally, sustained long-term observations with high temporal and spatial resolutions along with high-resolution modelling efforts are recommended for a deeper understanding of upwelling, its variability, and its impact on the ecosystem. 

   more » « less
5. Systematic Bias Correction in Ocean Mesoscale Forecasting Using Machine Learning

   https://doi.org/10.1029/2022MS003426  

   Liu, Guangpeng ; Bracco, Annalisa ; Brajard, Julien ( November 2023 , Journal of Advances in Modeling Earth Systems)

   The ocean circulation is modulated by meandering currents and eddies. Forecasting their evolution is a key target of operational models, but their forecast skill remains limited. We propose a machine learning approach that improves the output of an ocean circulation model by learning and predicting its systematic biases. This method can be applied a priori to any region, and is tested in the Gulf of Mexico, where the Loop Current (LC) and the large anticyclonic eddies that detach from it are major forecasting targets. The LC dynamics are recurrent and lie on a low‐dimensional dynamical attractor. Building upon the information gained analyzing this low dimensional attractor, we improve the representation of sea surface anomalies in model outputs through information from satellite altimeter data using a Sequence‐to‐Sequence model, which is a special class of Recurrent Neural Network. Building upon the HYCOM‐NCODA analysis system, we deliver a correction to the forecast at the observation resolution. For at least 15 days the proposed method learns to forecast the systematic bias in the HYCOM‐NCODA, outperforming persistence, and improving the forecast. This data‐driven approach is fast and can be implemented as an added step to any dynamical hindcasting or forecasting model. It offers an interesting avenue for further developing hybrid modeling tools. In these tools, fundamental physical conservations are preserved through the integration of partial differential equations which obey them. In addition, the method highlights specific deficiencies of the hindcast system that deserve further investigation in the future.

    

   more » « less