Attribution and Predictability of Climate‐Driven Variability in Global Ocean Color

Lim, Hyung‐Gyu; Dunne, John; Stock, Charles A.; Kwon, Min Serk

doi:10.1029/2022jc019121

Cited by 14 publications

(21 citation statements)

References 115 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Coupling the CNN submodels with Recurrent Neural Networks (RNNs) should help accounting for temporal dynamics/time history with a more sophisticated way than if adding time-lags as predictors within our current architecture. Indeed, while instantaneous environmental fluctuations are thought to explain much of the observed phytoplankton temporal variability, time-lag responses of weeks to a few months would also be expected (Ji et al, 2010;Feng et al, 2015;Schollaert Uz et al, 2017;Lim et al, 2022b). This would arise from biological processes mainly, such as dormancy and reproduction (Ji et al, 2010), or ecological interactions as species competition or grazing pressure (Feng et al, 2015).…”

Section: Discussion and Perspectivesmentioning

confidence: 99%

“…Thus, our training period [2003][2004][2005][2006][2007][2008][2009][2010][2011] mainly hosts CP events, whereas the strong 2015/2016 El Niño event is usually classified as an EP event, with different processes and related impact on primary production (Radenac et al, 2012;Racault et al, 2017). Finally, delayed effects of climate modes have been very recently shown to influenced Chl in large parts of the ocean (see Figure 6 of Lim et al, 2022b), and especially in the eastern tropical Pacific one, whereas time-lags are not considered into our model.…”

Section: Chl Reconstruction Improvement From Mono-mode Cnn 1 To Multi...mentioning

confidence: 99%

See 1 more Smart Citation

A Multi-Mode Convolutional Neural Network to reconstruct satellite-derived chlorophyll-a time series in the global ocean from physical drivers

et al. 2023

View full text Add to dashboard Cite

Time series of satellite-derived chlorophyll-a concentration (Chl, a proxy of phytoplankton biomass), continuously generated since 1997, are still too short to investigate the low-frequency variability of phytoplankton biomass (e.g. decadal variability). Machine learning models such as Support Vector Regression (SVR) or Multi-Layer Perceptron (MLP) have recently proven to be an alternative approach to mechanistic ones to reconstruct Chl synoptic past time-series before the satellite era from physical predictors. Nevertheless, the relationships between phytoplankton and its physical surrounding environment were implicitly considered homogeneous in space, and training such models on a global scale does not allow one to consider known regional mechanisms. Indeed, the global ocean is commonly partitioned into biogeochemical provinces (BGCPs) into which phytoplankton growth is supposed to be governed by regionally-”homogeneous” processes. The time-evolving nature of those provinces prevents imposing a priori spatially-fixed boundary constraints to restrict the learning phase. Here, we propose to use a multi-mode Convolutional Neural Network (CNN), which can spatially learn and combine different modes, to globally account for interregional variabilities. Each mode is associated with a CNN submodel, standing for a mode-specific response of phytoplankton biomass to the physical forcing. Beyond improving performance reconstruction, we show that the different modes appear regionally consistent with the ocean dynamics and that they may help to get new insights into physical-biogeochemical processes controlling phytoplankton spatio-temporal variability at global scale.

show abstract

Section: Discussion and Perspectivesmentioning

confidence: 99%

Section: Chl Reconstruction Improvement From Mono-mode Cnn 1 To Multi...mentioning

confidence: 99%

A Multi-Mode Convolutional Neural Network to reconstruct satellite-derived chlorophyll-a time series in the global ocean from physical drivers

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Climate change is recognized as a potential catalyst for the expansion of phytoplankton blooms, affecting the physical‐chemical environment and biological processes in water bodies (Lim et al., 2022; Paerl et al., 2011; Paerl & Huisman, 2008; Paerl & Paul, 2012). Temperature and rainfall changes, influenced by climate change, play a significant role in simulating the blooms (Xiao et al., 2019).…”

Section: Introductionmentioning

confidence: 99%

Teleconnection Between Coastal Phytoplankton Blooms Phenomenon in Western North Pacific and El Niño–Southern Oscillation by Time‐Frequency Analysis

Liu,

Wang,

Zhao

et al. 2024

JGR Oceans

View full text Add to dashboard Cite

Phytoplankton blooms in coastal marine environments are regard as a double‐edged sword, offering benefits to coastal ecosystems while also posing significant environmental challenges. The occurrence and severity of these blooms are influenced by environmental factors, with climate oscillations playing a crucial role. In this study, we employed two time‐frequency analysis methods to investigate the teleconnection between phytoplankton blooms (denoted by bloom intensity, BI) and El Niño–Southern Oscillation (ENSO) from 2003 to 2020 in eight large marine ecosystems (LMEs) of Western North Pacific which are particularly sensitive to the impacts of climate oscillations and reflect well the impacts on marine ecosystems at different latitudes. These analytical methods unveiled these time series exhibited similar periodicity and synchronization patterns, which indicated the teleconnection between ENSO and phytoplankton blooms in the Western North Pacific. The Hilbert‐Huang Transformation classified the eight LMEs into two categories based on their similarity to BI and MEI in the time‐frequency domain. Specifically, one category is similar to MEI with energy concentrated at low frequencies, while the other category exhibits significant dissimilarity with dispersed energy. Wavelet analysis uncovered a robust coherence between the ENSO and BI across all LMEs, with a 57‐month periodicity. Furthermore, phytoplankton blooms in the Western North Pacific Ocean strengthen with increasing Multivariate ENSO Index on the large spatiotemporal scale due to a decrease in sea surface temperatures. The findings will be useful as a reference for improving understanding of the teleconnection between coastal phytoplankton blooms and climate oscillations in other coastal zones.

show abstract

“…In details, CHL exhibits pronounced fluctuations in responses to changes in the climate system of the tropical Pacific, and can induce biofeedback onto the latter. On the one hand, CHL concentration in the upper ocean is controlled by physical properties of sea waters, such as light intensity, water temperature, and stratification strength, which influences the availability of nutrients in the upper layer (Gnanadesikan et al., 2004; Lim, Dunne, Stock, Ginoux, et al., 2022, Lim, Dunne, Stock, & Kwon, 2022; Mantua et al., 1997; Sugimoto & Tadokoro, 1997). On the other hand, CHL affects the penetration of solar radiation in the upper ocean, modifying the vertical distribution of penetrative solar radiation, which in turn influences the climate system by changing the thermo‐dynamical structure in the upper ocean (Murtugudde et al., 2002; Nakamoto et al., 2001; Park, Kug, & Park, 2014; Patara et al., 2012; Zhang et al., 2009).…”

Section: Introductionmentioning

confidence: 99%

Impact of the Deep Chlorophyll Maximum in the Equatorial Pacific as Revealed in a Coupled Ocean GCM‐Ecosystem Model

Shi

Tian

2023

JGR Oceans

View full text Add to dashboard Cite

A coupled ocean general circulation model (OGCM)‐ocean ecosystem model is used to investigate the effects of the deep chlorophyll maximum (DCM) on the ocean state in the equatorial Pacific Ocean. Climatological and interannual three‐dimensional (3‐D) chlorophyll (CHL) fields are captured well in control runs using the coupled ocean physics‐ecosystem model forced by prescribed atmospheric fields. Further sensitivity experiments are performed to assess the effects of 3‐D CHL structure using the OGCM with the prescribed CHL fields taken from the control runs: CHLclim and normalCnormalHnormalLnormalcnormallnormalinormalmnormalsnormalunormalrnormalf ${\mathrm{C}\mathrm{H}\mathrm{L}}_{\mathrm{c}\mathrm{l}\mathrm{i}\mathrm{m}}^{\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{f}}$ are runs in which climatological DCM effect is included or only surface CHL effect is included; CHLinter and normalCnormalHnormalLnormalinormalnnormaltnormalenormalrnormalsnormalunormalrnormalf ${\mathrm{C}\mathrm{H}\mathrm{L}}_{\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}}^{\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{f}}$ are runs with interannually varying DCM effects included or not. The differences in the simulated ocean conditions are analyzed to explore DCM effects on sea surface temperature (SST) and amplitude of El Niño‐Southern Oscillation (ENSO). Two competing mechanisms responsible for the DCM effects are revealed: an ocean biology‐induced direct heating (OBH) effect, and an indirect cooling effect due to dynamic processes associated with vertical mixing and shallow meridional overturning circulation. There are three major findings: (a) DCM acts to reduce mean SST by around 0.2°C in the eastern equatorial Pacific, being larger than the surface CHL effects. (b) DCM interannual variability increases the ENSO amplitude to a comparable degree as the surface CHL effects. (c) The total net impact of vertical mixing, currents, and net surface heat flux makes SST drop more under the DCM effect than the surface CHL effect in the eastern Pacific. These findings provide a new insight into the feedback mechanisms for the bioclimate interactions.

show abstract

Attribution and Predictability of Climate‐Driven Variability in Global Ocean Color

Cited by 14 publications

References 115 publications

A Multi-Mode Convolutional Neural Network to reconstruct satellite-derived chlorophyll-a time series in the global ocean from physical drivers

A Multi-Mode Convolutional Neural Network to reconstruct satellite-derived chlorophyll-a time series in the global ocean from physical drivers

Teleconnection Between Coastal Phytoplankton Blooms Phenomenon in Western North Pacific and El Niño–Southern Oscillation by Time‐Frequency Analysis

Impact of the Deep Chlorophyll Maximum in the Equatorial Pacific as Revealed in a Coupled Ocean GCM‐Ecosystem Model

Contact Info

Product

Resources

About