Retrieving the evolution of vertical profiles of Chlorophyll-a from satellite observations using Hidden Markov Models and Self-Organizing Topological Maps

Charantonis, Anastase Alexandre; Badran, Fouad; Thiria, Sylvie

doi:10.1016/j.rse.2015.03.019

Cited by 39 publications

(27 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our case, these are the vertical profiles of silts and sand concentrations extracted from a three-dimensional (3D) hydro-sedimentary numerical model. However, the PROFHMM approach by Charantonis et al [19] retained in the present study does not use direct outputs of the oceanographic numerical model. It instead uses inferred statistical properties in terms of conditional probabilities between two consecutive time steps and between hidden and observable data.…”

Section: Methodsmentioning

confidence: 99%

“…These observations can be provided either by sensors or by numerical models and are also assumed to be noisy or partly erroneous. The method PROFHMM works on transition and emission probabilities between hidden data, which corresponds to the unknown parameters (Section 3.1), and observable classes, which correspond to forcing parameters (Section 3.2) [19]. For this purpose, we have classified these hidden and observable data using SOMs [36] (Section 3.3).…”

Section: Methodsmentioning

confidence: 99%

“…Potential benefits of these methods include filling data gaps caused by cloud cover, inferring vertical distributions of biogeochemical variables observed at sea surface, emulate models of the upper ocean dynamics at high spatial resolution and developing relationships between environmental variables [9][10][11][12]. Further applications can be found in carbon dynamics [13][14][15], coastal turbidity [16,17], phytoplankton blooms [18][19][20][21], harmful algal blooms [22,23], and primary production [24].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Construction of Multi-Year Time-Series Profiles of Suspended Particulate Inorganic Matter Concentrations Using Machine Learning Approach

et al. 2017

View full text Add to dashboard Cite

Hydro-sedimentary numerical models have been widely employed to derive suspended particulate matter (SPM) concentrations in coastal and estuarine waters. These hydro-sedimentary models are computationally and technically expensive in nature. Here we have used a computationally less-expensive, well-established methodology of self-organizing maps (SOMs) along with a hidden Markov model (HMM) to derive profiles of suspended particulate inorganic matter (SPIM). The concept of the proposed work is to benefit from all available data sets through the use of fusion methods and machine learning approaches that are able to process a growing amount of available data. This approach is applied to two different data sets entitled "Hidden" and "Observable". The hidden data are composed of 15 months (27 September 2007 to 30 December 2008) of hourly SPIM profiles extracted from the Regional Ocean Modeling System (ROMS). The observable data include forcing parameter variables such as significant wave heights (Hs and Hs50 (50 days)) from the Wavewatch 3-HOMERE database and barotropic currents (Ubar and Vbar) from the Iberian-Biscay-Irish (IBI) reanalysis data. These observable data integrate hourly surface samples from 1 February 2002 to 31 December 2012. The time-series profiles of the SPIM have been derived from four different stations in the English Channel by considering 15 months of output hidden data from the ROMS as a statistical representation of the ocean for ≈11 years. The derived SPIM profiles clearly show seasonal and tidal fluctuations in accordance with the parent numerical model output. The surface SPIM concentrations of the derived model have been validated with satellite remote sensing data. The time series of the modeled SPIM and satellite-derived SPIM show similar seasonal fluctuations. The ranges of concentrations for the four stations are also in good agreement with the corresponding satellite data. The high accuracy of the estimated 25 h average surface SPIM concentrations (normalized root-mean-square error-NRMSE of less than 16%) is the first step in demonstrating the robustness of the method.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Construction of Multi-Year Time-Series Profiles of Suspended Particulate Inorganic Matter Concentrations Using Machine Learning Approach

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Salinity profiles can be estimated from surface satellite observations using a generalized regression neural network with the fruit fly optimization algorithm method [27]. The vertical profiles of chlorophyll-a can be retrieved from satellite observations using hidden Markov models and self-organizing topological maps [28]. An objective algorithm was proposed to reconstruct the 3D ocean temperature field based on Argo profiles and SST data [29].…”

Section: Introductionmentioning

confidence: 99%

Estimating Subsurface Thermohaline Structure of the Global Ocean Using Surface Remote Sensing Observations

Yang

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Retrieving multi-temporal and large-scale thermohaline structure information of the interior of the global ocean based on surface satellite observations is important for understanding the complex and multidimensional dynamic processes within the ocean. This study proposes a new ensemble learning algorithm, extreme gradient boosting (XGBoost), for retrieving subsurface thermohaline anomalies, including the subsurface temperature anomaly (STA) and the subsurface salinity anomaly (SSA), in the upper 2000 m of the global ocean. The model combines surface satellite observations and in situ Argo data for estimation, and uses root-mean-square error (RMSE), normalized root-mean-square error (NRMSE), and R2 as accuracy evaluations. The results show that the proposed XGBoost model can easily retrieve subsurface thermohaline anomalies and outperforms the gradient boosting decision tree (GBDT) model. The XGBoost model had good performance with average R2 values of 0.69 and 0.54, and average NRMSE values of 0.035 and 0.042, for STA and SSA estimations, respectively. The thermohaline anomaly patterns presented obvious seasonal variation signals in the upper layers (the upper 500 m); however, these signals became weaker as the depth increased. The model performance fluctuated, with the best performance in October (autumn) for both STA and SSA, and the lowest accuracy occurred in January (winter) for STA and April (spring) for SSA. The STA estimation error mainly occurred in the El Niño-Southern Oscillation (ENSO) region in the upper ocean and the boundary of the ocean basins in the deeper ocean; meanwhile, the SSA estimation error presented a relatively even distribution. The wind speed anomalies, including the u and v components, contributed more to the XGBoost model for both STA and SSA estimations than the other surface parameters; however, its importance at deeper layers decreased and the contributions of the other parameters increased. This study provides an effective remote sensing technique for subsurface thermohaline estimations and further promotes long-term remote sensing reconstructions of internal ocean parameters.

show abstract

“…Among others, it has been used in climate sciences (Cavazos et al, 2002), in genetics, working with DNA sequences (Nikkilä et al, 2002), or ecological sciences applications (Chon, 2011). In the physical and biological oceanography context SOM has also been used in several studies (Charantonis et al, 2015;Liu and Weisberg, 2005;Liu et al, 2016;Richardson et al, 2003;Hales et al, 2012;Hernández-Carrasco and Orfila, 2018b). However, to our knowledge, applications of SOM analysis on the 20 reconstruction of HF radar velocity fields have not been addressed.…”

mentioning

confidence: 99%

Impact of HF radar currents gap-filling methodologies on the Lagrangian assessment of coastal dynamics

Hernández‐Carrasco

Solabarrieta

Rubio

et al. 2018

Preprint

View full text Add to dashboard Cite

Abstract. In coastal basins HF radar, HFR, is a cost-effective monitoring technique that allows to obtain high-resolution continuous surface currents, providing new insights for understanding small-scale coastal ocean transport and dispersion processes. In the last years the use of Lagrangian diagnosis to study mixing and transport properties is growing in importance. A common condition among all the Lagrangian techniques is that complete spatial and temporal velocity data is required to compute the trajectories of virtual particles in the flow. However, hardware or software failures in HFR can compromise the availability of data, resulting in incomplete spatial coverage fields or non-data periods. In this regard, several methods have been widely used to fill spatio-temporal gaps in HFR measurements. Despite the growing relevance of these systems there are still many open questions concerning the reliability of the gap-filling methods for the Lagrangian assessment of the coastal ocean dynamics. In this paper, we first develop a new methodology to fill gaps in the HFR velocity field based on Self-Organizing Maps (SOM). Then a comparative analysis of this method with the most extended available methods for gap-filling in HFR systems, i.e., Open-Boundary Modal Analysis (OMA) and Data Interpolating Empirical Orthogonal Functions (DINEOF), is performed. The error introduced by each approach is quantified in the Lagrangian computations, i.e., finite size Lyapunov exponents, Lagrangian Coherent Structures and Residence Times. We determine the limit of applicability of each method regarding four experiments based on the typical temporal and spatial gaps distribution observed in HFR systems unveiled by a k-means clustering analysis. The data sets used for this study are the hourly radial velocities obtained from the HFR system located in the SE Bay of Biscay, with a spatial resolution of 5 km in range and 5º in angle. Our, results, show that even when the number of missing data is large, the Lagrangian diagnosis still give an accurate description of the oceanic transport properties.

show abstract

Retrieving the evolution of vertical profiles of Chlorophyll-a from satellite observations using Hidden Markov Models and Self-Organizing Topological Maps

Cited by 39 publications

References 17 publications

Construction of Multi-Year Time-Series Profiles of Suspended Particulate Inorganic Matter Concentrations Using Machine Learning Approach

Construction of Multi-Year Time-Series Profiles of Suspended Particulate Inorganic Matter Concentrations Using Machine Learning Approach

Estimating Subsurface Thermohaline Structure of the Global Ocean Using Surface Remote Sensing Observations

Impact of HF radar currents gap-filling methodologies on the Lagrangian assessment of coastal dynamics

Contact Info

Product

Resources

About