Abstract
Subsurface chlorophyll maxima (SCMs), commonly occurring beneath the surface mixed layer in coastal seas and open oceans, account for main changes in depth-integrated primary production and hence significantly contribute to the global carbon cycle. To fill the gap of previous methods (in situ measurement, remote sensing, and the extrapolating function based on surface-ocean data) for obtaining SCM characteristics (intensity, depth, and thickness), we developed an improved deep neural network (IDNN) model using a Gaussian radial basis activation function to retrieve the vertical profile of chlorophyll a concentration (Chl a) and associated SCM characteristics from surface-ocean data. The annually averaged SCM depth was further incorporated into the bias term and the Gaussian activation function to improve the estimation accuracy of the IDNN model. Based on the Biogeochemical-Argo (BGC-Argo) data acquired for three regions in the northwestern Pacific Ocean, vertical Chl a profiles produced by our improved DNN model using sea surface Chl a and sea surface temperature (SST) were in good agreement with the observations, especially in regions with low surface Chl a. Compared to other neural-network-based models with one hidden layer and a sigmoid activation function, the IDNN model retrieved vertical Chl a profiles well in more eutrophic subpolar regions. Furthermore, the application of the IDNN model to infer vertical Chl a profiles from remote-sensing information was validated in the northwestern Pacific Ocean.
Highlights
Licensee MDPI, Basel, Switzerland.The ocean plays an important role in the global carbon cycle, with marine phytoplankton accounting for ~50% of the global primary production [1]
We develop a deep neural network (DNN) model to retrieve the vertical profiles of Chl a from surface-ocean data which are equivalent to the average value within the first 20 m water depth [4]
A DNN model with at least two hidden layers was applied due to its a DNN model with at least two hidden layers was applied due to its availability in capturing the nonlinear relationships, and the annual averaged subsurface chlorophyll maximum (SCM) depth availability in capturing the nonlinear relationships, and the annual averaged SCM depth is incorporated into both the bias term and the Gaussian radial basis activation function is incorporated into both the bias term and the Gaussian radial basis activation function to improve the capability in retrieving the SCM characteristics
Summary
Sammartino et al [24] trained an artificial neural network (ANN) model with one hidden layer ( known as a shallow ANN) to infer Chl a vertical profiles in the Mediterranean Sea via in situ measurements together with remote-sensing data (Chl a and temperature). Because previous neural-network-based models used only one genu in the hidden layer with the sigmoid activation function, their accuracy needs to improve for the vertical Chl a profiles. We develop a deep neural network (DNN) model to retrieve the vertical profiles of Chl a from surface-ocean data which are equivalent to the average value within the first 20 m water depth [4]. SCM characteristics of our DNN model in retrieving vertical Chl a profiles from remote-sensing data inwith the observations different regions and seasons.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
