EGU22-7197
https://doi.org/10.5194/egusphere-egu22-7197
EGU General Assembly 2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Dynamics of the euphotic zone in the Black Sea: The synergy of data from profiling floats, machine learning and numerical modeling 

Emil Stanev, Kathrin Wahle, and Joanna Staneva
Emil Stanev et al.
  • Helmholtz-Zentrum Geesthacht, Institute for Coastal Research, Geesthacht, Germany (emil.stanev@hereon.de)

Data from profiling floats in the Black Sea revealed complex temporal and spatial relationships between physical variables on the one hand and oxygen, chlorophyll (Chl-a) and the backscattering coefficient at 700 nm (bbp) on the other. It was demonstrated that the temperature-stratified upper layer and salinity-stratified layers below 50 m provide the ecological niche responsible for the major variability of the BGC system in the euphotic layer. There, around the depth of the minimum potential vorticity, the subsurface chlorophyll maximum presents a major BGC characteristics. Data analysis revealed some limits in understanding the details of BGC dynamics, as well as their dependence on physical drivers. One example is the fact that while bbp and Chl-a weakly follow the long-term changes in temperature and salinity, their responses to individual cooling events appear much stronger than what is observed in the physical fields. To fully account for different interdependences, a feedforward backpropagation neural network (NN) was used. The NN learnes from data recorded by profiling floats and predicts BGC states using physical measurements only. The skill was very high, particularly for oxygen, but it reduced when the NN was applied to older data because the interrelationships between the physical and BGC properties have changed recently. One indication of such change is the missing overshooting of either Chl-a or bbp penetration depth in winter reported in earlier studies. The BGC states reconstructed by the NN from physical data produced by a coupled physical-BGC operational model outperform the BGC output of the same coupled model. Therefore, the use of data from profiling floats, physical properties from numerical models and NN appeared a powerful tool to reconstruct the 4D dynamics of euphotic zone. Basin wide patterns and temporal variabilities of oxygen, bbp and Chl-a are also analyzed. Of particular interest is the reconstruction of short-living BGC features, particularly in the area of coastal anticyclones, which are difficult to observe basin-wide with available floats.

How to cite: Stanev, E., Wahle, K., and Staneva, J.: Dynamics of the euphotic zone in the Black Sea: The synergy of data from profiling floats, machine learning and numerical modeling , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7197, https://doi.org/10.5194/egusphere-egu22-7197, 2022.