материалы конференции

En

Advances in Science, Technology and Innovation. 2nd International conference on Mediterranean Geosciences Union, MedGU 2022. Marrakech. 27 November 2022 до 30 November 2022

ISSN:25228714

2024

10.1007/978-3-031-48715-6_34

The work proposes a salinity recovery method for the Sea of Azov based on implementation of a general regression compiled from archival in situ data and regional biooptical parameters obtained from standard MODIS L2 products. The observational data from the open Internet services were acquired directly from data providers. The authors’ procedures for quality control and merging were implemented for these data. We researched the following biooptical parameters: aph(678) is the absorption coefficient by phytoplankton at 678 nm, Tchl is the sum concentration of chlorophyll-a and pheophetin-a, atot(438) is the total absorption coefficient by all optically active components at 438 nm, aCDM(438) is the absorption coefficient by colored detrital matter at 438 nm, and bbp(438) is the particulate backscattering coefficient at 438 nm. We chose these variables because they are the operational satellite ocean color products of the MODIS (NASA). Each measurement of satellite data with 1 km spatial resolution processed to spatial maps of five biooptical parameters on a regular grid of the Sea of Azov. Based on the linear regressions satisfying the condition R ≥ 0.5, general equations of the following form were compiled y = (aaver ± σ1) ⋅ x + (baver ± σ2), where aaver and baver—averaged linear coefficients a и b, σ1 and σ2—standard deviations, x—regional biooptical products, y—salinity (‰). The results of the study showed the possibility of using different approaches to building generalized empirical regressions for the spring and summer. The result is merged regressions for spring and summer designed for reconstruction of salinity and obtaining the data sets for spatiotemporal variability analysis of salinity. Average values of salinity recovered indirectly using our proposed method are within the 95% confidence bands for the long-term average seasonal trends for periods 1986–2018 and 2000–2018 from in situ data. The main result is merged regressions for spring and summer seasons designed for reconstruction of salinity and obtaining the data sets for spatiotemporal variability analysis of salinity in the Sea of Azov. In addition, the values of salinity reconstructed from aCDM(438) are found to reflect its changes most realistically within the observed salinity range (1–18‰). These results allow us to use in the future reconstructed salinity datasets in the assimilation procedures of the 3D hydrodynamic model and for retrospective salinity recovery.

Biooptical regional products, Ocean-color data, Regression analysis, Salinity, Sea of Azov

2024-04-01 11:26:48