A new approach to aquatic ecosystem monitoring of the Ob river

Aleksandr Andreevich Tskhai,Vladimir Viktorovich Kirillov,Aleksandr Vasilievich Puzanov,Nelley Mikhailovna Kovalevskaya

doi:10.5194/piahs-383-375-2020

Aleksandr Andreevich Tskhai, Vladimir Viktorovich Kirillov + Show 2 more

Open Access

https://doi.org/10.5194/piahs-383-375-2020

Copy DOI

Abstract

Abstract. An integrated approach to aquatic ecosystem monitoring is being developed for the Ob river. The novelty of the approach is in transition to high-performance computing (HPC) under processing of remote sensing data and modeling of aquatic ecosystem degradation. Classes of optically complex surface waters of the Gulf of the Kara Sea – the Ob Bay are studied on the basis of the fuzzy cluster analysis of water objects emission. A ten-year dynamics of concentrations of optically active substances was estimated using the neural network simulation. The application of the structural-dynamic model allowed explaining the long-term trend in the development of the ecosystem of the Novosibirsk reservoir, the largest one in the Ob river basin.

Highlights

It includes (1) effective neural models for processing satellite data on ecosystem component characteristics; (2) advanced algorithms for big remote sensing data management; (3) original structural-dynamic adaptive models of biogeochemical cycles of transformation of limiting elements able to reconstruct the development of hydrobionts communities when the data are incomplete and uncertain
The noted earlier tendency (Gaevsky et al, 2009) in phytoplankton decrease is observed as it moves from the Ob delta towards the Kara Sea (Fig. 6)
This is used for the neural model validation as a confirmation of the zonal CHL distribution depending on a salinity gradient and a biogenic element concentration (Sukhanova et al, 2010) in case of desalinated and frontal zones as well as a shallow shelf occurrence

Summary

Introduction

It includes (1) effective neural models for processing satellite data on ecosystem component characteristics; (2) advanced algorithms for big remote sensing data management; (3) original structural-dynamic adaptive models of biogeochemical cycles of transformation of limiting elements able to reconstruct the development of hydrobionts communities when the data are incomplete and uncertain. Such components ensure the implementation of the new approach to real-time environmental monitoring

Methods and materials

Results and discussion

Summary