A review of cloud computing frameworks and distributed systems in remote sensing data processing for marine pollution monitoring.

A review of cloud computing frameworks and distributed systems in remote sensing data processing for marine pollution monitoring.

Negin Darzi1 Milad Tashakori2 Mohsen Falah rad3

1) PhD student in Computer Engineering, Islamic Azad University, Lahijan Branch, Iran -
2) Expert in charge of coordination, Water and Wastewater Company of Gilan Province, Rasht, Iran -
3) Assistant Professor, Department of Computer and Information Technology, Islamic Azad University, Lahijan Branch, Iran -

IOI Code : AAHD-ACZZZG
Publication : National Conference on Information Technology, Nanotechnology, Artificial Intelligence and Technological Futures Studies(itnaf.ir)
Abstract :
    In recent decades, marine pollution monitoring has shifted from traditional local processing approaches to advanced cloud computing infrastructures and distributed systems, driven by the explosive growth of satellite data volumes and the increasing complexity of environmental modeling. This review article analyzes and systematically examines the role of powerful cloud platforms (such as Google Earth Engine and AWS) and their integration with modern artificial intelligence algorithms (such as random forests, convolutional neural networks, and deep models) in processing multispectral and hyperspectral data. The main goal of this analysis is to explain the effectiveness of these frameworks in accurately monitoring key marine pollutants including algal blooms, chlorophyll-a, suspended solids, turbidity, and oil spill detection in large water areas. The findings of this study show that cloud architectures have not only overcome the hardware, memory, and time constraints in the analysis of long-term satellite time series, but also significantly improved the accuracy of pollutant identification and the speed of water quality monitoring map generation by providing a platform for online and scalable processing. This study, while extracting successful patterns in integrating multi-source data and analyzing the effectiveness of distributed architectures, has explained the technical gaps in the scalability of complex algorithms for operational monitoring at national and regional levels. Finally, this review provides a knowledge infrastructure for the transition from static observation to dynamic monitoring and management decision-making based on big data by drawing future paths for the development of intelligent, automated, and reliable systems in marine environmental risk management.

Keywords : artificial intelligence cloud computing remote sensing water quality monitoring satellite data fusion