Multi-sensor Satellite Image Fusion, Data Merging, and Machine Learning For Monitoring Changing Earth Environment

The Earth's total surface area is made up of various flow regimes of reservoirs, bays and lakes as well as soil environment, which are considered important natural resources for the maintenance of ecosystem integrity and human consumption. As the condition of environment deteriorates throughout the world, it necessitates the scientific work of monitoring environmental quality in response to its dynamic changes of quality status or flow conditions and feedbacks to our society. For this purpose, satellite remote sensing techniques with multiple in-situ ground-based sensors may be applied to collectively capture a much larger spatial coverage within relatively short time periods through various traditional or non-traditional algorithms. To improve the overall efficiency there is a tradeoff in spectral, spatial and temporal resolution of different sensors when monitoring the water, soil, and air pollutants in the changing environment. The goal of this presentation is to introduce the latest forefronts in the field and demonstrate green, smart, and sustainable management of our changing Earth environment by integrating multi-sensor satellite image fusion, data merging, and machine learning – an emerging area of importance in systems engineering. The following scientific questions are explored in this study: (1) Are fused image reflectance bands and machine-learning techniques able to accurately carry out the estimation of target environmental quality parameters under different challenges? (2) Is it feasible to have an integrative and innovative process for updating the environmental condition for early warning? (3) How microwave signals work with optical remote sensing for monitoring changing earth environment?