Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
8867818 | International Journal of Applied Earth Observation and Geoinformation | 2018 | 11 Pages |
Abstract
Initially, we ran a numerical ocean model to simulate the movement of hypothetical green tides for last 20 years. The model was driven by remote sensing data of sea surface winds, surface temperatures, and tracers representing macroalgae that were created on certain dates so that drift paths could be traced. Ocean color remote sensing data were employed to determine the drift parameters. Next, the relationship between the displacement of tracers, including directions and distances of movement during certain periods were then analyzed along with the corresponding values of a set of six climate indices. A forecasting algorithm based on an artificial neural network was then established and trained with these data. Using this algorithm, the drift path of green tides could be predicted from the values of certain climate indices of the previous year. The model assessment with satellite ocean color remote sensing images indicated the effectiveness and practicability of this method.
Related Topics
Physical Sciences and Engineering
Earth and Planetary Sciences
Computers in Earth Sciences
Authors
Po Hu, Yahao Liu, Yijun Hou, Yuqi Yin,