Abstract

Polar regions have the most productive ecosystems in the global ocean but are vulnerable to global climate changes. Traditionally, the long-term changes occurred in an ecosystem are studied by using satellite-derived estimates of passive ocean color remote sensing measurements. However, this technology is severely limited by the inability to observe high-latitude ocean areas during lengthy polar nights. The spaceborne lidar can address the limitations and provide a decade of uninterrupted polar observations. This paper presents an innovative feed-forward neural network (FFNN) model for the inversion of subsurface particulate backscatter coefficients (bbp), chlorophyll concentration (Chl), and total particulate organic carbon (POC) from the spaceborne lidar. Non-linear relationship between lidar signal and bio-optical parameters was estimated through FFNN. The inversion results are in good agreement with biogeochemical Argo data, indicating the accuracy of the method. The annual cycles of Chl and POC were then analyzed based on the inversion results. We find that Chl, bbp, and POC have similar interannual variability but there are some subtle differences between them. Light limitation appears to be a dominant factor controlling phytoplankton growth in polar regions according to the results. Overall, the combined analysis of bbp, Chl, and POC contributes to a comprehensive understanding of interannual variability in the ecosystem in polar regions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call