Estimating the carbon content of oceans using satellite sensor data
Abstract The impact of chemical processes in ocean surface waters is far-reaching. Recently, increased significance has been placed on the concentration of Carbon and its compounds and the effects these may have on climate change. Remote-sensing enables near real-time measurement of key sea-surface...
| Published in: | Journal of Big Data |
|---|---|
| Main Authors: | , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
SpringerOpen
2022
|
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s40537-022-00647-7 https://doaj.org/article/93446a72b170464c8f4d20f0eb92f626 |
| Summary: | Abstract The impact of chemical processes in ocean surface waters is far-reaching. Recently, increased significance has been placed on the concentration of Carbon and its compounds and the effects these may have on climate change. Remote-sensing enables near real-time measurement of key sea-surface data which can be used to estimate Carbon levels. We illustrate with the use of hybrid Satellite sensor data. To validate our results we use data collected from cruise ships as the ground truth when training our algorithms. The error rate of our predictor is found to be small and hence the proposed approach can be used to estimate Carbon levels in any ocean. This work improves upon previous research in many ways including the use of sea water salinity as a proxy for Carbon estimates. Binary combinations of typically unary predictor attributes are used for the purposes of predicting the Carbon content of surface water and an inherently non-linear model is used to quantify the relationship. |
|---|