NSF Award #1429940 set out to develop a new application to improve the measurement of biological gases, like carbon dioxide and oxygen, that are dissolved in seawater. These biological gases are important to the ocean ecosystem and to the habitability of earth. The new application utilizes a gas measuring instrument, the SWIMS underwater mass spectrometer, to make an inventory of gases dissolved in the upper ocean. The SWIMS is unique because it can tolerate water depths up to 2000 m (~5500 ft), and it measures gases in the water every 2 seconds, creating a very detailed picture. We mounted the SWIMS onto a vehicle that can be towed behind a ship, while undulating in a yo-yo fashion between the ocean surface and 200 m (~700 ft) of water depth (see Primary Image for a picture of the SWIMS and tow vehicle).
The goal of this application is to estimate the net (gain-loss) production of oxygen in the region of the ocean that receives sunlight. However, the SWIMS is highly influenced by the changes that happen during towing, namely the changes in water temperature and water pressure. These influences make the oxygen estimates inaccurate, until the instrumental effects are accounted for. We developed a correction using machine learning methods, similar to the methods made famous for voice and picture recognition, as artificial intelligence. These methods are able to recognize how the SWIMS will be influenced by seawater temperature and pressure, and we use them to improve the accuracy by correcting for the predicted instrumental effects (see Image #3, which shows the workflow for correcting the SWIMS measurement of oxygen in seawater).
The method has produced high resolution maps of oxygen and other biogenic gases in the upper ocean; here we show a section of oxygen that was measured while crossing the Gulf Stream (see Image #2, map). The image of oxygen in the upper ocean reveals a region of exceedingly low oxygen, directly beneath the Gulf Stream, where two water currents converge. These structures reveal how water motion and other properties influence the distribution of oxygen in seawater. The work to estimate net production of oxygen using the SWIMS data is ongoing, and the results will be published in the public domain under Open Access license.
Participants in this project, also developed an interactive engagement entitled "Carbon, oxygen and our breathing ocean" that demonstrates how the ocean behaves as a "reverse lung", compared to the human lung. The display includes a carbon dioxide analyzer to measure the amount of carbon dioxide in human breath. It also includes a video and fluid dynamics tank that illustrates how the ocean can inhale carbon dioxide through its overturning currents, and exhale oxygen through the action of aquatic photosynthesizing organisms. This display is popular and is used during week-long science engagements in Newport, RI as well as the annual Open House at the URI Bay Campus.
Last Modified: 07/06/2020
Modified by: Brice Loose
| Dataset | Latest Version Date | Current State |
|---|---|---|
| CTD and underwater mass spectrometer data acquired in 2017 aboard the RV/Endeavor (cruise EN602) in the the subtropical Atlantic using a Triaxus tow vehicle. | 2021-03-04 | Final no updates expected |
| CTD and underwater mass spectrometer data acquired in 2016 aboard the RV/Endeavor (cruise EN575) in the the subtropical Atlantic using a Triaxus tow vehicle. | 2021-03-03 | Final no updates expected |
Principal Investigator: Brice Loose (University of Rhode Island)