How DeepSee Helps Scientists Plan and Adapt in the Deep Ocean

Table of Links

Abstract and 1 Introduction

2 Related Work

3 Methodology

4 Studying Deep Ocean Ecosystem and 4.1 Deep Ocean Research Goals

4.2 Workflow and Data

4.3 Design Challenges and User Tasks

5 The DeepSea System

• 5.1 Map View
• 5.2 Core View

5.3 Interpolation View and 5.4 Implementation

6 Usage Scenarios and 6.1 Scenario: Pre-Cruise Planning

• 6.2 Scenario: On-the-Fly Decision-Making

7 Evaluation and 7.1 Cruise Deployment

7.2 Expert Interviews

7.3 Limitations

7.4 Lessons Learned

8 Conclusions and Future Work, Acknowledgments, and References

6 USAGE SCENARIOS

In this section, we illustrate how DeepSee can be used to study deep ocean microbial ecology via two example usage scenarios (Fig. 6).

6.1 Scenario: Pre-Cruise Planning

A graduate researcher is planning a follow-up cruise with their lab to collect cores at a site visited previously. Based on prior sampling, they want to determine where cores with a high abundance of both ANME-2c archaea and JS1 bacteria might be (Fig. 6A).

\ They start by selecting all cores in the Map View and “ANME-2c” in the Core View, then examining the ANME-2c distribution down each core (T4). Using the Core View instead of the filters in the Map View, they locate four cores with high ANME-2c abundance (T2, T8). They select these cores in the Map View (T6) and find them clustered in pairs of two. Wondering about the interactions between ANME-2c and JS1, they select “JS1” in the Core View and notice that the two cores to the North also have a high abundance of JS1 (T1). Using the menus of the Map View to load a photomosaic map of seafloor imagery on the fly while maintaining the current viewport, they note that this region has a white microbial mat where the cores were taken (T5). They annotate the Map View by drawing circles around the two cores to the North and write a short message: “Area of interest #1” (T3). They then screenshot the image and place the image in a document for future reference (T9).

\ The researcher decides that for the upcoming dive, retrieving cores from areas near these two cores with ANME-2c and JS1 are a high priority for their project, and focus should be on areas with a white microbial mat visible on the seafloor in this region. DeepSee enabled the researcher to maximize the value of limited samples by integrating several different data types (maps, tables, annotations) to make informed decisions about where to sample in the future.

\

6.2 Scenario: On-the-Fly Decision-Making

A chief scientist (PI) is currently on a cruise, working with a team of submersible operators to visit their next sampling location. Unfortunately, once on the seafloor, they discover the site no longer hosts the characteristic white and orange microbial mats and adjacent chemosynthetic clams observed a few years prior. The chief scientist must quickly select another nearby seep sampling site with a high probability of similar microbial composition to collect sediment cores (Fig. 6B).

\ Using DeepSee, the PI quickly loads a photomosaic map in the Map View from the last cruise to this site (T9) and uses the Map View to filter for the local cluster of previous cores collected here (T6). Looking at the spread of previous cores and the photomosaic map together, they realize the submersible is at the far end of the previously sampled area, which is ≈ 50 meters from an area that may still feature microbial activity (T5, T7). To test this hypothesis, they select all of the cores and choose “Sulfide” in the Core View, identifying which cores have measured sulfide gradients which act as a proxy for active seeps (T2, T10). The PI eliminates cores without sulfide concentration data and identifies several cores that span the width of a previous sampling area (T8, T9). Switching to the Interpolation View, they select a natural neighbor interpolation at 77 cm resolution and a VSUP color palette [11]. The PI sees that there is a predicted sharp subsurface gradient towards higher sulfide concentrations heading northeast across the previous sampling area (T11). Switching the predicted attribute to microbial taxa identified at the site, they also observe a similar trend in the relative abundance of a methane-oxidizing archaeon, which are found in areas of methane seepage (T1, T2, T11). Throughout, they annotate the Map View to support their decision-making, document the process, and share the results with the team (T3, T8).

\ The PI decides that the submersible should navigate northwest to the sampled area and sample at, if not past, the location of core collections from the previous cruise located in the region with the highest sulfide concentrations. Having DeepSee’s ability to visualize unseen gradients quickly with limited data in the field can help researchers to make smarter, more data-driven decisions.

\ \

\ \ \

:::info This paper is available on arxiv under CC BY 4.0 DEED license.

:::

:::info Authors:

(1) Adam Coscia, Georgia Institute of Technology, Atlanta, Georgia, USA ([email protected]);

(2) Haley M. Sapers, Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California, USA ([email protected]);

(3) Noah Deutsch, Harvard University Cambridge, Massachusetts, USA ([email protected]);

(4) Malika Khurana, The New York Times Company, New York, New York, USA ([email protected]);

(5) John S. Magyar, Division of Geological and Planetary Sciences, California Institute of Technology Pasadena, California, USA ([email protected]);

(6) Sergio A. Parra, Division of Geological and Planetary Sciences, California Institute of Technology Pasadena, California, USA ([email protected]);

(7) Daniel R. Utter, [email protected] Division of Geological and Planetary Sciences, California Institute of Technology Pasadena, California, USA ([email protected]);

(8) John S. Magyar, Division of Geological and Planetary Sciences, California Institute of Technology Pasadena, California, USA ([email protected]);

(9) David W. Caress, Monterey Bay Aquarium Research Institute, Moss Landing, California, USA ([email protected]);

(10) Eric J. Martin Jennifer B. Paduan Monterey Bay Aquarium Research Institute, Moss Landing, California, USA ([email protected]);

(11) Jennifer B. Paduan, Monterey Bay Aquarium Research Institute, Moss Landing, California, USA ([email protected]);

(12) Maggie Hendrie, ArtCenter College of Design, Pasadena, California, USA ([email protected]);

(13) Santiago Lombeyda, California Institute of Technology, Pasadena, California, USA ([email protected]);

(14) Hillary Mushkin, California Institute of Technology, Pasadena, California, USA ([email protected]);

(15) Alex Endert, Georgia Institute of Technology, Atlanta, Georgia, USA ([email protected]);

(16) Scott Davidoff, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA ([email protected]);

(17) Victoria J. Orphan, Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California, USA ([email protected]).

:::

\