September 2025

Oceanography | Vol. 38, No. 3

86

agreement. The non-​commercial unit sampled at a faster rate. In

general, most of the drifters followed similar patterns. The drifter

built by the instructors had a drogue depth of 15 m, which may

explain its different trajectory at the start of the release. The drifter

from Team 3 had a larger surface area protruding out of the water,

which may explain why it deviated from the other trajectories,

potentially being more strongly carried by wind.

Figure 7b,c shows the location of the drifter from Team 1 with

the non-commercial GPS at three timestamps. Figure 7b shows

the position over a contour plot of the bathymetry with an orange

arrow denoting the wind speed and direction and the tide height

listed in the panel header. In Figure 7c, the position is shown over

the flow field obtained from the +Atlantic CoLAB MOHID model.

While the temporal resolution of the model is low (one snapshot

every three hours), it is still a helpful tool for estimating the trajec-

tory of the drifter and teaching students about data assimilation.

POSSIBLE MODIFICATIONS

COURSE MODIFICATIONS

Most student designs were variations of the example drifter intro-

duced during the course. To encourage greater diversity in design,

we recommend presenting multiple example drifters. Additionally,

we suggest documenting and evaluating how design differences

affect water-following performance. Increasing testing opportuni-

ties at the deployment site would also be beneficial; for instance,

teams could deploy multiple drifters over several days to better

track current patterns. To further engage students, they could be

challenged to select their own deployment locations and develop,

then validate, current models using their drifter data (Champenois

et al., 2025). Finally, the curriculum could be enhanced by includ-

ing a lesson focused on assessing and quantifying uncertainty in

both measurements and models.

DRIFTER DESIGN MODIFICATIONS

Incorporating environmental sensors, such as those measuring

salinity, temperature, and pH, into the drifter’s custom electronic

stack can capture a broader range of oceanographic data, which

helps contextualize drifter trajectories and contributes to under-

standing ocean processes. Switching to a commercial satellite-​

based GPS tracker removes the 4G network range limitation, pro-

viding much larger coverage at a higher cost. Achieving longer

deployments (>1 week) would require a significant redesign, focus-

ing on a more durable frame and drogue, extended battery life, and

a robust communication system. Because this would involve more

permanent materials and a higher cost, drifter retrieval should also

be considered.

OCEAN DRIFTER DEVELOPMENT

AS A TEACHING TOOL

We developed and tested a curriculum to teach undergraduates

and graduate students in oceanography and engineering about

ocean sensor design, ocean sensing, and ocean hydrodynamics.

The project challenged students to design, build, and deploy an

ocean drifter for measuring near-surface ocean currents, as well

as analyze and compare the measured data. The hands-on curric-

ulum was paired with lectures on oceanography, marine robot-

ics, and ocean monitoring. Understanding the movement of drift-

ers required students to engage with the Lagrangian perspective, in

which sensors follow the flow, as they tracked drifter trajectories

to study how ocean currents transport material, in contrast to the

Eulerian perspective that observes stationary points. This hands-on

experience helped them explore real-world processes such as dis-

persion and mixing in coastal environments—concepts central

to physical oceanography.

Students faced several challenges, including limited oppor-

tunities for at-sea testing and deployment due to weather and

ocean conditions. Additionally, limitation in materials and tools

restricted each team to building just one drifter. This single deploy-

ment meant students launched their final designs under condi-

tions that were different from any prior testing. Specifically, winds

and currents were stronger at the deployment location than they

were in the protected harbor used for initial testing. This project

revealed several challenges with remote sensing in marine envi-

ronments. Drifters with higher buoyancy tended to maintain GPS

signal transmission longer by better protecting electronics from

wave impacts. While most drifters followed similar trajectories,

those with deeper drogues or larger surfaces exposed to wind

diverged, illustrating design impacts on current-following perfor-

mance. The custom GPS units may have failed due to degradation

of the water-resistant shellac coatings and coconut wax potting

from saltwater and wind erosion.

The construction and design process was largely student-driven,

with instructors checking in regularly to ensure that teams stayed

on track. For many students, this was their first time deploying

equipment at sea. Instructors provided support by guiding them

through the deployment process, logistics, and design consider-

ations for launching from a boat. In the second week, an addi-

tional lecture on ocean modeling was introduced, which enabled

students to analyze how their drifters’ movements were influenced

by ocean phenomena.

The design challenge proved to be an engaging and educational

experience for the students. In the course evaluation, one student

shared: “It was a great experience to learn material outside of my

degree classes. I feel like I have a much more holistic understand-

ing of marine robotics now that I understand oceanography and

some marine biology.”

REFERENCES

Anderson, T. 2015. Educational experiences in oceanography through hands-on

involvement with surface drifters: An introduction to ocean currents, engineering,

data collection, and computer science. Paper presented at the fall meeting of the

American Geophysical Union, December 14–18, 2015, San Francisco, California,

abstract ED21A–0823, https://ui.adsabs.harvard.edu/abs/2015AGUFMED21A0823A/

abstract.

Beardsley, R.C., and S.J. Lentz. 1987. The Coastal Ocean Dynamics Experiment collec-

tion: An introduction. Journal of Geophysical Research: Oceans 92(C2):1,455–1,463,

https://doi.org/10.1029/JC092iC02p01455.