Daniel Rudnick, Daniel Costa, Ken Johnson, Craig Lee, and Mary-Louise Timmermans
The fundamental observational problem in oceanography is sampling a global, turbulent fluid where physical, biological, and chemical processes act over a wide range of scales. Relevant length scales range from the size of ocean basins down to millimeters where turbulent dissipation occurs. Time scales of interest are as small as seconds and as large as decades or centuries. An approach to this daunting problem is to use autonomous platforms, defined here as being unconnected either to a ship or the seafloor. This approach relies on many relatively small, inexpensive platforms. The wide range of scales favors observational systems that are scalable. Intermittence and regionality require observational systems to be portable.
The notion of an observing system of small, scalable, and portable devices was the driver of the first Autonomous and Lagrangian Platforms and Sensors (ALPS) meeting in 2003. This meeting took place during a time in the early 2000s when there were several competing ideas on how to observe the ocean. Resources for observing were relatively abundant at the time, and there were many planning exercises based around the turn of the millennium. There were already a number of successes in the early 2000s, with the Global Drifter Program and the Argo profiling float array getting underway. Underwater gliders were just beginning to be used for science as opposed to engineering tests. Propeller driven autonomous underwater vehicles (AUVs) were starting to see wide use. The trend toward miniaturization was leading to sensors for a wide range of physical and biogeochemical variables. Whether by design or luck, the ALPS meeting presaged the rapid growth in autonomous observation that has fundamentally changed observational oceanography.
The ALPS-II meeting took place in early 2017, 14 years after the first ALPS meeting. Given the growth in the ALPS enterprise, the topics of interest had grown to include autonomous surface vehicles, unmanned aerial vehicles, and animal borne sensors. Applications of ALPS had also grown, especially in concert with the improvement in numerical ocean forecasts and state estimates. The topics covered in ALPS-II were thus much broader than 14 years ago. The collection of brief articles in this report reflects the breadth of discussion at the meeting.
The articles are roughly grouped into collections on ALPS Technologies, Global and Regional scientific issues, and Infrastructure. This introduction includes a distillation of the ideas about these topics derived from breakout groups at the meeting. The appendices include the workshop agenda, participants, and a list of white papers that were solicited from workshop participants prior to the meeting.
ALPS technologies include both platforms and sensors. Lagrangian platforms move with the water, including drifters that track the surface horizontal flow, and neutrally buoyant floats that are capable of three-dimensional trajectories (D’Asaro). Unmanned aerial vehicles (Reineman) and animals as platforms (Roquet and Boehme) have exploded in use in recent years, and were not considered during the original ALPS meeting in 2003. Optical sensors find special application in ALPS for biological studies as of the carbon pump (Estapa and Boss).
Lessons from the past 14 years focus around the importance of sustained observations to establish reliability. Experimental tools are often tried first in more targeted studies. Technology development for sensors must extend through quality control and data management to achieve the greatest impact.
In general, platform development has outpaced improvements in sensors. Needed investments in sensors should target Essential Ocean Variables (www.goosocean.org/eov). Devoted centers might be considered to encourage sensor development. Sensors for measuring throughout different trophic levels would contribute to marine resource management. Finally, education in the use of new sensors could be improved through summer schools or webinar series.
A major challenge for sensors is the continuing need to improve quality and accuracy. Progress requires cooperation between manufacturers and practicing scientists. This ongoing quest for improvement is sometimes not as attractive for funding, but is essential. While a fine goal is a set of standardized protocols for each sensed variable, an open question is whether this is an oversimplification or an impediment to creativity.
The use and value of ALPS on a global scale have grown significantly over the past decade. Key applications include global maps and trends of physical parameters (Gray), numerical state estimates and network design (Nguyen and Heimbach), global-scale assessments of small-scale processes (Cole), and air-sea interactions (Thomson).
The most effective employment of ALPS for global assessments requires filling regional sampling gaps. Essential undersampled areas include coastal shelves, boundary currents, polar regions, the deep ocean, the near-surface atmospheric boundary layer, and remote environments such as at ice-sheet ocean boundaries. Filling these gaps also requires higher sampling resolution for the global array in some cases, and a committed integration effort to ensure connectivity between boundary regions and the interior ocean to produce a single global data set. It is important to recognize the value of multi-platform experiments, which require making the distinction between programs (e.g., Argo) and sensor platforms (e.g., floats).
In the coming decade, global ALPS systems will be invaluable tools for event detection and resolution. For example, Argo data enabled the detection (in 2013) and monitoring of a large mass of warm water in the Pacific Ocean. Sustained systems for identification of such global anomalies will be key to understanding climate processes and making reliable projections. Adaptive sampling needs to be an important capability of ALPS platforms in the global array.
The biggest achievements with respect to global ALPS have been largely physical. There is an immediate need to extend global maps and trends to properties like biomass and inorganic carbon. Plans for biogeochemical studies on global scales (BGC-Argo; biogeochemical-argo.org) are presently being implemented. Global standards for biogeochemical sensing remain to be fully developed. In the coming decade, it is anticipated that there will be significant progress using ALPS to link biogeochemical changes to changing physics on a global scale.
Other key focus areas over the coming decade should include identifying and maintaining core parameters for global ALPS systems (e.g., the physical ocean data set is critical for continued monitoring of climate change and viable projections). Community needs should be defined for individual sensors, encompassing physical, biological, and chemical properties; for example, air-sea fluxes, waves and velocity measurements are immediate needs for global ALPS. Other focus areas should be continued improvements in data services for better accessibility of ALPS data, and robust uncertainty estimates (both for global maps and trends as well as for individual data). Novel and unanticipated uses of these global ALPS will continue to be made possible by open-access quality-controlled data. Along with essential public access to data for advancing science, there is the need to educate users by providing guidance on appropriate use and limitations. Finally, there is a continual obligation for training of early career scientists to maintain quality and reliability of data over the duration of an observational system.
Because ALPS are scalable and portable, they are uniquely suited to regional studies. The scientific and societal motivations depend on the region, as do the mix of platforms and approaches. Because the time and length scales of regional processes can cover such a wide range, a mix of platforms is often required. Among the regions considered in this report are high latitudes in both the Arctic (Timmermans et al.) and Antarctic (Purkey and Dutrieux). Shallow coastal areas are energetic and biologically active, with many ALPS technologies finding application (Nidzieko et al.). The western boundary currents that drive oceanic heat transport and eastern boundary regions where the effects of global climate variability are felt by society are targets for ALPS networks (Todd et al.). Targeted deployments of ALPS are an active component of observations for studies of hurricanes (Goni et al.).
The specific observational requirements of regions prompt the use of certain ALPS approaches. Fast, propeller-driven AUVs are ideal for the short time and space scales near coasts. Underwater gliders find special application in boundary currents, and to connect the coast and open ocean. Surface drifters are especially useful to identify circulation patterns and to quantify dispersion. Profiling floats excel at broad coverage, for example, in the equatorial region. Instrumented animals are perfect for high-density observations where the animals live. Ice-based systems are essential for collecting collocated measurements of the upper ocean, ice, and atmosphere at high latitudes.
Special challenges in regional settings revolve around the merging of data and strengths of different platforms. In this respect, data services are key to successful regional observing systems. Assimilative modeling and state estimation yield optimized fields and forecasts for research and decision-making, and assessments of network design. Local logistical issues including Exclusive Economic Zones must be respected in regional studies.
With the growth of ALPS over the last decade and a half, there are new requirements for infrastructure for support. Indeed, ALPS systems should begin to be appreciated as infrastructure as much as ships have been during the last several decades. Wynn and White present an approach to providing ALPS services as infrastructure in the UK. The massive amounts of data created by thousands of ALPS presents challenges and opportunities for data services (Zykov and Miller).
ALPS may improve observational capability in environments where resources are constrained, presenting an opportunity as well as a challenge. A key to moving forward is to broaden the user base by lowering barriers of expertise. At the same time, existing expertise must be maintained to continue progress. Improved data services would increase the use of ALPS data, creating additional justification for technological development.
Opportunities exist for educational efforts in platform and sensor use at sea, and in data analysis on land. Communities of practice must be built and supported. This is an area where cooperation between agencies may help to identify viable models and to craft pilot efforts.
With robotics a growing field, ALPS may especially benefit from focusing on partnerships between academia, government, and the private sector. With a number of private foundations focusing on the ocean and climate, new ideas for support may arise in the coming years. A future network of connected ALPS covering the global ocean and extending into societally important regions is and exciting possibility.