Timescales that Matter: A Case for Faster Feedback in Environmental Science

It’s a level of detachment that we rarely get to experience: Seeing the Grand Canyon from above, an illuminated coastline at night, or recognizing the familiar skyline of your home city from a distance. Looking out the window of an airplane can give you a perspective that makes you wonder about the nature of your planet, if you care to look. If you’re on a plane headed over the American Midwest and fortunate enough to be sitting in a window seat during the daytime–you're bound to notice the circular patterns dotting the landscape beneath you. Despite their eerily perfect geometric proportions, these circles are not signs of alien visitors. These are crop circles, formed by center-pivot irrigation structures whose galvanized steel pipes rotate in a circle to apply water to a crop, often commodity crops like corn or soybean. What you may also not consider as you fly over these circles, is that the entirety of the Earth’s troposphere lies beneath you. At an altitude of 35,000 feet, your plane is above most of the weather that occurs on earth.

The Most Important Layer of the Troposphere, the Boundary Layer

At the lowest layer of the troposphere, near the Earth’s surface, is the boundary layer. The boundary layer is like that coveted, thin crust at the base of a layer cake that gets caramelized by the pan. Entire scientific journals, conferences, research programs, and many scientific careers are dedicated to understanding the inner workings of this caramelized crust. Why? What’s so interesting about this thin layer? The boundary layer is the place where most of life takes place. It’s where you and your family (and maybe your dog or cat) exist. It’s where you met your spouse, if you have one, and it’s where your credit card company’s office is located among many (read: all) other things. So, it’s not too hard to understand why it’s important.

Life Happens Here: Monitoring How Our Plants–and Planet–Breathe

The boundary layer is also the place where the earth breathes: Plants breathe in carbon dioxide, and exhale oxygen; humans drive in their cars and trucks, releasing tons of CO₂ into the atmosphere. Now, you might be wondering: How do we know that plants breathe carbon dioxide? Or how do we know how much a field of corn is breathing over a growing season? You might also be wondering why that matters, and how you could even be sure when it was happening. To start, we know that plants breathe carbon dioxide because we understand (mostly) how they take energy from light through photosynthesis.

How is carbon dioxide CO₂ in the atmosphere measured?

We can see evidence of carbon dioxide gas concentration on a global scale through the examination of the famous Keeling Curve plot, created by instrumentation at the National Oceanic and Atmospheric Administration (NOAA)’s Global Monitoring Lab, that since 1958 has been monitoring global atmospheric gases from atop Mauna Loa on the island of Hawaii. We also see evidence of carbon dioxide processes on local scales, using sophisticated tools and data processing techniques employed as part of a scientific methodology known as eddy covariance.

What is the Eddy Covariance Method

Eddy covariance is a technique that measures the amount of `something` and the wind speed at one time and the amount of that same `something` and the wind speed at another time. The times are very rapid (10 to 20 measurements per second!) and the measurements very precise. If the right conditions are met, such as stability of the atmosphere and topography of the surface, then these measurements and wind speeds can be aggregated and analyzed over a longer period, often 30 minutes, to see how much of that `something` went up or down. If that “something” is CO₂…this is the “breath” of the corn field.

One of the key requirements for eddy covariance is that turbulent conditions are present. Turbulence (which you likely experienced on that flight over the Midwest, as it jostled your can of Sprite), consists of eddies of air moving over and across the surface. These eddies vary in size, speed, and direction, and are the mechanisms that carry the “breath” of the planet into the troposphere or down to the surface, whatever the case may be.

A High-Altitude Perspective of Human Impacts on Our Planet

As you are flying on that same plane, you can see the magnitude of the impact that humans have had on our planet, and not just with things like litter and blight. Think for a moment what the landscape below you would look like if humans did not exist. The magnitude of our impact: streets, lights, houses, airports, massive city skylines, rural routes, mini malls, movie theaters, parking lots, drive-thru coffee places. We created all of this! Then, there’s an entire set of that exist things that you can’t see: Pollutants, greenhouse gases, noises, temperatures, humidity…All of these things are impacted and influenced by humans, too.

Can modern research tools keep up with today's rate of change?

The tools we have available to understand these impacts on our world--and make decisions to alter their course--evolve over time. Surely, you wouldn’t use your grandparents’ Rolodex to find the nearest sushi restaurant while you were in Berlin for work. Nor would you dig up your family’s ancient Encyclopedia Britannica from the depths of your basement to learn about the latest in convolutional neural networks. Fortunately, today, we have modern tools to cover us in each of these situations. But what happens when we’re not covered–when the tools can’t change fast enough? Because the rate of change in the world is not constant, sometimes, even in today’s modern age, the tools we have at our disposal for understanding and acting are out-of-date for the task at-hand.

Use Case #1: The Pace of AI Evolution Eclipses the Peer-Review Process

In the past, the rate of change during studies was slow enough that the traditional peer-review process was the best tool to share research findings with the community and learn from others. There was enough of a time delay from the sharing of the research to the start of the next cycle, that the feedback could be integrated into producing the next work. That is no longer the case. As a recent modification to the research publication process, preprint servers now allow authors to publish early versions of their papers, prior to the peer-review process. However, this pace pales in comparison to the rate of change in AI research.

Since 2024, the rate of change in artificial intelligence research continues to accelerate, to the point where the pace of the traditional peer-review publication process just couldn’t keep up. When the pace of development is so rapid, there’s no way around it–the tools for sharing the state of research just don’t match the situation. Even prior to the current generative AI phase, there was a rapid increase in the rate of research within deep learning, a subset of AI. In this deep learning phase, the rate of change could still fit within the traditional paradigm of research, but as the generative AI phase has taken over, the tools for understanding and communicating the state of the art in research increasingly do not match the situation.

Use Case #2: The Alarming Rate of Change in Atmospheric Carbon Dioxide

Swap the pace of AI research for CO₂ concentration in the atmosphere, and the paradigm is not particularly different: Beyond the monumental visible impact we humans have had on our environment, the invisible impact is just as important. In April 2024, the global concentration of carbon dioxide reported from Mauna Loa was ~ 420 ppm, well exceeding pre-industrial levels. Based on the best record keeping available, this level is significantly higher than at any point in the past 800,000 years (3). We are approaching a critical threshold that, if crossed, could result in changes to our planet that are difficult to predict. Regardless of whether we cross this point of no-return or not, we need tools and processes to help us understand what is happening, to make decisions, and take actions. We need a network that keeps a pulse on our entire planet, every forest, city, and cropland. When you examine the scale of geologic time, humanity is at a stage in Earth’s history where the rate of change in concentration of carbon dioxide in the atmosphere has accelerated to the point where there is no comparison.

Use Case #3: The USDA’s Cooperative Extension Program:

How the U.S. Boosted Agricultural Productivity and Efficiency in the Face of Farms’ Dramatic Decline

When looking at a much more recent time in our planet’s history, we can study the conditions in the U.S. just prior to World War I, which gave rise to the modern Cooperative Extension program in the United States. The mission of the Cooperative Extension program today is to “empower farmers, ranchers, and communities of all sizes to meet the challenges they face, adapt to changing technology, improve nutrition and food safety, prepare for and respond to emergencies, and protect our environment.” (1) As we progressed through the Great Depression, World War II, and all the way to today, the U.S. experienced a dramatic increase in output from American agriculture, while the number of farms declined from 5.4 million to 1.9 million. In 1950, one farmer in the U.S. grew enough food to feed ~ 15.5 people. In 1997, one farmer grew enough food to feed 140 people (5).

The Time is Now: The Need for Transformative Environmental Science Tools

Just as the US agricultural sector dramatically transformed in response to the dire need to increase productivity and efficiency–environmental science needs a transformative solution to help bring the critically important research that occurs into practice. Lacking a cooperative extension program tailored to environmental science, what can we do?

We possess the foundation of the Internet of Things (IoT) in addition to tools and technologies that allow us to understand the environmental processes taking place. IoT as an idea and phrase was first coined in 1998, but the underlying technology was first used to remotely monitor a Coca-Cola vending machine at Carnegie Mellon University. The core capability of IoT is the ability to directly measure something, to gain some knowledge about what is happening so you can decide something. In near real-time. Since 1998, IoT has undergone some shifts in application, including the subfield of the Industrial IoT (IIoT). In industrial settings, the IIoT-enabled temperature sensors can trigger a machine to shut off if threshold is exceeded. Or, if a dangerous gas is detected in an enclosed space, an alarm can be configured to immediately sound. For immediate feedback–IoT is the tool for the job.

The tool we need now, to help address the situation we find ourselves in, is analogous to network of digital environmental science extension offices. This virtual digital network will be fueled by data and science that is shared rapidly, and informed actions taken based on timescales that matter for our planet. The global population needs a new IoT–an Internet of the Environment (IoE).

This Internet of the Environment is one that constantly reports what is happening at the local scales of our planet, tracking how they react and contribute to a changing earth climate system. There already are multiple national-scale or international-scale networks that are funded by governments or intergovernmental panels; these are critically important as monitoring infrastructure–but we need more than that. Just as the internet has given people the tool to share information freely and immediately, a globally connected Internet of the Environment will give people the ability to share data immediately and inform decisions quickly.

The traditional paradigm of data collection, followed by analysis, followed by publication, followed by dissemination, just takes too long. We need to efficiently integrate the knowledge gained from such exercises in a reasonable amount of time and affect change when and how it’s most needed. We need answers to questions such as:

“How much carbon was taken up by that corn field?”“How much water was lost by those mangroves?” and most importantly: “What does it mean?”, “How much carbon was sequestered?” (2), “How much water do I need to apply and when?” (6), “How much longer until this forest is at risk?”, “Will my business be at risk of flood or fire damage if I don’t move?”, “What will the future look like and how can we prepare for it?” (7).

This virtual, IoT environmental monitoring network would be the blueprint for translating the critically important environmental science research that occurs into actions for people who need to act, understanding for people who need to understand, and decisions for people who need to decide.

What good is it to understand that your building is on fire, if you can’t then act to put that fire out?