Predicting and Preventing Peatland Fires: Aalto University Develops Groundbreaking Neural Network Model ‘FireCNN’

August 27, 2023 / activist360

Military might. Army officers try to extinguish fires in peat land areas, outside Palangka Raya, Central Kalimantan. Photo by Aulia Erlangga/CIFOR.

Aalto University researchers have developed a neural network model that can predict peatland fires in Central Kalimantan, Indonesia. The model performs consistently well, with ranges about the medium values of 95% for accuracy, and 78% for precision.

FireCNN, First-Ever Model Capable of Predicting Future Fire Locations

The researchers developed ‘FireCNN’, the first-ever model that can accurately predict the locations of future fires. FireCNN uses a type of machine learning algorithm called CNN (convolutional neural network) to analyze various factors that can predict fire occurrences (e.g., weather conditions, land use) before the start of fire season. The model allows researchers to test how different land management and restoration strategies, such as blocking canals, reforestation, and converting land to plantations, might impact the number of fires in the future without any bias. Researchers also simulated the effects of ongoing deforestation, converting swamp forests into degraded scrublands and plantations, to understand its potential impact on future fires.

The Focus of the Research

Indonesian peatlands face recurrent fires due to human-induced degradation, increasing recurrent fires since the late 1990s. These fires release CO2, equivalent to 30% of global fossil fuel emissions in 2020, and negatively impact the environment, economy, public health, agriculture, and social structure. In 2015, this resulted in a loss of over $16 billion to the Indonesian economy. Despite prohibitions, most ignitions are anthropogenic, started for agricultural expansion.

The investigation focused on the ex-Mega Rice Project (EMRP) area in central Kalimantan, Borneo, which has the highest density of peatland fires in Southeast Asia, recurring since 1997 due to logging, oil palm plantation development, and a failed rice cultivation scheme. This scheme inadvertently transformed swamp forests into degraded peatlands by digging 4000 km of drainage canals and clearing 1 million hectares of swamp forest. The area has distinct dry and wet seasons but a consistent mean monthly temperature of 28°C. Fire season hotspots peak around 11,000 but vary significantly yearly.

Study area map. Land cover map showing the whole study area (edge of map) circa 2015 as well as the ex-Mega Rice Project (EMRP) area (black outline). Inset map of Borneo provided by OpenStreetMap. Horton, A.J., Lehtinen, J. & Kummu, M. Targeted land management strategies could halve peatland fire occurrences in Central Kalimantan, Indonesia. Commun Earth Environ 3, 204 (2022).

Researchers found that converting degraded swamp shrubland to swamp forest or plantations could reduce fire occurrences by 40-55%. Blocking most canals could reduce fire occurrences by 70%. Effective strategies can reduce carbon emissions and enable sustainable ecosystem management.

Reducing peatland fires is essential for global carbon emission reduction, economic productivity, biodiversity safeguarding, and protecting vulnerable communities. However, efforts in Central Kalimantan have been unsuccessful due to corruption, poor governance, and lack of accountability. Previous studies lacked clear links between restoration efforts and future fire reductions.

Hope for the Development of an Early-Warning System

The findings demonstrate the potential impacts of future peatland restoration efforts, providing much-needed evidence for the potential success of these strategies, which may benefit similar projects currently underway. Postdoctoral researcher Alexander Horton noted that while the methodology could apply to other contexts, the model would need retraining on new data. Researchers hope to improve the model’s performance to serve as an early-warning system.

We tried to quantify how the different strategies would work. It’s more about informing policy-makers than providing direct solutions.
—Professor Matti Kummu, study team’s leader

University of Central Florida Researchers Unveil Breakthrough in Greenhouse Gas Recycling

August 13, 2023 / activist360

UCF researchers Richard Blair (left) and Laurene Tetard (right) are long-time collaborators and have developed new methods to produce energy and materials from the harmful greenhouse gas, methane.

In a significant step toward sustainable energy, researchers from the University of Central Florida (UCF) have innovated methods to convert the potent greenhouse gas, methane, into green energy and advanced materials.

Methane, with an impact 28 times greater than carbon dioxide over a century, is a notable contributor to global warming. Its emissions predominantly arise from energy sectors, agriculture, and landfills. Now, UCF’s groundbreaking methods might turn this environmental challenge into an opportunity, as they utilize methane for producing green energy and crafting high-performance materials for smart devices, solar cells, and biotech applications.

Behind these inventions are UFC researchers, nanotechnologist Laurene Tetard and catalysis specialist Richard Blair. Tetard is an associate professor and associate chair of UCF’s Department of Physics. He is also a researcher with the NanoScience Technology Center. Blair is a research professor at UCF’s Florida Space Institute. The two have been collaborating on research projects for the past decade.

Their pioneering technique produces hydrogen from methane without carbon gas emission. Utilizing visible light sources, like lasers or solar energy, and defect-engineered boron-rich photocatalysts, the process emphasizes the advanced potential of nanoscale materials.

Blair highlights the dual benefits: You get green hydrogen, and you remove — not really sequester — methane. You’re processing methane into just hydrogen and pure carbon that can be used for things like batteries.” Traditional methods, Blair notes, produce CO2 along with hydrogen. Their innovation not only tackles methane emissions but also transforms it into valuable hydrogen and carbon. Market applications include possible large-scale hydrogen production in solar farms and methane capture and conversion.

“Our process takes a greenhouse gas, methane and converts it into something that’s not a greenhouse gas and two things that are valuable products, hydrogen and carbon. And we’ve removed methane from the cycle.”
—Richard Blair, research professor at UCF’s Florida Space Institute

Additionally, this technology from Tetard and Blair offers the ability to manufacture carbon structures at nano and micro scales using light and a defect-engineered photocatalyst. Envisioning it as a “carbon 3D printer,” Blair notes the dream is to make high-performance carbon materials from methane.

“It took a while to get some really exciting results,” Tetard says. “In the beginning, a lot of the characterization that we tried to do was not working the way we wanted. We sat down to discuss puzzling observations so many times.”

Countries lacking significant power sources could potentially benefit, requiring only methane and sunlight to leverage the innovation. As Blair summarizes, the process takes a greenhouse menace and turns it into precious, non-polluting commodities.

Ants – with their wise farming practices and efficient navigation techniques – could inspire solutions for some human problems

November 20, 2022 / activist360

By Scott Solomon, The Conversation

King Solomon may have gained some of his famed wisdom from an unlikely source – ants.

According to a Jewish legend, Solomon conversed with a clever ant queen that confronted his pride, making quite an impression on the Israelite king. In the biblical book of Proverbs (6:6-8), Solomon shares this advice with his son: “Look to the ant, thou sluggard, consider her ways and be wise. Which having no guide, overseer, or ruler, provideth her meat in the summer, and gathereth her food in the harvest.”

While I can’t claim any familial connection to King Solomon, despite sharing his name, I’ve long admired the wisdom of ants and have spent over 20 years studying their ecology, evolution and behaviors. While the notion that ants may offer lessons for humans has certainly been around for a while, there may be new wisdom to gain from what scientists have learned about their biology.

Ants have evolved highly complex social organizations.

Lessons from ant agriculture

As a researcher, I’m especially intrigued by fungus-growing ants, a group of 248 species that cultivate fungi as their main source of food. They include 79 species of leafcutter ants, which grow their fungal gardens with freshly cut leaves they carry into their enormous underground nests. I’ve excavated hundreds of leafcutter ant nests from Texas to Argentina as part of the scientific effort to understand how these ants coevolved with their fungal crops.

Much like human farmers, each species of fungus-growing ant is very particular about the type of crops they cultivate. Most varieties descend from a type of fungus that the ancestors of fungus-growing ants began growing some 55 million to 65 million years ago. Some of these fungi became domesticated and are now unable to survive on their own without their insect farmers, much like some human crops such as maize.

Ants started farming tens of millions of years before humans.

Ant farmers face many of the same challenges human farmers do, including the threat of pests. A parasite called Escovopsis can devastate ant gardens, causing the ants to starve. Likewise in human agriculture, pest outbreaks have contributed to disasters like the Irish Potato Famine, the 1970 corn blight and the current threat to bananas.

Since the 1950s, human agriculture has become industrialized and relies on monoculture, or growing large amounts of the same variety of crop in a single place. Yet monoculture makes crops more vulnerable to pests because it is easier to destroy an entire field of genetically identical plants than a more diverse one.

Industrial agriculture has looked to chemical pesticides as a partial solution, turning agricultural pest management into a billion-dollar industry. The trouble with this approach is that pests can evolve new ways to get around pesticides faster than researchers can develop more effective chemicals. It’s an arms race – and the pests have the upper hand.

Ants also grow their crops in monoculture and at a similar scale – after all, a leafcutter ant nest can be home to 5 million ants, all of which feed on the fungi in their underground gardens. They, too, use a pesticide to control Escovopsis and other pests.

Yet, their approach to pesticide use differs from humans’ in one important way. Ant pesticides are produced by bacteria they allow to grow in their nests, and in some cases even on their bodies. Keeping bacteria as a living culture allows the microbes to adapt in real time to evolutionary changes in the pests. In the arms race between pests and farmers, farming ants have discovered that live bacteria can serve as pharmaceutical factories that can keep up with ever-changing pests.

Whereas recent developments in agricultural pest management have focused on genetically engineering crop plants to produce their own pesticides, the lesson from 55 million years of ant agriculture is to leverage living microorganisms to make useful products. Researchers are currently experimenting with applying live bacteria to crop plants to determine if they are effective at producing pesticides that can evolve in real time along with pests.

Improving transportation

Ants can also offer practical lessons in the realm of transportation.

Ants are notoriously good at quickly locating food, whether it’s a dead insect on a forest floor or some crumbs in your kitchen. They do this by leaving a trail of pheromones – chemicals with a distinctive smell ants use to guide their nest mates to food. The shortest route to a destination will accumulate the most pheromone because more ants will have traveled back and forth along it in a given amount of time.

In the 1990s, computer scientists developed a class of algorithms modeled after ant behavior that are very effective at finding the shortest path between two or more locations. Like with real ants, the shortest route to a destination will accumulate the most virtual pheromone because more virtual ants will have traveled along it in a given amount of time. Engineers have used this simple but effective approach to design telecommunication networks and map delivery routes.

Not only are ants good at finding the shortest route from their nests to a source of food, thousands of ants are capable of traveling along these routes without causing traffic jams. I recently began collaborating with physicist Oscar Andrey Herrera-Sancho to study how leafcutter ants maintain such a steady flow along their foraging paths without the slowdowns typical of crowded human sidewalks and highways.

We are using cameras to track how each individual ant responds to artificial obstacles placed on their foraging trails. Our hope is that by getting a better understanding of the rules ants use to respond to both obstacles and the movement of other ants, we can develop algorithms that can eventually help program self-driving cars that never get stuck in traffic.

Look to the ant

To be fair, there are plenty of ways ants are far from perfect role models. After all, some ant species are known for indiscriminate killing, and others for enslaving babies.

But the fact is that ants remind us of ourselves – or the way we might like to imagine ourselves – in many ways. They live in complex societies with division of labor. They cooperate to raise their young. And they accomplish remarkable engineering feats – like building structures with air funnels that can house millions – all without blueprints or a leader. Did I mention their societies are run entirely by females?

There is still a lot to learn about ants. For example, researchers still don’t fully understand how an ant larva develops into either a queen – a female with wings that can live for 20 years and lay millions of eggs – or a worker – a wingless, often sterile female that lives for less than a year and performs all the other jobs in the colony. What’s more, scientists are constantly discovering new species – 167 new ant species were described in 2021 alone, bringing the total to more than 15,980.

By considering ants and their many fascinating ways, there’s plenty of wisdom to be gained.