From translating data into tools, to improving hazard management: Fund for Climate Solutions awards five new grants
The first round of 2025 Fund for Climate Solutions (FCS) awardees has been announced. The FCS advances innovative, solutions-oriented climate science through a competitive, internal, and cross-disciplinary funding process. Generous donor support has enabled us to raise more than $10 million towards the FCS, funding 74 research grants since 2018. Many of the latest cohort of grantees are translating data into tools for amplified impact. One project is bringing climate-related hazard expertise to at-risk communities, empowering them to co-create hazard management plans with their government officials.
A generic climate AI framework for multi-domain time series prediction
Lead: Dr. Yili Yang
Collaborators: Dr. Elchin Jafarov, Dr. Brendan Rogers, Dogukan Teber, Dr. José Lucas Safanelli, Dr. Andrea Castanho, Dr. Christopher Schwalm, Dominick Dusseau, Dr. Marcia Macedo, Dr. Jonathan Sanderman, Dr. Anna Liljedahl, Dr. Sue Natali, Dr. Michael Coe
Current climate science relies heavily on physics-based numerical models to make a wide range of predictions—from thawing permafrost in the Arctic, to tropical fires in the Brazilian Amazon, to soil carbon budgets of pastures, and risks associated with flooding. These types of models require an intense amount of computing power, and are expensive to run enough times to test a variety of detailed scenarios and assumptions. Researchers often need to simplify the models or the questions they are asking, limiting the insights that they can extract. To address this problem, the project team will use deep learning to develop a Center-wide AI framework for climate data that encompasses data from across regions and projects, reducing computational costs and energy demands. This new framework will have the potential to transform the ability of all Woodwell Climate researchers to provide policymakers with rapid, accurate predictions that support urgent climate response strategies.
Climate and Indigenous-centered boreal wildfire risk assessment
Lead: Dr. Kayla Mathes
Collaborators: Dr. Brendan Rogers and Dr. Peter Frumhoff
While fire has always been an important part of boreal ecosystems, fires that reach beyond historical patterns on the landscape are posing widespread consequences for climate, Indigenous sovereignty, and public health. Through our partnerships with land managers and community leaders in Alaska, Woodwell Climate researchers have identified two key barriers to responsive management and policy action: 1) Fire managers currently lack maps that identify areas with both a high probability of wildfire, and a high carbon emission potential from burning and permafrost thaw. 2) Current fire management priorities do not adequately include Indigenous knowledge and community needs. This project will generate two maps to address boreal fire management knowledge gaps. The team will create one map representing wildfire carbon vulnerability, and will also work with Yukon Flats Indigenous communities to co-produce a regionally-specific map that identifies their wildfire management needs and priorities.
Forms and functions of soil organic carbon
Lead: Dr. Jonathan Sanderman
Collaborators: Dr. José Lucas Safanelli, Dr. Ludmila Rattis, Dr. Christopher Neill
Not all carbon is created equal—some forms of carbon are easier for microbes to break down, while others are more persistent. In soils, scientists are typically interested in a few specific forms of organic carbon, but each type has a different decay rate. Measuring the amount of each type of carbon—referred to as fractions—in a soil sample is currently labor intensive, sometimes requiring highly specialized equipment. Woodwell Climate researchers recently proved that a 60-second, low-cost spectroscopy scan can provide similar information on soil carbon fractions to days of work with traditional methods. However, this scan relies on machine learning or deep learning algorithms and high-quality, geographically appropriate training datasets. This project will build an open-source database of soil carbon fraction data, along with freely-available models to predict soil carbon fractions using spectroscopy, hosted by the Woodwell-led Open Soil Spectral Library and Estimation Service. This groundbreaking solution offers a transformative approach to soil carbon monitoring—making soil carbon fraction prediction more widely accessible to labs around the world.
Empowering the Tropical Forests Forever Facility with a tool for informed decision-making
Lead: Dr. Glenn Bush
Collaborators: Kathleen Savage, Patrick Fedor, Emily Sturdivant, Dr. Wayne Walker, Dr. Ludmila Rattis, Dr. Michael Coe
The Tropical Forests Forever Facility (TFFF), is an initiative spearheaded by the Government of Brazil to establish a US$125 billion global investment fund. If successfully established, TFFF can generate long-term finance to provide ongoing annual compensation to tropical forest nations to conserve intact tropical forests. The fund now needs to build confidence amongst potential sponsors to demonstrate feasible pathways to impact. The project team will create a new location-based dataset of cost-effective forest conservation options for the Democratic Republic of Congo (DRC) and Brazil, where Woodwell researchers have a long history of relationships and expertise. This dataset will provide valuable information around what forest conservation strategies will be effective, and where, based on financial and social benefits to people in the target landscapes. Ultimately, the team will be able to concretely identify how much forest conservation can be achieved with a given budget, and where to target efforts to resolve the highest-priority risks. With this information, the TFFF can demonstrate the effectiveness of the program and incentivize participation of sponsor countries and tropical forest nations.
Landslide hazard management workshops in Homer and Seward, Alaska
Lead: Dr. Anna Liljedahl
Collaborator: Dr. Jennifer Francis
Extreme rain events, glacier retreat, and permafrost thaw are making landslides and landslide-generated tsunamis in Alaska more likely. However, these hazards are not well-integrated into land and emergency management—for example, warning systems are relatively non-existent. The project team will host workshops in two Alaska communities at risk for landslides and landslide-generated tsunamis to raise awareness about the threats among residents and public agencies, and to identify landslide hazard management practices. The workshops will bring together experts in science and hazard mitigation; city, borough, and state officials; and community members to jointly develop recommendations for action. This effort builds on the ongoing work of Dr. Liljedahl’s project, Arctic Tsunamigenic Slope Instabilities Partnership (Arctic T-SLIP), and will support a future group of research proposals to the National Science Foundation on landslides and landslide-generated tsunami hazards. Insights gained from these workshops will also add detail to a Woodwell Climate Just Access risk report completed for Homer in 2021.
Air quality monitoring to machine learning: Fund for Climate Solutions awards six new grants
The second round of 2024 Fund for Climate Solutions (FCS) awardees has been announced. The FCS advances innovative, solutions-oriented climate science through a competitive, internal, and cross-disciplinary funding process. Generous donor support has enabled us to raise more than $10 million towards the FCS, funding 69 research grants since 2018. The latest cohort of grantees includes three projects focused on driving impact through collaboration and community-building, and three projects exploring new horizons in technology with timely policy relevance.
Arctic wildfire pollutants: Towards improving emissions estimates and developing tribally-led monitoring
Lead: Scott Zolkos
Collaborators: Brendan Rogers, Sue Natali, Kyle Arndt, Elise Sunderland (Harvard University)
Increasing wildfire activity in northern high-latitude regions is threatening global climate goals and public health. When organic matter in soils and vegetation burns, greenhouse gasses, fine particulates (PM2.5), and contaminants including mercury are released to the environment. Currently, there is sparse data for understanding how wildfires contribute to the northern mercury cycle, as well as gaps in infrastructure for monitoring PM2.5 in Alaska Native communities. This project will develop a network to measure and monitor the release of mercury and PM2.5 from wildfire, with an emphasis on peatlands. Leveraging ongoing work by Permafrost Pathways, the team will install mercury sampling equipment on existing eddy covariance flux towers across Alaska and Canada. Alongside Permafrost Pathways and their tribal partners, the team will also consult with Alaska Native communities in the Yukon-Kuskokwim Delta to co-develop a tribally-led air quality monitoring program.
Workshop: Innovative sensors and applications in environmental research
Lead: Kathleen Savage
Collaborators: Zoë Dietrich, Marcia Macedo
Many of the Woods Hole science community’s cutting-edge researchers, including several scientists at Woodwell Climate, are developing creative, do-it-yourself (DIY) tools using relatively simple components to further explore their research questions. However, despite the six institutions’ similar applications and geographic proximity, there are few opportunities for exchange and knowledge, both across Woods Hole institutions and more broadly with Cape Cod educational institutions. The project team will convene a one-day workshop to bring together aquatic, atmospheric, and terrestrial science researchers and educators from the Woods Hole science community and local community colleges. The event will focus on three main themes: development of new sensor systems that use existing technologies in novel ways; new data storage or transmission solutions; and community initiatives to facilitate continued creation and sharing of new technologies. Sessions will foster knowledge exchange, build networks, and develop community resources focused on innovative DIY research solutions, and a hybrid virtual option will be offered for oral presentations to broaden participation.
Soil Spectroscopy for Global Good network
Lead: José Lucas Safanelli
Collaborators: Jonathan Sanderman
The Soil Spectroscopy for Global Good (SS4GG) initiative is a collaborative network of hundreds of soil scientists and others focused on using soil spectroscopy as a means to generate high-quality soil data at significantly reduced costs. It was created in 2020 by the Woodwell Climate Research Center, the University of Florida, and the OpenGeoHub Foundation (the Netherlands) with support from many national and international institutions and researchers. SS4GG created and supports the Open Soil Spectral Library (OSSL), an open source of soil spectroscopy data, and a broad community of practitioners uses the library and collaborates on related science. This award will extend the activities of the SS4GG initiative with a focus on training and further engagement with the soil science community. The project team will continue to add data sets and new models to the OSSL, as well as engage with the soil science community by attending international conferences and providing a training workshop. The funds will also support hosting a visiting soil biogeochemist at the Woodwell Climate campus—Dr. Raj Setia from the Punjab Remote Sensing Center.
Pathways of carbon metabolism under cover crops
Lead: Taniya RoyChowdhury
Collaborator: Jonathan Sanderman
Sequestering, or capturing carbon in soils has a high potential to mitigate climate change. It is challenging to specifically predict how successful carbon sequestration may be, as current models used to evaluate agronomic management oversimplify soil microbial properties. This project will test for the key pathways of carbon transformations using soil samples taken under cover crops from a long-term study site. The team will quantify the chemical diversity of carbon substrates that microbes in the soil take up, and use data mining to predict the impacts of that diversity on soil carbon sequestration and nutrient cycling. The research outcomes will also lay a foundation for future collaborative research within the Department of Energy scientific community, and the soil health research community more broadly.
Bringing confidence to carbon markets through improved monitoring
Lead: Seth Gorelik
Collaborator: Wayne Walker
The protection, improved management, and restoration of forests are key nature-based solutions to the climate crisis, yet implementation and maintenance of these forest-based solutions requires sustainable and substantial financing. The voluntary carbon market (VCM) has the potential to deliver the necessary level of financing; however, a significant gap exists between its potential and actual performance. Improving the accuracy of forest carbon monitoring is crucial for the VCM to deliver effective, meaningful climate change mitigation. This project will enhance the credibility and effectiveness of forest carbon markets by evaluating new remote sensing methods for measuring forest carbon and showing that these methods provide more robust data than the conventional approach. Research findings could lead to updated global standards and policies for issuing carbon credits, which would increase market confidence and promote sustainable forest management.
Applying machine learning models to link river hydrology and fire risk forecasting in the Amazon
Lead: Andrea D. de Almeida Castanho
Collaborators: Michael Coe, Marcia Macedo
In recent decades, extreme drought events have increased forest flammability, fire severity, and the likelihood of fire escaping and spreading into adjacent forests and working lands, as illustrated by the wildfires seen throughout Amazonia during the 2023-24 drought. The project team will explore the potential of using river stage (water level) data as a proxy for landscape dryness, to ultimately reveal the short-term risk of wildfires spreading into forests. If confirmed, this innovative hypothesis could provide the scientific basis for developing new metrics of river stage to improve early-warning systems that forecast high fire risk days to weeks in advance. These improvements would create benefits not only for tropical forest protection, but also for biodiversity, greenhouse gas emissions, and human health.
Research Assistant Colleen Smith crouches low to the ground over a tray of crumbled soil. Using a boxy grey device that looks like a heavy-duty flashlight, she presses the flat glass end against the soil and fires a beam of infrared energy that bounces off the soil and back into the device’s sensor.
In moments, a readout pops up on a tablet screen, showing a spectrum of reflected light. With some analysis, Smith will have data on the chemical makeup of this patch of ground. With enough data points, she could estimate the soil properties of an entire field, pasture, ranch or farm, and how it might be changing over time.
Soil spectroscopy is a newer but fast-growing technique employed by scientists studying soil composition. At Woodwell Climate Research Center, a group led by Carbon Program Director Dr. Jonathan Sanderman has been spearheading its use to help improve the availability and affordability of reliable soil quality information, which is essential if we want to get serious about soil carbon sequestration as a natural climate solution.
Why soil spectroscopy?
“The heart of the technology is essentially getting the fingerprint of the soil, which tells us something about the overall chemical makeup of that sample,” says Dr. Sanderman.
The principles of soil spectroscopy are based in nuclear physics. Elements in the soil react in unique ways to the energy from the electromagnetic spectrum, reflecting some wavelengths and absorbing others. The reflected wavelengths give scientists clues to which minerals and elements are present and in what quantities.
That information can then be related to certain soil properties, like whether it’s suitable for certain crops, or whether it’s effectively sequestering carbon. The former is valuable information for producers like ranchers or farmers who need to make land management decisions. The latter is what climate researchers are most interested in. Soil spectroscopy represents an opportunity to marry the interests of both.
In a single scan, soil spectroscopy can estimate carbon, nitrogen, phosphorus, moisture, pH levels, and more. Traditional methods rely on multi-step chemical analyses to get you the same information— a time consuming and expensive process that could involve grinding, drying, weighing, mixing with reagents, and other steps to extract information on just one or two indicators of soil quality.
“With soil spectroscopy, you can get a pretty large suite of properties from one sixty second scan. A lab needs easily $2 million worth of instruments to be able to make all the same measurements using traditional methods,” says Dr. Sanderman. The most precise soil spectrometers can cost $100,000, but lower resolution and portable ones are substantially cheaper. “The speed and cost of spectroscopy are unmatched.”
Soil Spectroscopy for Global Good
These benefits make soil spectroscopy a method with big potential, but according to Dr. Sanderman there is still work to be done in refining the methodology to get universally accurate data. Alongside collaborators from the University of Florida and OpenGeoHub, he started the Soil Spectroscopy for the Global Good project (SS4GG) to jumpstart that work.
The project focused on two main efforts. The first was an extensive inter-laboratory comparison to understand how much the accuracy of scans varies between different instruments. Twenty laboratories across the globe participated, scanning identical samples which were then compared to the output from a lab widely regarded as the gold-standard in accuracy. The results were published in Geoderma late last year.
“We demonstrated that there is lab-to-lab variability, but also that there are procedures we can use to correct for differences between laboratories and get better integration of data,” says Postdoctoral Researcher, Dr. José Safanelli, who coordinated the study.
The second goal was to pool data from different labs into one accessible and open-source resource that also provides tools to analyze the data. The Open Soil Spectral Library (OSSL) now hosts over 100,000 soil spectra from across the globe that scientists can incorporate into their research and offers an engine for analysis. The idea is that with more people using and contributing soil spectral data, the faster the technology and the information gained from it will advance.
“We hope that the OSSL will be a driver of the soil spectroscopy community, advancing the pace of scientific discovery, and promoting innovation,” says Dr. Safanelli.
Building a community of soil scientists
Throughout the project, SS4GG efforts remained dedicated to transparency.
“We were always available to answer questions. We shared best practices and gave advice on which instruments are better, which manufacturers are the best in the market, and which procedures to use to collect spectra,” says Dr. Safanelli.
According to Dr. Sanderman, that openness fostered trust and collaboration— in both contributing data to the OSSL and participating in the inter-laboratory study— strengthening the community of scientists using soil spectroscopy.
“As we built momentum, more groups began to contribute,” says Dr. Sanderman. “It’s been great to see people realizing the value of collaborative, open science. People are now taking advantage of the foundation we’ve built.”
The soil spectroscopy community convened this past year for several webinars and presentations, including the Agronomy, Crop, and Soil Science Society meeting, where Drs. Sanderman and Safanelli hosted a training workshop and symposium on spectroscopy, as well as a two-day immersive workshop on the future of the field.
“We all benefit when this technology is more widely used,” says Smith.
Soil carbon as a climate solution
Speeding up the pace of soil science is key for developing climate solutions. Agricultural soils represent a large potential carbon sink; changes in farming and ranching practices can encourage sequestration of carbon in the soils. Soil carbon markets, and other payment for ecosystem services schemes could incentivise producers to make sustainable management decisions and soil spectroscopy could be a useful tool to track their contributions.
“The ultimate goal is to better monitor soils across landscapes to make food production more sustainable,” says Dr. Safanelli.
The handheld device that Smith was using is a test case for the speed and convenience of soil spectroscopy for analyzing soil carbon. If testing the quality of your soils can be as simple as a 60 second measurement with a low-cost piece of portable equipment, and the scan can get you additional information about soil fertility, then why not participate?
“We are trying to verify that we actually are sequestering carbon, and that requires lots and lots of measurements. So this is where we start moving into field-based spectroscopy,” says Dr. Sanderman. “If we can eliminate bringing the sample back to the lab altogether, we’re cutting our costs by another order of magnitude and could potentially scan several hundred points in a field in a day.”
Smith theorizes that cost could be further diffused through farming cooperatives or extension offices offering soil testing using inexpensive spectrometers. “Soil spectroscopy could be an easier way to get answers to big questions,” says Smith. “And that’s exciting.”
With the OSSL now up and running, the team is now focusing efforts on maintaining the growing network of interested soil researchers, pursuing new opportunities for collaboration as they arise.
“The network is getting stronger,” says Dr. Safanelli. “More people are coming and reaching out to us. That’s our biggest contribution: creating a network and sharing information across the community.”