Spotlighting Novel AI-Environmental Data Innovations and Climate Datasets on Earth Day
By Alyson Marks
Photo by Getty
Last month marked the release of the latest Intergovernmental Panel on Climate Change’s (IPCC) Sixth Assessment Report (AR6), which draws on the climate change-related research of hundreds of scientists. The report offers a bleak analysis – demonstrating that “climate impacts on people and ecosystems are more widespread and severe than expected, and future risks will escalate rapidly with every fraction of a degree of warming.” Furthermore, the report underscores the importance of comprehensive data on the environment and the need to continue to develop data-driven responses to address the complexity of the climate crisis. Without these, the implementation of climate adaptation options, particularly for vulnerable groups, will fail.
Despite the grim reality that the report presents, there have been many recent positive developments in the field of environmental data. For instance, in December, after years of preparation and negotiation, the parties of the Convention on Biological Diversity (CBD) adopted the “Kunming-Montreal Global Biodiversity Framework” (GBF), an inclusive and transformative agenda for protecting, restoring, and sustainably using and managing biodiversity and ecosystems. The GBF includes a monitoring framework designed to ensure that implementation of the GBF is results-oriented and can be monitored transparently. And last month, the UN Environmental Programme’s report on ‘Measuring Progress: Water-related Ecosystems and the SDGs’ highlighted that global data availability for environment-related SDG indicators has steadily increased over the past few years – rising from 34% in 2018 and 42% in 2020 to 59% in 2022. In recent months, there have also been a number of environmental data innovations in the field of Artificial Intelligence and new datasets, and we’ve spotlighted a handful of these in honor of Earth Day.
Mobilizing AI to Address Climate Challenges
With the rise of ChatGPT and a slew of other AI advancements, discussions around AI have been ubiquitous in the news, popular culture, and across the global development community. While there are a number of ethical challenges that AI brings into question, the technology offers great potential for positive societal impact in the field of sustainable development. In particular, AI combined with traditional data sources allows for significantly more granular, quality, lower-cost, less resource-intensive, and timely data than ever before to improve decision-making and planning. Some recent noteworthy innovations include:
Helping Cities Adapt to Extreme Heat - Google’s Tree Canopy project combines AI and aerial imagery to enable policymakers to better understand their city’s current tree coverage and plan urban forestry initiatives, including prioritizing planting trees in areas more susceptible to extreme heat and ensuring equitable canopy coverage. Last month, Google announced that it has expanded the project to nearly 350 cities globally.
Tackling Agricultural Challenges - Google is also using AI to enable farmers to more accurately understand field performance and environmental conditions to improve their crop yields. The project leverages satellite imagery and machine learning to draw boundaries between fields, allowing the model to determine the acreage of farm fields, forest, and woodland areas and to identify irrigation structures (e.g., farm wells and dug pounds) to build tools for drought preparedness. In parallel, the project is also developing ‘landscape monitoring’ models to provide a more granular picture of individual fields’ performance and needs, including crop type, distance to water, last harvest date, and more, which are essential to improving drought management and water security strategies.
Identifying Broken Water Systems in Water-Stressed Areas – In East Africa, access to clean and safe water remains a challenge. In recent years, many communities have begun to rely on solar-powered water systems that pump water from a borehole and store it in a collection tank. Yet, reporting when these systems break (which occurs frequently) can prove challenging due to the remoteness of the location and poor internet connectivity. Consequently, an international social enterprise, Virridy, has partnered with several local and international partners to provide innovative water monitoring technologies through a Drought Resilience Impact Platform (DRIP). The system uses satellite-connected sensors to identify broken water systems by transmitting collected data through an open-access dashboard to response teams and reporting on whether a system is in use or has reduced its capacity to deliver water to communities.
Improving Climate Change Research – In February of this year, IBM and NASA announced a new project that will use IBM’s AI technology with NASA’s Earth and geospatial science data to more seamlessly draw insights from these large datasets. The collaboration will also generate several new technologies, including an easy-to-search collection of Earth Science literature and a foundation model for weather and climate prediction.
New Environmental Datasets
Coupled with the above environmental data innovations, several new datasets to address climate issues have been released and/or updated in the past few months. These include, but are not limited to the following:
NASA’s Socioeconomic Data and Applications Center (SEDAC)’s Groundswell Spatial Population Dataset – A new dataset from SEDAC released this past March will help users to better understand how climate change affects internal migration in low- to middle-income countries in the coming decade. The dataset provides projections of future population distribution and migration at a resolution of one-eighth degree for more than 110 countries.
The NOAA’s National Centers for Environmental Information (NCEI) Global Climate Dataset – In February 2023, the NCEI updated its global climate dataset (one of the most widely used datasets to assess the global climate) to include more data for the Arctic region and new scientific methods for monitoring climate in other locations with limited climate data. It also features an improved methodology to analyze NCEI’s archival land and ocean observations.
A Global Dataset of Reservoir and Lake Surface Area Variations – Last July, a team of University of Minnesota Twin Cities data scientists published ReaLSAT, the first-of-its-kind comprehensive global dataset of the Earth’s lakes and reservoirs. The dataset offers new information about land and freshwater use, as well as how lakes and reservoirs have been impacted by climate change over the past 30+ years.
New Indicators to Analyze Climate Risks in the Financial Sector –The European Central Bank (ECB) published a first set of climate-related statistical indicators this January 2023 that will assess the impact of climate-related risks on the financial sectors, as well as monitor the development of sustainable and green finance (e.g., the number of debt-related instruments labeled as ‘green’ or ‘sustainable’ issued and carbon emissions financed). While the indicators are ‘experimental’ or ‘analytical’ in nature, the ECB, alongside national central banks, will continue to work to improve the methodology and the data used.
Looking Ahead
The recent data innovations demonstrate a promising trend in the fight against climate change, yet it remains to be seen if policymakers will fully adopt them to address the crisis. Next week’s World Data Forum presents an important opportunity to continue to showcase these innovations, particularly in the field of AI, and connect stakeholders across the data ecosystem to work together toward developing solutions to improve the state of our planet.