Departure Amid Political Pressure
Climate scientist and author Kate Marvel resigned last month from NASA’s Goddard Institute for Space Studies, where she had worked for more than a decade modeling Earth’s warming climate. In her resignation letter, she wrote that she expected political pushback against her work but did not anticipate what she described as broader attacks on science itself.
Her departure comes during a period of sweeping changes across federal science agencies. Since January 2025, more than 10,000 PhD-level scientists have reportedly left federal positions. The administration has also dismantled the U.S. Global Change Research Program, dismissed authors of the upcoming National Climate Assessment, and withdrawn from major international climate agreements.
Scientific Speech and Self-Censorship
Marvel said her decision was driven less by partisan disagreement than by restrictions on discussing scientific findings publicly. She argued that the mission of publicly funded research depends on communicating results to the public.
According to Marvel, researchers increasingly engaged in self-censorship, avoiding terms such as “climate change” in grant proposals in favor of less politically charged language. Independent analyses of federal grant databases have shown sharp declines in traditional climate terminology after 2025, alongside increases in substitute phrases such as “extreme weather.”
She described a fragmented work environment following the eviction of the Goddard Institute from its longtime building, limiting collaboration and cohesion among researchers.
Global Implications for Climate Modeling
Marvel emphasized that climate science is inherently global, relying on international data-sharing and coordinated modeling efforts. The U.S. has historically played a leading role in building climate modeling infrastructure and attracting top researchers. She warned that retreating from global scientific cooperation could diminish American influence and shift leadership elsewhere.
Publicly funded research, she noted, operates under different incentives than private-sector initiatives. In areas such as geoengineering, where research into solar radiation management is advancing, she expressed concern about reduced transparency if oversight shifts away from public institutions.
Carbon-Cycle Feedback: A Critical Unknown
Much of Marvel’s recent work focused on carbon-cycle feedback, a key uncertainty in climate projections. While physical feedbacks such as melting ice and cloud dynamics are relatively well understood, carbon-cycle feedback examines how much of human-emitted carbon dioxide remains in the atmosphere.
Currently, natural systems absorb roughly half of human emissions. Whether that capacity will persist under escalating heat, wildfire and ecosystem stress remains uncertain. Understanding this dynamic requires interdisciplinary collaboration across plant biology, ecology, meteorology and climate modeling.
Emotion, Motivation and the Future
Marvel has written about the emotional toll of modeling worst-case climate scenarios. She argues that acknowledging emotion does not undermine scientific credibility, provided researchers remain rigorous in analyzing data and updating conclusions when evidence changes.
Looking ahead, she expressed concern about early-career scientists facing shrinking opportunities. At the same time, she suggested that widespread frustration within the scientific community could fuel renewed engagement and institutional rebuilding.
Despite uncertainty, Marvel said she does not see giving up as an option. Climate science, she argued, will continue — though its institutional landscape may look very different in the years ahead.