A theoretical index linking precipitation and atmospheric evaporative demand explains historical land humidity changes across observations, reanalyses, and climate models.

Observations, reanalysis, and E3SMv3 experiments linked inter-annual variability of Congo Basin surface temperature to the zonal position of precipitation over the Indo-Pacific maritime continent region.

Quasi-stationary Rossby waves shape North American temperature extremes, but higher model resolution does not always improve their modeled representation.

Tests show which low-cost 3D method best matches field measurements to minimize errors in mapping headwaters.

Watershed-based representation discretized into topographic subgrid units better captured small-scale land cover, precipitation, temperature, and snow variability across multiple spatial scales.

Numerical simulations reveal how well-controlled wall temperature and moisture can generate a stratocumulus-like cloud top in a laboratory cloud chamber.

A new tracer model tracked how May 2015 storm water re-entered the atmosphere and contributed to later rainfall.

Physics-informed and data-driven neural network models enhance Earth system models in simulating compound flooding at river-ocean interfaces.

Global exercise advances nuclear forensic science and investigative capabilities.

Tracking water age using a new precipitation tagging method within a hydrologic model generates key information to support hydropower energy and other water-dependent systems.

More rain and less snow shorten how long water stays in watersheds.

​​Tests show which low-cost 3D method best matches field measurements to minimize errors in mapping headwaters.​

Five PNNL innovations earned 2026 R&D 100 Awards for advances in biotechnology, manufacturing, critical minerals, and nuclear security.

A new report highlights a public workshop hosted by the National Academies Forum on Microbial Threats and features PNNL's Lauren Charles and the AI-Driven One Health Security program.

Researchers at PNNL have developed an interpretable, lightweight AI model that can easily predict weld microstructure features using only basic machine sensor inputs.

Integrating new functions into a genome-scale model enabled researchers to simulate and experimentally validate viral hijacking pathways.

Distributed science is thriving at PNNL, where scientists share data and collaborate with researchers around the world to increase the impact of the work.

A study by researchers at PNNL assessed the feasibility of using strontium isotope ratios and an existing machine learning–based model to predict and verify a product’s source—in this case, honey.

Researchers have developed the first coastal transport model for microplastics to understand how particles move in waterways and coastal currents.

High-resolution hydrodynamic-sediment modeling shows that inundation, suspended sediment concentration in the Amazon River, and floodplain hydrodynamics drive sediment deposition in Amazonian floodplains.