A probabilistic deep learning methodology that learns from climate simulation big data offers advantageous seasonal forecasting skill and crucial climate model diagnosis information at a global scale.
Editors’ Highlights
Posted inEditors' Highlights
A Glimpse at Planet Formation at the Dawn of the Solar System
The low density of Kuiper Belt Object Arrokoth sheds light on the formation of planetesimals in the early solar system.
Posted inEditors' Highlights
Radar Diagnosis of the Thundercloud Electron Accelerator
Altitude-resolved S-band radar observations of graupel are used to decipher thunderstorm ground enhancements in surface electric field and gamma ray flux.
Posted inEditors' Highlights
Simulating Surface Impacts of Stratospheric Sudden Warmings
New evaluations of climate model simulations show how the stratosphere polar vortex couples to surface weather.
Posted inEditors' Highlights
Time-Step Filtering in Holocene Global Magnetic Field Models
Through a local fixed time-step filter, global Holocene magnetic field models remain mathematically tractable refining our insight into field variability and improving archeological dating.
Posted inEditors' Highlights
Using Artificial Intelligence to Study Convection
Machine learning techniques are used to examine relationships between the large-scale state of the atmosphere, the convection total area, and the degree of organization in northern Australia.
Posted inEditors' Highlights
Uncovering the Mysterious STEVE Aurora
Scientists present the first direct observations on the rapid evolution of a bright red auroral arc into a thin white-mauve arc known as STEVE.
Posted inEditors' Highlights
Cloud Feedbacks in CMIP6 Models Versus Expert Synthesis
Models with smallest feedback errors are found to have moderate cloud feedbacks and equilibrium climate sensitivity of 3 to 4 degrees of warming when the concentration of carbon dioxide is doubled.
Posted inEditors' Highlights
Watching the Substorms Grow
Updated procedures enable consistent use of a wide network of polar magnetometers to monitor energy flow into the tail of Earth’s magnetosphere during the growth phase of substorms.
Posted inEditors' Highlights
Good or Bad Jam? Modeling Boulders’ Fate at Constrictions
A new modeling framework to assess the likelihood of jamming at constrictions can be used to support the design of effective mitigation measures and reduce risk in debris flow prone areas.