Recent advances in machine learning hold great potential for converting a deluge of data into weather forecasts that are fast, accurate, and detailed.
data management
Posted inResearch Spotlights
Updating a Crucial Source of Sea Surface Temperature Data
A new version of a major sea surface temperature data set reduces systematic errors in measurements of one of the most important indicators of the state of Earth’s climate system.
Posted inScience Updates
Updated Temperature Data Give a Sharper View of Climate Trends
The latest version of NOAA’s Global Surface Temperature Dataset improves coverage over land and sea and improves the treatment of historical changes in observational practices.
Posted inGeoFIZZ
Rock On with a Group That Makes Music from Geophysical Data
Musical numbers include an Italian fault and a tour of an African greenstone belt.
Posted inEditors' Vox
Age and Speed Matter in the Formation of New Oceanic Crust
A synthesis of data from studies in different ocean basins reveals that the characteristics of oceanic crust are shaped by age and spreading rate.
Posted inEditors' Highlights
Linking Regional Weather and Climate to Remote Events
A new index for quantifying regional sensitivities to the influence of periodic events.
Posted inResearch Spotlights
Data Mining Reveals the Dynamics of Auroral Substorms
An analysis of 5 decades of satellite data has pieced together the most comprehensive picture yet of substorms, the magnetic disturbances that cause surges of aurora.
Posted inResearch Spotlights
Big Data Help Paint a New Picture of Trace Element Cycling
A new framework for understanding the suite of processes acting on marine particulate trace metals exemplifies how alternative analyses can maximize the information that large data sets provide.
Posted inResearch Spotlights
A Novel Method for Assessing Model Sensitivity
This newly developed approach to assessing the quality of sensitivity analyses can be applied to any method without bootstrapping or additional model runs.
Posted inEditors' Highlights
X-Ray Computed Tomography Detects Resolution Scale Gold Grains
A method combining partial-volume and blurring effects can be used to measure small features in computed tomography data volumes.