Scientists propose development of new models that use machine learning techniques to reduce uncertainties in climate predictions.
Modeling
Posted inEditors' Highlights
Improving Temperature Forecasts in the Upper Atmosphere
Scientists are blending output from multi-year model runs to improve temperature forecasts in regions where satellites experience “drag,” in the hopes of avoiding future spacecraft collisions.
Posted inScience Updates
Modeling Global Change Ecology in a High–Carbon Dioxide World
Ignite-style Session, Ecological Society of America Annual Meeting; Portland, Oregon, 11 August 2017
Posted inResearch Spotlights
Modeling Storm Evolution
A “moist shell” makes all the difference in how some storms evolve.
Posted inEditors' Highlights
A Powerful New Tool to Analyze and Calibrate Earth System Models
Polynomial chaos and Bayesian compressive sensing are applied to a land surface model to understand how large numbers of tunable parameters interact and may be optimized.
Posted inOpinions
Climate Models Are Uncertain, but We Can Do Something About It
Model simulations of many climate phenomena remain highly uncertain despite scientific advances and huge amounts of data. Scientists must do more to tackle model uncertainty head-on.
Posted inEditors' Vox
The Challenges of Drought Prediction
Advances in dynamical modeling and the use of hybrid methods have improved drought prediction, but challenges still remain to improve the accuracy of drought forecasting.
Posted inResearch Spotlights
Medieval Temperature Trends in Africa and Arabia
A synthesis of paleotemperature reconstructions from published case studies suggests warm onshore temperatures persisted across most of Afro-Arabia between 1000 and 1200 CE.
Posted inEditors' Highlights
Ocean Wind Satellites Observe an Amazonian Drought
Satellites designed to observe ocean winds can also be used to map both forest structure and water content, allowing researchers to disentangle factors of carbon loss due to drought in the Amazon.
Posted inResearch Spotlights
A New Model of Drumlin Formation
Observations from the surge-type glacier Múlajökull in Iceland underpin new modeling results that suggest the glacier’s drumlins grow during quiet intervals of normal flow between glacial surges.