Initialized Earth System prediction from subseasonal to decadal timescales
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Initialized Earth System predictions are made by starting a numerical prediction model in a state as consistent as possible to observations and running it forward in time for up to 10 years. Skilful predictions at time slices from subseasonal to seasonal (S2S), seasonal to interannual (S2I) and seasonal to decadal (S2D) offer information useful for various stakeholders, ranging from agriculture to water resource management to human and infrastructure safety. In this Review, we examine the processes influencing predictability, and discuss estimates of skill across S2S, S2I and S2D timescales. There are encouraging signs that skilful predictions can be made: on S2S timescales, there has been some skill in predicting the Madden–Julian Oscillation and North Atlantic Oscillation; on S2I, in predicting the El Niño–Southern Oscillation; and on S2D, in predicting ocean and atmosphere variability in the North Atlantic region. However, challenges remain, and future work must prioritize reducing model error, more effectively communicating forecasts to users, and increasing process and mechanistic understanding that could enhance predictive skill and, in turn, confidence. As numerical models progress towards Earth System models, initialized predictions are expanding to include prediction of sea ice, air pollution, and terrestrial and ocean biochemistry that can bring clear benefit to society and various stakeholders. Initialized climate predictions offer distinct benefits for multiple stakeholders. This Review discusses initialized prediction on subseasonal to seasonal (S2S), seasonal to interannual (S2I) and seasonal to decadal (S2D) timescales, highlighting potential for skilful predictions in the years to come.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.010 | 0.009 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it