Ensemble precipitation down-selection methods using Continuous Ranked Probability Score (CRPS): Balancing accuracy and spread under computational constraints
Bibliographic record
Abstract
        An ensemble down-selection method is proposed to improve the analysis and forecast with a small ensemble,  reducing computational needs. A usual problem with ensemble down-selection is that, despite of the reduction of forecast error, ensemble spread sharply decrease. To limit ensemble spread collapse, this study introduces two variations of a novel down-selection method seeking to minimize the sub-ensemble’s Continuous Ranked Probability Score (CRPS), thereby preserving ensemble spread while minimizing forecast error. The approaches are then tested with a regional-scale model whose precipitation forecast we seek to improve. The precipitation forecast performance of sub-ensembles obtained by these CRPS-based methods is evaluated against the full ensemble, and 100 randomly down-selected sets using various verification metrics measuring precipitation forecast skill. Results demonstrate that the CRPS-based sub-ensembles improve probabilistic forecast accuracy by achieving lower CRPS with the lowest Root Mean Square Error (RMSE) value, especially for short forecasts, without increasing false alarms. Additionally, the Brier Score shows improved forecasts, while Fraction Skill Score (FSS) confirms the improved spatial accuracy in light precipitation. These findings suggest that CRPS-based methods are viable sub-ensembling approaches for balancing accuracy, reliability, and computational efficiency in operational forecasting. By preserving ensemble spread, they improve the sub-ensemble's capacity to represent uncertainty, offering a practical and robust solution for ensemble down-selection.
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.
How this classification was reachedexpand
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 | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".