Applying machine learning for drought prediction using a large ensemble of climate simulations
Bibliographic record
Abstract
There is strong scientific and social interest to understand the factors leading to extreme events in order to improve the management of risks associated with hazards like droughts. Recent events like the summer 2018 drought in Germany already had severe und unexpected impacts, e.g. forest fires and crop failures; in order to increase preparedness robust prediction tools are urgently required. In this study, machine learning methods are applied to predict the occurrence of a drought with lead times of one to three months. The approach takes into account a list of thirty atmospheric and soil variables as predictor input parameters from a single regional climate model initial condition large ensemble (CRCM5-LE). The data was produced the context of the ClimEx project by Ouranos with the Canadian Regional Climate Model (CRCM5) driven by 50 members of the Canadian Earth System Model (CanESM2) for the Bavarian and Quebec domains. Drought occurrence was defined using the Standardized Precipitation Index. The training and test datasets were chosen from the current climatology (1955-2005) for the Munich and Lisbon subdomain within the CRCM5-LE. The best performing machine learning algorithms managed to obtain a correct classification of drought or no drought for a lead time of one month for around 60 % of the events of each class for the both domains. Explainable AI methods like feature importance and shapley values were applied to gain a better understanding of the trained algorithms. Physical variables like the North Atlantic Oscillation Index and air pressure one month before the event proved to be of high importance for the prediction. The study showed that better accuracies can be obtained for the Lisbon domain, due to the stronger influence of the North Atlantic Oscillation Index on Portugal’s climate.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| 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 source (direct Gemma or distilled Codex), 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".