CSTools: the MEDSCOPE Toolbox for Climate Forecasts postprocessing
Bibliographic record
Abstract
Climate forecasts need to be postprocessed to obtain user-relevant climate information, to develop and implement strategies of adaptation to climate variability and to trigger decisions. Several postprocessing methods are gathered into CSTools (short for Climate Service Tools) for forecast calibration, bias correction, statistical and stochastic downscaling, optimal forecast combination and multivariate verification, as well as basic and advanced tools to obtain tailored products. Besides an overview of the methods and documentation available in CSTools, a practical example is demonstrated. The objective of this practical example is to postprocess a seasonal forecast with a set of CSTools functions in order to obtain the required data to produce forecasts of mountain snow resources. Quantile mapping bias-correction and RainFARM stochastic downscaling methods are applied to raw seasonal forecast daily precipitation data to derive 1 km resolution fields. Bias-adjusted and downscaled precipitation data are then employed to drive a snow model, SNOWPACK, and generate snow depth seasonal forecasts at selected high-elevation sites in North-Western Italian Alps. The computational resources required by CSTools to process the forecasts will be discussed. This assessment is relevant given the memory requirements for the use case: while seasonal forecast data occupies ~10MB (8 x 8 grid cells, 215 forecast time steps for 30 different initializations with 25 members each), the data post-processed reaches ~1TB (the RainFARM downscaling requires a refinement factor 100 for the SNOWPACK model increasing the spatial resolution to 800 x 800 grid cells and creating 10 stochastic realizations for each ensemble member). In addition to one strategy using conventional loops, startR is introduced as an efficient alternative. startR is an R package that allows implementing the MapReduce paradigm, i.e. chunking the data and processing them either locally or remotely on high-performance computing systems, leveraging multi-node and multi-core parallelism where possible.
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 | 0.007 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.184 | 0.100 |
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".