Graph-based multivariable bias correction of meteorological variables under strict temporal validation in a western mediterranean region prone to atmospheric instability
Journal of Computational Science, article 102988, 2026
← Research programme
03
Research area
Environmental data quality
The effect of missingness, observational uncertainty, aggregation and measurement inconsistencies on learned models and the decisions built on them.
Rationale
The research problem.
Sensor and reference data are part of the model-selection system, not neutral inputs. This line examines how missingness, substitution, spatial mismatch and measurement degradation propagate into evaluation results and deployment recommendations.
Research questions
Questions that guide this area.
- 01
Which data-quality failures are credible for a particular deployment?
- 02
How sensitive are reported conclusions to aggregation and reference choice?
- 03
Can provenance and perturbation assumptions be made auditable?
Methods, systems and sources
Concrete research elements.
- Reference perturbation
- Missing-data analysis
- Provenance
- Sensitivity analysis
- Data quality
- Uncertainty
- Provenance
Evidence
Publications in this area.
Out-of-sample statistical correctability limits under an uncertain operational reference: the case of IMERG sub-daily areal precipitation extremes
Stochastic Environmental Research and Risk Assessment, vol. 40, no. 8, article 202, 2026
A practical decision-support system for robust post-training model selection in spatiotemporal forecasting
Knowledge-Based Systems, article 116673, 2026
Spatio-temporal graph neural network for inter-city air quality forecasting
International Journal of Environmental Science and Technology, vol. 23, no. 1, article 63, 2026