Predictia

METACLIP Project overview

The objective of METAdata for CLImate Products (METACLIP) is to encode the needed metadata to ensure the traceability and reproducibility of any kind climate product (data files, plots, maps…), thus requiring a comprehensive framework to track the operations undertaken through often complex data workflows. Metaclip is based on RDF and follows a modular approach (different modules for different thematic fields) implementing semantics. The current state of metaclip includes the following modules: Data sources, Transformations, (Model Output) Adjustment, Seasonal Forecast Verification and Climate Validation. For each of these modules a both a visual schema and full documentation of the vocabulary elements and their dependencies is provided. Moreover, the vocabulary definition is provided in a OWL file and a sample case-study is given (in a RDF file) to illustrate the usage of the vocabulary. The modules are described in the following:

  1. Data sources: The provenance of the input data need to be clearly identified (source, version, model documentation etc.). These applies to any type of data source (observations, operative/retrospective forecasts, reanalysis, climate projections…). Links to [vocabulary doc] [OWL file] [Visual schema]
  2. Transformations: Any operations transforming the original data source that do not entail a second dataset (may entail a different subset of the same dataset though; e.g. temporal/spatial aggregation, ensemble means, spatial interpolation/regridding, calculation of climate indices and anomalies …). Links to [Visual schema] [vocabulary doc] [OWL file] [usage example (RDF file)]
  3. (Model Output) Adjustment: Bias adjustment and downscaling techniques. Links to [Visual Schema] [vocabulary doc] [OWL file] [usage example (RDF file)]
  4. Seasonal Forecast Verification: Standard performance measures for different aspects of seasonal forecast quality (association, reliability, etc.). This vocabulary has been developed for the QA4Seas Project (Quality Assurance for Multi-model Seasonal Forecast Products). Links to [Visual schema] [vocabulary doc] [OWL file] [usage example (RDF file)]
  5. Climate Validation: Validation scores for different aspects (marginal, temporal, extremes, etc.) typically used in climate studies. This vocabulary is being developed in the framework of the COST Action VALUE (Validating and Integrating Downscaling Methods for Climate Change Research). Links to [Visual schema] [vocabulary doc] [OWL file] [usage example (RDF file)]
  6. Products: Final products of the workflow as maps, time series, plumes or verification plots. [In construction]

A workflow illustrating the generation of a common seasonal verification product (AUC) is given in the figure below.


Example Figure. Schematic representation of a data workflow to generate a verification map (Area under the ROC Curve, based on tercile categories) of a seasonal forecasting system (ECMWF System-4) of mean JJA global temperature. The verifying reference is the ECMWF ERA-Interim reanalysis. All the necessary metadata for the reconstruction of the figure is encoded in RDF (Resource Description Framework) and embedded in the final outcome (in this case a jpeg file, but any other type may serve as well). See this demo for a graphical representation of the metadata schema associated with this figure.


Key strengths

The main advantages of The Climate Ontology approach are next summarised:

Linked International Projects and Initiatives

The Climate Ontology Project is aligned with currently on-going initiatives facing the problems of data provenance and metadata encoding of climate products: