The Marble climate informatics platform: data discovery and data access
Bibliographic record
Abstract
Advances in remote sensing and computing infrastructure, and demands of modern climate research are driving the production of new climate datasets at a breathtaking pace. It is increasingly felt by researchers, that the growing volume of climate datasets is challenging to store, analyze or generally "shepherd" through their analysis pipelines. Quite often, the ability to do this is limited to those with access to government or institutional facilities in wealthier nations, raising important questions around equitable access to climate data.The Data Analytics for Canadian Climate Services (DACCS) project has built a cloud based network of federated nodes, called Marble, that allows anyone seeking to extract insights from the large volumes of climate data to undertake their study without concerning themselves with the logistics of acquiring, cleaning and storing data. The aspiration for building this network is to provide a low-barrier entry not only to those working in core climate change research, but also to those involved in climate mitigation, resilience and adaptation work and to policy makers, non-profits, educators and students. Marble is one of the platforms selected to contribute to the 'Open Science Platform' component of the OGC’s OSPD initiative.The user-facing aspect of the platform is comprised of two components: (i) the Jupyter compute environment and (ii) the data server and catalogue. Here, we focus on the latter and present details of the infrastructure, developed on top of proven open-source software and standards (e.g. STAC), that allows for discovery and access of climate datasets stored anywhere on the network by anyone on the network. We will also discuss the publication capability of the platform that allows a user to host their own data on the network and make it quickly available to others.
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 | 0.007 | 0.021 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.006 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.009 | 0.011 |
| Open science | 0.005 | 0.009 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.025 | 0.034 |
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".