EnviStor logo · FIU

Data

This page lists FIU EnviStor's environmental datasets and their FAIR (Findable, Accessible, Interoperable, Reusable) publication status. Every number below is checked against the live Dataverse record — not copied from a slide or a report. See how the platform and its human-in-the-loop agentic-AI workflow behind these numbers work on the Home page and in the PEARC 2026 paper ↗.

Last verified: 2026-07-25 How we verify this →

5

Published projects

with a RELEASED Dataverse DOI

16

Released DOIs

across those 5 projects

3

In the publication pipeline

real data, DOI pending

43.6 GB

Data volume in Dataverse

excludes bulk corpora hosted on Pelican/OSDF (e.g. BirdNest’s 113 GB image set)

Data at a Glance

How our published datasets stack up — by project, by size, by publication stage — and how that compares to the rest of FIU's Dataverse over time. We're not drawing conclusions here; the charts just show what's there.

Number of released Dataverse DOIs per project. 🔍 Click to enlarge
Number of released Dataverse DOIs per project.
Size of each project’s Dataverse record, log scale — bulk corpora hosted on Pelican/OSDF instead of Dataverse are not included. 🔍 Click to enlarge
Size of each project’s Dataverse record, log scale — bulk corpora hosted on Pelican/OSDF instead of Dataverse are not included.
Where each tracked project currently stands: published, migrated-no-DOI, in-pipeline, or raw holdings. 🔍 Click to enlarge
Where each tracked project currently stands: published, migrated-no-DOI, in-pipeline, or raw holdings.
Subject-area mix across the whole dataverse.fiu.edu instance (122 datasets), with Earth & Environmental Sciences — our projects’ home subject — highlighted. 🔍 Click to enlarge
Subject-area mix across the whole dataverse.fiu.edu instance (122 datasets), with Earth & Environmental Sciences — our projects’ home subject — highlighted.
Our tracked projects hold 16 of the ~122 datasets currently in FIU’s Research Data Portal (Dataverse instance launched 2020). 🔍 Click to enlarge
Our tracked projects hold 16 of the ~122 datasets currently in FIU’s Research Data Portal (Dataverse instance launched 2020).
Monthly published-dataset counts, dataverse.fiu.edu, full history since 2019 — our tracked DOIs highlighted. Of the 26 datasets published FIU-wide between 2026-03 and 2026-06 (the months our tracked DOIs appear in), 15 were from these tracked projects. 🔍 Click to enlarge
Monthly published-dataset counts, dataverse.fiu.edu, full history since 2019 — our tracked DOIs highlighted. Of the 26 datasets published FIU-wide between 2026-03 and 2026-06 (the months our tracked DOIs appear in), 15 were from these tracked projects.

Spatial coverage

Where the data underlying these DOIs comes from — point/region markers, not exact georeferenced footprints. Click a marker to jump to that project's card below.

Published (has a DOI) In progress

Published

Every project below has at least one Dataverse record with the status RELEASED, confirmed live against the Dataverse API (not a report describing it).

National Park Service · FIU GIS Center

  • · 15,759 aerial images
  • · 161,744 AI-assisted annotations
  • · 117 helicopter surveys (2010–2024)

Dataverse record (~30 MB) holds the model/metadata package; the full 113 GB image corpus is hosted on Pelican/OSDF.

FIU GIS Center

  • · 844 geospatial layers
  • · 7 Dataverse DOIs (1 master + 6 regional)
  • · ~38 GB in Dataverse

Legacy FIU GIS Center geoportal, migrated to the EnviStor FAIR pipeline (2025–2026).

Obey Research Group

  • · 2,452 monitoring stations
  • · 8,557 files, 1985–2024
  • · ~4.4 GB in Dataverse

6 Dataverse DOIs (1 master + 5 station types: water, groundwater, gate, pump, rainfall), all RELEASED.

Everglades National Park DTM

✅ Verified

Daniel Gann, Gann Spatial Ecology Lab

  • · 1 bias-adjusted lidar-derived GeoTIFF
  • · 2 m resolution
  • · ~1.1 GB

The simplest end-to-end case: single file, single PI, fully FAIR-published with checksums.

Paulo Olivas, sensor PI

  • · 5 instruments (PurpleAir ×3, Tempest, Pandora 256)
  • · Air quality + atmospheric chemistry, growing daily

FIU Miami campus. A companion coastal-buoy pipeline is validated on test data and moving toward production with the NSF CREST Center for Aquatic Chemistry & Environment.

This work is described in more depth in "Exploring Robust Multi-Agent Workflows for Environmental Data Management" (PEARC 2026) ↗ and "A Scalable Framework for Heterogeneous Environmental Data Management Using Smart Data Pipeline" (PEARC 2025) ↗ — see the full list on Publications.

In progress

Real data, real pipelines — just not a released Dataverse DOI yet. No external link, because there's nothing published to link to.

FIU GIS Center

4,025 files · ~157 GB (1924–1954)

Migrated + already served on maps.fiu.edu — DOI not yet requested

LOCA2-CMIP6 Florida (climate projections)

🔶 In progress

Obey Research Group

~3.1 TB, full scope

Sequencing into 3 variable-specific Dataverse records (precipitation → Tmax → Tmin); first pilot staged

Global Sea Level Change

🔶 In progress

Obey Research Group

~1.2 TB · GTSM/ERA5 reanalysis + CMIP6 projections

Planned as 1 master record + product-specific sub-records

Coastal Buoy Pilot

🔶 In progress

Obey Research Group · NSF CREST Center for Aquatic Chemistry & Environment

Pipeline mechanics validated on real test data

Formal onboarding of live buoy readings pending CREST Center collaboration

The Obey Research Group's two in-progress climate products above are part of a 54.7 TB / 369,857-file NAS holdings scan — most of that scan is raw storage, not yet slated for publication. See the full project list on Projects & Data Showcase.

How we verify these numbers

Every number on this page traces to one of these sources, checked in order of strength (strongest first):

  1. Live Dataverse record — the actual RELEASED dataset, queried directly via the Dataverse API. Outranks any doc, task, or report describing it.
  2. Dataverse instance & search metrics — FIU's own Make Data Count / COUNTER-compliant metrics API and search index, for platform-wide context (total datasets, publication dates, subject mix).
  3. Project & task records — structured facts from the platform's project management system.
  4. Curated knowledge base & message history — lowest-reliability tier; treated as "someone reported this," re-confirmed against a live system before being shown here.

The charts above come from fair_publication_stats.ipynb, a notebook that queries the live Dataverse record and Dataverse's own metrics/search APIs directly, so they stay in sync with the same numbers above. This page is regenerated on demand (not on a fixed schedule) whenever a project publishes something new. See AI Governance & Data Access for how the underlying agentic publication workflow is reviewed and gated by named humans before anything goes public.