National Reef Monitoring Network - Global benthic cover data (photoquadrat image annotations) - Darwin Core

Provider
AODN
License
http://creativecommons.org/licenses/by/4.0/
Size
20 objects, 2,097,563 bytes
Records
97,922 rows (exact, full read 2026-10-04); occurrenceID is unique per row
Format
Parquet (in sampled files)
Stored at
Amazon S3, s3://aodn-cloud-optimised/diver_photoquadrat_qc_DwC.parquet (ap-southeast-2)
Access
Free and public. No account with the data provider is needed.
Category
Life Sciences
Source
https://catalogue-imos.aodn.org.au/geonetwork/srv/eng/catalog.search#/metadata/0a65be6d-1c76-49ac-a151-80acf123612c

About this dataset

The National Reef Monitoring Network (NRMN) brings together shallow reef surveys conducted around Australia into a centralised database; the IMOS NRMN Sub-Facility collates, cleans, stores and makes this data available from Reef Life Survey, Parks Australia, the IMOS Understanding of Marine Imagery Sub-Facility and state partners. This dataset contains habitat composition scored from photoquadrats taken by RLS and ATRC divers along 50 m transects on shallow rocky and coral reefs: 20 images per transect, annotated with a point-overlay approach using standardised label schemes, with Squidle+ as the image and annotation platform. This store is the Darwin Core (DwC) occurrence re-expression of the quality-controlled photoquadrat score table: one row per habitat label per transect analysis, with eventID equal to the score table's dataset_id, RLSIdentification / catamiIdentification carrying the original labels, scientificName and scientificNameID resolved against WoRMS, individualCount the annotation point count, plus an H3 index and IMCRA bioregion tags. It holds 97,922 occurrences from 8,575 events, 2008-01 to 2026-03, restricted to the eastern hemisphere (Australia, NZ, Indonesia, PNG, Solomon Islands), partitioned by event year.

How to load this dataset in Python

import os
import pyarrow.parquet as pq
pf = pq.ParquetFile(os.path.join(os.environ['SCIGANTIC_MOUNT_PATH'], 'timestamp=1199145600/polygon=01030000000100000005000000000000000000000000000000008056C0000000000080664000000000008056C00000000000806640000000000080564000000000000000000000000000805640000000000000000000000000008056C0/nrmn.parquet-0.parquet'))
print(pf.metadata)
df = pf.read_row_group(0).to_pandas()
df.head()

Run in a Scigantic notebook on 2026-10-04 against timestamp=1199145600/polygon=01030000000100000005000000000000000000000000000000008056C0000000000080664000000000008056C00000000000806640000000000080564000000000000000000000000000805640000000000000000000000000008056C0/nrmn.parquet-0.parquet: Parquet 2,733 rows, 26 columns, 1 row groups; columns ['class:string', 'genus:string', 'order:string', 'family:string', 'phylum:string', 'eventID:string', 'h3Index:string', 'kingdom:string', 'eventDate:date32[day]', 'taxonRank:string'].

From any machine

# Task: read the schema and dataset identity without touching a data file (_common_metadata at the root)
import pyarrow.parquet as pq, s3fs, ast
fs = s3fs.S3FileSystem(anon=True, client_kwargs={'region_name': 'ap-southeast-2'})
for store in ['diver_benthic_cover_in_situ_qc', 'diver_photoquadrat_score_qc', 'diver_photoquadrat_qc_DwC', 'diver_site_information_qc']:
    schema = pq.read_schema(f'aodn-cloud-optimised/{store}.parquet/_common_metadata', filesystem=fs)
    meta = {k.decode(): v.decode() for k, v in schema.metadata.items()}
    print(store, '|', len(schema), 'columns |', ast.literal_eval(meta['global_attributes']))
    print('   ', ast.literal_eval(meta.get('latitude') or meta.get('decimalLatitude')))

Lists the files with anonymous access. Runs on any machine, with no cloud account.

Data structure

File types in sampled files: Parquet

Fields

classstring; dwc:class; null 42%
genusstring; dwc:genus; null 86%
orderstring; dwc:order; null 50%
familystring; dwc:family; null 83%
phylumstring; dwc:phylum; null 25%
eventIDstring NOT NULL; 8,575 distinct; EQUALS diver_photoquadrat_score_qc.dataset_id (100% of eventIDs are score dataset_ids)
h3Indexstring NOT NULL; H3 cell derived from the coordinates; metadata says resolution 5 but every index begins with '84'…
kingdomstring; Animalia 44,276 | Plantae 16,818 | Chromista 11,353 | Bacteria 1,107; null 24,368 (substrate and unresolved…
eventDatedate32[day] (pandas object of datetime.date, NOT datetime64: wrap in pd.to_datetime); 2008-01-09 to 2026-03-12; equals…
taxonRankstring; Order 32,443 | Genus 10,306 | Superdomain 9,440 ('Biota') | Class 7,153 | Phylum 6,574 | Phylum (Division)…
RLSLineagestring; = score rls_lineage; null 1%
occurrenceIDstring NOT NULL; unique per row (97,922 distinct), a plain integer serial 1330..120560, not a persistent global id
basisOfRecordstring; single value 'MachineObservation'
catamiLineagestring; = score catami_lineage
scientificNamestring; WoRMS-matched name at the resolvable rank ('Ecklonia radiata', 'Corallinales', 'Phaeophyceae', 'Biota'); null…
associatedMediastring; Squidle+ deployment viewer URL https://squidle.org/iframe/api/deployment/<id>?template=...; null 31% (legacy…
decimalLatitudedouble NOT NULL, EPSG:4326; site coordinate at 2 dp (equals diver_site_information_qc latitude for 99.9% of events)…
individualCountint32 (float64 in pandas because of nulls); = score label_count = number of annotation points with this label; null…
decimalLongitudedouble NOT NULL; 105.56 to 174.93 (eastern hemisphere only)
occurrenceStatusstring; single value 'present'

and 8 more

Analyze this dataset without downloading it

On Scigantic, National Reef Monitoring Network - Global benthic cover data (photoquadrat image annotations) - Darwin Core opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Life Sciences datasets