DEA Geometric Median and Median Absolute Deviation, Landsat 8/9 Annual (GeoMAD, Australia)

Provider
Geoscience Australia (Digital Earth Australia)
License
CC-BY-4.0 (source)
Format
JSON, TIFF (in sampled files)
Stored at
Amazon S3, s3://dea-public-data/derivative/ga_ls8cls9c_gm_cyear_3 (ap-southeast-2)
Access
Free and public. No account with the data provider is needed.
DOI
10.26186/149305
Category
Earth Observation & Geospatial
Source
https://doi.org/10.26186/149305

About this dataset

Annual cloud-free surface reflectance composites of Australia built from all Landsat 8 and Landsat 9 observations in each calendar year, using the geometric median (Roberts et al. 2017), which preserves the spectral relationships between bands, together with three median absolute deviation layers that describe how variable the surface was during the year. Measured 2026-11-15: version 4-0-0 is tiled on the DEA Australian Albers (EPSG:3577) grid as x/y tiles (e.g. 4-0-0/x25/y41/2013--P1Y/); each tile-year has six 30 m NBART bands (blue, green, red, nir, swir_1, swir_2) plus sdev (spectral), edev (Euclidean) and bcdev (Bray-Curtis) deviation layers and a clear-observation count, all as Cloud Optimised GeoTIFFs with STAC and ODC metadata. Sibling folders hold the same product for Landsat 5 (ga_ls5t_nbart_gm_cyear_3) and Landsat 7 (ga_ls7e_nbart_gm_cyear_3), so the series runs 1986 to 2025. The composites are inputs to land cover mapping, change detection and model calibration. Load with odc-stac or rioxarray. Bucket region ap-southeast-2. Licensed CC BY 4.0; cite Geoscience Australia (2022), doi:10.26186/149305, and Roberts, Mueller and McIntyre (2017, IEEE TGRS).

How to load this dataset in Python

import os
import rasterio
with rasterio.open(os.path.join(os.environ['SCIGANTIC_MOUNT_PATH'], '4-0-0/x68/y42/2025--P1Y/ga_ls8cls9c_gm_cyear_3_x68y42_2025--P1Y_final_nbart_swir_1.tif')) as src:
    print(src.profile)
    band = src.read(1, window=((0, 512), (0, 512)))

Run in a Scigantic notebook on 2026-11-17 against 4-0-0/x68/y42/2025--P1Y/ga_ls8cls9c_gm_cyear_3_x68y42_2025--P1Y_final_nbart_swir_1.tif: rasterio: 1 band(s) 3200x3200 int16, crs EPSG:3577, driver GTiff.

From any machine

import boto3
from botocore import UNSIGNED
from botocore.client import Config
s3 = boto3.client('s3', region_name='ap-southeast-2', config=Config(signature_version=UNSIGNED))
s3.list_objects_v2(Bucket='dea-public-data', Prefix='derivative/ga_ls8cls9c_gm_cyear_3/', Delimiter='/')

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

Data structure

File types in sampled files: JSON, TIFF

Files

Example file: derivative/ga_ls8cls9c_gm_cyear_3/4-0-0/x25/y41/2013--P1Y/ga_ls8cls9c_gm_cyear_3_x25y41_2013--P1Y_final.odc-metadata.yaml

Analyze this dataset without downloading it

On Scigantic, DEA Geometric Median and Median Absolute Deviation, Landsat 8/9 Annual (GeoMAD, Australia) 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 Earth Observation & Geospatial datasets