ITS_LIVE¶

NASA MEaSUREs ITS_LIVE (Inter-mission Time Series of Land Ice Velocity and Elevation) glacier and ice-sheet surface velocity — image pairs, datacubes and regional annual mosaics

Available

At a glance¶

Note on period: The earliest image pairs are Landsat scenes from 1982 (Gardner et al., 2025), but the first years are very sparse and coverage is nearly complete only after 2013. The ITS_LIVE website says "1985-present" and the NSIDC-0776 V2 metadata gives 1984-01-01 to 2022-12-31. On AWS, the V02.1 annual mosaics for Alaska (RGI01A) run from 1982 to 2024 (checked 2026-10-13).

Overview¶

ITS_LIVE, a NASA MEaSUREs project, measures glacier surface displacement with autoRIFT feature tracking on pairs of Landsat 4–9 and Sentinel-2A/B/C optical images and Sentinel-1A/B/C radar images, giving surface velocities for all of the world's land ice. It provides more than 36 million image-pair velocities (L2, netCDF), cloud-optimised Zarr datacubes that stack them in 100 km tiles, and regional annual and climatology mosaics (L3) from an error-weighted least-squares fit, all on a 120 m grid. New scenes are processed within a day, so surges, calving and seasonal speed-up can be followed in near-real time, and the data are widely used for ice-sheet discharge estimates and glacier dynamics.

What changes with the successor¶

ITS_LIVE V1 → ITS_LIVE V2¶

Transition: 2025-07

V2 extends the Landsat-only V1 to Sentinel-1, Sentinel-2 and Landsat 9, refines the grid from 240 m to 120 m and reprocesses the full record with improved autoRIFT feature tracking and error estimates. Image pairs and datacubes on AWS moved to V2 first; the NSIDC regional mosaics followed with V2 (NSIDC-0776) in 2025-07, when V1 was retired.

Sources: NSIDC — NSIDC-0776 V2 release and V1 retirement (2025-07-23) · NSIDC-0776 V2 user guide · Gardner et al. (2025), The Cryosphere · Lei et al. (2022), ESSD (Sentinel-1 V2 processing)

Available variables¶

Image-pair variables follow Lei et al. (2022) and the variable definitions in the ITS_LIVE production code (its_live_production); datacube variables follow the STAC catalogue; annual and climatology variables follow the NSIDC-0776 V2 user guide. vx and vy are components along the map projection axes (x, y), not east–west and north–south.

instantaneous mean accumulated maximum minimum static — instantaneous: value at the time stamp · mean/accumulated/maximum/minimum: over the indicated interval · static: does not change with time

Image-pair velocities (L2 netCDF, one file per pair)¶

Datacubes (Zarr, 100 km tiles, time axis mid_date)¶

Annual mosaics (L3, NSIDC-0776 V2 yearly files)¶

Climatology summary mosaics (L3, NSIDC-0776 V2 2014–2022 files)¶

The full variable list is in the official documentation.

How to get the data¶

• AWS Open Data — its-live-data bucket (us-west-2) — S3 (s3://its-live-data, anonymous) or direct HTTPS download. Image pairs in velocity_image_pair/ (landsatOLI, sentinel1, sentinel2 /v02/), datacubes in datacubes/, annual mosaics in velocity_mosaic/ (v2, v2.1), annual composites in composites/ No login needed
  • The bucket holds more than a billion objects, so the README discourages recursive listing. The datacube index is datacubes/catalog_v02.json (GeoJSON), and an SNS topic announces new files. Antarctic (RGI19A) annual files are about 8–9 GB per year and the climatology file about 21 GB.
• ITS_LIVE STAC API — STAC search (pystac-client) — collections itslive-granules (image pairs), itslive-cubes (datacubes), velocity-mosaics (annual mosaics V2) No login needed
  • Filter image pairs by location, time, platform, date_dt and percent_valid_pixels to get S3 and HTTPS links.
• NSIDC DAAC — NSIDC-0776 V2 regional annual and climatology mosaics — NSIDC Data Access Tool / Earthdata Search / HTTPS / earthaccess / Earthdata Cloud (direct access in us-west-2)
  • Requires a free NASA Earthdata Login. This is the DOI-registered release, covering 1984–2022.
• NSIDC ITS_LIVE web app — Web map for point time series and data selection

Example code¶

Finds the datacube covering one point through STAC and extracts its time series. No login is needed. The coordinates are an example point in Victoria Land, Antarctica, near Terra Nova Bay; change them to your glacier.

# pip install pystac-client xarray zarr fsspec aiohttp pyproj
from pystac_client import Client
from pyproj import Transformer
import xarray as xr

lon, lat = 164.5, -74.4                         # example point (lon, lat)
cat = Client.open("https://stac.itslive.cloud")
items = cat.search(collections=["itslive-cubes"],
                   intersects={"type": "Point", "coordinates": [lon, lat]}).item_collection()
cube = xr.open_zarr(items[0].assets["zarr"].href)
print(list(cube.data_vars)[:10])

epsg = items[0].properties["proj:code"]          # e.g. "EPSG:3031"
x, y = Transformer.from_crs("EPSG:4326", epsg, always_xy=True).transform(lon, lat)
v = cube["v"].sel(x=x, y=y, method="nearest").sortby("mid_date")
print(v.dropna("mid_date").to_series().tail())   # mean speed per pair (m/yr), stamped at the midpoint of the two images

Practical tips¶

• An image-pair value is the mean speed over the acquisition interval. Pick pairs to suit the question: short date_dt for seasonal changes, long pairs or the annual mosaics for long-term means.
• The producers say errors such as v_error and vx_error are formally propagated and usually too small, especially in snowy accumulation zones where clear images are scarce. Treat the errors and count as relative quality indicators.
• V2 velocities are measured in the map projection plane, so scale errors of up to a few percent occur depending on location. Flux-gate calculations need no correction, but comparisons with GPS do.
• The High Mountain Asia (RGI14A) mosaic uses Lambert (EPSG:102027), not UTM. In this region Landsat 4/5 TM velocities have been reported about 5 m/yr slower than Landsat 8 (Dehecq et al., 2019), so take care when comparing 1980s–90s values with recent ones.
• In UTM regional mosaics the mapping variable does not use the CF name, so tools such as Panoply may draw the data a second time on the other side of the globe. The values themselves are fine.
• At glacier edges, icefalls and flow bends, feature tracking can lock onto terrain and return near-zero speeds, and Sentinel-2 is noisy where contrast is low. Annual and climatology products filter much of this, but surges can distort trend and seasonal values, so look at the full time series.
• Some annual files for certain regions and years hold little or no valid data, especially in the 1980s and 1990s. Check count and the quicklook images first.
• The DOI release at NSIDC (NSIDC-0776 V2) stops in 2022, while velocity_mosaic/v2.1 on AWS (file names V02.1) extends to 2024. State the release and path you actually used in your paper.
• Antarctic annual files exceed 8 GB per year. If you need only a few glaciers, find the matching 100 km datacubes through STAC and read just the pixels you need instead of the whole file.

Notices and change history¶

References for your paper¶

Dataset citation¶

Regional annual and climatology mosaics V2 (NSIDC-0776) — DOI 10.5067/JQ6337239C96

@misc{gardner2025itslive0776,
  author    = {Gardner, A. S. and Fahnestock, M. and Greene, C. A. and Kennedy, J. H. and Liukis, M. and L{\'o}pez, L. A. and Scambos, T.},
  title     = {{MEaSUREs} {ITS\_LIVE} Regional Glacier and Ice Sheet Surface Velocities ({NSIDC-0776}, Version 2)},
  year      = {2025},
  publisher = {NASA National Snow and Ice Data Center Distributed Active Archive Center},
  version   = {2},
  doi       = {10.5067/JQ6337239C96}
}

Landsat image-pair velocities V1 (NSIDC-0775, earlier release) — DOI 10.5067/IMR9D3PEI28U

Reference papers¶

Full description of ITS_LIVE V2 (image pairs, datacubes, mosaics) — DOI 10.5194/tc-19-3517-2025

@article{gardner2025itslive,
  author  = {Gardner, Alex S. and Greene, Chad A. and Kennedy, Joseph H. and Fahnestock, Mark A. and Liukis, Maria and L{\'o}pez, Luis A. and Lei, Yang and Scambos, Ted A. and Dehecq, Amaury},
  title   = {{ITS\_LIVE} global glacier velocity data in near-real time},
  journal = {The Cryosphere},
  volume  = {19},
  number  = {9},
  pages   = {3517--3533},
  year    = {2025},
  doi     = {10.5194/tc-19-3517-2025}
}

Paper requested in the auto-RIFT credit line — DOI 10.5194/tc-12-521-2018

@article{gardner2018discharge,
  author  = {Gardner, Alex S. and Moholdt, Geir and Scambos, Ted and Fahnstock, Mark and Ligtenberg, Stefan and van den Broeke, Michiel and Nilsson, Johan},
  title   = {Increased {West} {Antarctic} and unchanged {East} {Antarctic} ice discharge over the last 7 years},
  journal = {The Cryosphere},
  volume  = {12},
  number  = {2},
  pages   = {521--547},
  year    = {2018},
  doi     = {10.5194/tc-12-521-2018}
}

Sentinel-1 image-pair processing (V2) — DOI 10.5194/essd-14-5111-2022

@article{lei2022itslives1,
  author  = {Lei, Yang and Gardner, Alex S. and Agram, Piyush},
  title   = {Processing methodology for the {ITS\_LIVE} {Sentinel-1} ice velocity products},
  journal = {Earth System Science Data},
  volume  = {14},
  number  = {11},
  pages   = {5111--5137},
  year    = {2022},
  doi     = {10.5194/essd-14-5111-2022}
}

Acknowledgements¶

The ITS_LIVE website asks for "Velocity data generated using auto-RIFT (Gardner et al., 2018) and provided by the NASA MEaSUREs ITS_LIVE project (Gardner et al., 20XX)." (replace 20XX with the year of the dataset citation you used). Also credit USGS for Landsat imagery and ESA (Copernicus) for Sentinel imagery.

Information on the datasets used in this study was compiled with the help of the Korean Climate Data Navigator (KCDN) (Sim, 2026; https://doi.org/10.5281/zenodo.23300923).

Links¶

• ITS_LIVE website (NASA JPL)
• NSIDC — NSIDC-0776 V2 regional annual mosaics
• NSIDC-0776 V2 user guide (PDF)
• AWS Open Data — ITS_LIVE
• ITS_LIVE regional velocities known issues (PDF)
• ITS_LIVE tools and notebooks (GitHub)
• This site — GRACE / GRACE-FO (ice sheet and glacier mass change)