MODIS Land Cover (MCD12)¶

MODIS/Terra+Aqua Land Cover Type Yearly L3 Global, Collection 6.1 (annual global land cover classification, 500 m tiles and 0.05° grid)

Being replaced

At a glance¶

Note on period: Annual maps from 2001. On 2026-10-13 NASA CMR held MCD12C1 (0.05°) through the 2025 map (produced 2026-09-26) and MCD12Q1 (500 m tiles) through the 2024 map (produced July 2025); no 2025 MCD12Q1 tiles were found yet. The LP DAAC and Earthdata product pages still state an extent of 2001–2022, but files for later years are online.

Overview¶

NASA's annual global land cover maps, produced by supervised classification of Terra and Aqua MODIS reflectance. Each file carries five legends — IGBP, UMD, LAI, BIOME-BGC and PFT — plus three FAO LCCS-based layers for land cover, land use and surface hydrology, which makes the product a common source of vegetation types and plant functional types for land surface models and of regional land cover statistics. There is a 500 m tiled product (MCD12Q1) and a 0.05° aggregate (MCD12C1) that keeps the IGBP, UMD and LAI legends together with sub-grid class fractions.

What changes with the successor¶

MODIS Land Cover C6 (V006) → MODIS Land Cover C6.1 (V061)¶

Transition: 2022

A full reprocessing from 2001 that keeps the science algorithm and picks up the improved MODIS L1B calibration. The file format and layers are unchanged, and two faulty classes in the UMD legend were cleaned up.

Sources: MODIS C6.1 Land Cover Type Product User Guide (2022-04-20) · NASA Earthdata — MCD12Q1 V061 product page

Available variables¶

HDF4 (HDF-EOS) science dataset (SDS) names as given in the LP DAAC layer table and NASA CMR variable records. All layers are 8-bit unsigned integers with fill value 255 and no scale factor. Classification layers hold class codes, so never average or interpolate them; decode them with the legend tables (Tables 3–10 of the user guide).

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

MCD12Q1 legends (500 m, classification for the year)¶

MCD12Q1 FAO LCCS layers (500 m, classification for the year)¶

MCD12Q1 quality and mask (500 m)¶

MCD12C1 0.05° aggregate (IGBP, UMD and LAI, classification for the year)¶

The full variable list is in the official documentation.

How to get the data¶

• NASA LP DAAC — Earthdata Cloud (MCD12Q1 V061) — HTTPS download / earthaccess / direct AWS S3 access (us-west-2)
  • Files live under https://data.lpdaac.earthdatacloud.nasa.gov/lp-prod-protected/MCD12Q1.061/ and require a free NASA Earthdata login. Distribution of MODIS data through the legacy LP DAAC Data Pool (e4ftl01) ended on 30 June 2025.
• NASA LP DAAC — Earthdata Cloud (MCD12C1 V061) — HTTPS download / earthaccess
  • One global file per year (the 2025 file is about 1.2 GB, HDF4).
• NASA Earthdata Search — Web search and download
  • The MCD12C1 collection is C2484078896-LPCLOUD.
• AppEEARS — Web and API area/point extraction (GeoTIFF or NetCDF output)
  • Supports MCD12Q1.061 (checked October 2026). Handy for mosaicking tiles and reprojecting on the fly.

Example code¶

Downloads the 2024 MCD12Q1 tiles covering the Korean Peninsula (h27v05, h28v05, etc.) with earthaccess and reads the IGBP layer with pyhdf. The first run asks for your NASA Earthdata username and password.

# pip install earthaccess pyhdf numpy
import earthaccess
import numpy as np
from pyhdf.SD import SD, SDC

earthaccess.login()
results = earthaccess.search_data(
    short_name="MCD12Q1", version="061",
    temporal=("2024-01-01", "2024-12-31"),
    bounding_box=(124, 33, 132, 39),   # (west, south, east, north)
)
files = earthaccess.download(results, "mcd12q1_2024")

for f in files:
    hdf = SD(str(f), SDC.READ)
    igbp = hdf.select("LC_Type1")[:]          # 2400 x 2400, classes 1–17, 255 = fill
    qc = hdf.select("QC")[:]
    valid = igbp != 255
    classes, counts = np.unique(igbp[valid], return_counts=True)
    print(f, dict(zip(classes.tolist(), counts.tolist())))

# 0.05° aggregate: short_name="MCD12C1", version="061"
# Layer names, e.g. "Majority_Land_Cover_Type_1", "Land_Cover_Type_1_Percent"

Practical tips¶

• MCD12Q1 comes in Sinusoidal-projection tiles. Overlaying it on latitude–longitude data requires reprojection, and because the values are class codes you must use nearest-neighbour resampling; bilinear interpolation invents class codes that do not exist.
• The IGBP water code differs between products: 17 in MCD12Q1 LC_Type1, 0 in MCD12C1 Majority_Land_Cover_Type_1. Harmonize the legends before mixing the two.
• There is an official warning that the 2021-and-later maps were made without updated training samples and show larger year-to-year swings, especially among grassland, shrubland and forest. Do not read class differences between recent years as real change.
• The user guide discourages detecting change by differencing annual maps because labels fluctuate from year to year. For change studies, pool several years or use a dataset designed for change detection.
• Known misclassifications include permanent sea ice and shaded glaciers mapped as water, small tropical fields labelled as natural vegetation, and some temperate evergreen needleleaf forests (e.g. Japan) labelled as evergreen broadleaf. Wetlands are under-represented.
• Use the QC layer to screen post-processed pixels (switched by the water mask, relabelled snow/ice, backfilled with the pre-stabilization result). 255 is the fill value.
• Terra and Aqua orbit drift has shifted observation times, and both satellites are projected to end science operations in 2027. Keep the changing inputs in mind when adding years after 2022 to long-term trend analyses.
• Scripts that still use the old Data Pool host (e4ftl01.cr.usgs.gov) stopped working after 30 June 2025. Switch to earthaccess or data.lpdaac.earthdatacloud.nasa.gov URLs.

Notices and change history¶

References for your paper¶

Dataset citation¶

MCD12Q1 V061 (500 m) — DOI 10.5067/MODIS/MCD12Q1.061

@misc{mcd12q1_v061,
  author    = {Friedl, Mark and Sulla-Menashe, Damien},
  title     = {{MODIS/Terra+Aqua} Land Cover Type Yearly {L3} Global 500m {SIN} Grid {V061}},
  year      = {2022},
  publisher = {NASA Land Processes Distributed Active Archive Center},
  doi       = {10.5067/MODIS/MCD12Q1.061}
}

MCD12C1 V061 (0.05°) — DOI 10.5067/MODIS/MCD12C1.061

@misc{mcd12c1_v061,
  author    = {Friedl, Mark and Sulla-Menashe, Damien},
  title     = {{MODIS/Terra+Aqua} Land Cover Type Yearly {L3} Global 0.05Deg {CMG} {V061}},
  year      = {2022},
  publisher = {NASA Land Processes Distributed Active Archive Center},
  doi       = {10.5067/MODIS/MCD12C1.061}
}

Reference papers¶

MODIS Collection 6 land cover algorithm paper — DOI 10.1016/j.rse.2018.12.013

@article{sullamenashe2019mcd12,
  author  = {Sulla-Menashe, Damien and Gray, Josh M. and Abercrombie, S. Parker and Friedl, Mark A.},
  title   = {Hierarchical mapping of annual global land cover 2001 to present: The {MODIS} Collection 6 Land Cover product},
  journal = {Remote Sensing of Environment},
  volume  = {222},
  pages   = {183--194},
  year    = {2019},
  doi     = {10.1016/j.rse.2018.12.013}
}

Acknowledgements¶

LP DAAC data are shared without restriction under the NASA EOSDIS data use and citation guidance; cite the DOI of the product you used (MCD12Q1 or MCD12C1) with an access date. Cite Sulla-Menashe et al. (2019) for the algorithm, and state in the paper that you used Collection 6.1 (V061) and which legend (IGBP, etc.).

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¶

• LP DAAC — MCD12Q1 V061 product page (layer table)
• MODIS C6.1 Land Cover Type User Guide (legends, QC, known issues)
• MCD12 Algorithm Theoretical Basis Document (ATBD)
• MODIS Land validation status — MCD12
• MODIS Land news (Terra and Aqua orbit drift, Collection 7 plans)
• LP DAAC — Aqua MODIS acquisition plan through August 2026