MODIS LAI/FPAR (MCD15)¶

MODIS Leaf Area Index / FPAR (MCD15A2H, MCD15A3H, MOD15A2H, MYD15A2H) Version 6.1 (500 m, 8-day and 4-day composites)

Being replaced

At a glance¶

Note on period: Terra-only MOD15A2H starts on 2000-02-18; Aqua-only MYD15A2H and the combined Terra+Aqua MCD15A2H (8-day) and MCD15A3H (4-day) start on 2002-07-04. A CMR check of the Korean tile (h28v05) on 2026-10-13 found the latest MCD15A2H and MOD15A2H composite was 2026-09-22 to 29, and the latest MCD15A3H composite 2026-09-30 to 10-03. MODIS data collection is expected to end when Terra and Aqua end operations.

Overview¶

Global leaf area index (LAI) and fraction of photosynthetically active radiation (400–700 nm) absorbed by vegetation (FPAR) derived from MODIS on NASA's Terra and Aqua satellites. The main algorithm inverts red and near-infrared surface reflectance against look-up tables built from a 3D radiative transfer model; when that fails, an empirical NDVI-based backup is used. Data come as 500 m sinusoidal tiles in 8-day (combined and single-sensor) and 4-day (combined) composites, and the record back to 2000 makes it a standard reference for ecology, carbon-cycle studies and land-model evaluation. The successor products apply the same algorithm to VIIRS (VNP15, VJ115, VJ215).

What changes with the successor¶

MODIS LAI/FPAR C6.1 (MCD15A2H etc.) → VIIRS LAI/FPAR V2 (VNP15A2H, VJ115A2H, VJ215A2H) Planned¶

Transition: 2027

The successor ports the same Boston University algorithm (3D radiative-transfer look-up tables plus an NDVI backup) to the VIIRS red and near-infrared channels. In a 2012–2016 comparison the global mean differences were small (LAI −0.006, FPAR −0.002), but variable names, quality-bit layout and file format differ, so code must be updated. MODIS collection ends with Terra and Aqua (late 2026 to 2027).

Sources: Transition from MODIS to VIIRS (NASA Earthdata notice, 2025-05-02) · VNP15 User Guide V1.2 (2018) · Xu et al. (2018) — consistency of VIIRS and MODIS LAI/FPAR (Forests) · Román et al. (2024) — continuity between MODIS C6.1 and VIIRS C2 land products (RSE)

Compare the two side by side

MODIS LAI/FPAR C6 → MODIS LAI/FPAR C6.1¶

Transition: 2023-07

A full-record reprocessing that kept the science algorithm and changed only the Terra and Aqua L1B calibration approach and the polarization correction. Format and variable names are the same as C6. Distribution of C6 (006) land products ended on 31 July 2023.

Sources: MOD15 C6.1 User Guide (2020) · Final notice — MODIS Version 6.0 land products decommissioned (2023-07-25)

Available variables¶

From the MOD15 C6.1 User Guide and LP DAAC collection information. The 8-day and 4-day, Terra, Aqua and combined products all use the same SDS names. Values are stored as unsigned 8-bit integers that must be multiplied by the scale factor, and 249–255 are non-vegetation and fill codes (see tips).

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

LAI and FPAR (500 m, 8-day and 4-day composites)¶

The full variable list is in the official documentation.

How to get the data¶

• NASA Earthdata Search (LP DAAC, Earthdata Cloud) — Web search and HTTPS download / earthaccess
  • The link opens MCD15A2H V061. Short names are MCD15A2H, MCD15A3H, MOD15A2H and MYD15A2H, version 061; files are HDF-EOS2 (.hdf) tiles. LP DAAC Data Pool distribution of MODIS data ended on 30 June 2025, so scripts that used Data Pool URLs should switch to Earthdata Cloud (for example via earthaccess).
• AppEEARS (area and point extraction) — Web and API — area (GeoTIFF, NetCDF4) or point (CSV) extraction with optional reprojection
  • Lets you receive just your region on a geographic grid without handling tiles or the sinusoidal projection, which is convenient for new users. Quality-bit lookup tables are included.
• Earthdata catalog (MCD15A2H V061) — Web (documents, citation, layer information)

Example code¶

Downloads MCD15A2H for July 2024 with earthaccess and reads LAI from the central Korean tile (h28v05). The files are HDF4, so pyhdf is used. Stored values of 249 and above are non-vegetation or fill codes and are removed first, and only main-algorithm retrievals (SCF_QC bits 5–7 equal to 0 or 1) are kept.

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

earthaccess.login()  # free NASA Earthdata account
results = earthaccess.search_data(
    short_name="MCD15A2H", version="061",
    temporal=("2024-07-01", "2024-07-31"),
    bounding_box=(126.5, 35.5, 128.5, 37.5),   # (west, south, east, north)
)
files = earthaccess.download(results, "./mcd15a2h")

for f in sorted(map(str, files)):
    if ".h28v05." not in f:          # central Korea tile only
        continue
    sd = SD(f, SDC.READ)
    raw = sd.select("Lai_500m")[:].astype("float32")
    qc = sd.select("FparLai_QC")[:]
    lai = np.where(raw <= 100, raw * 0.1, np.nan)   # 249-255 = non-vegetation / fill
    scf = (qc >> 5) & 0b111                          # 0 or 1 = main (RT) algorithm
    lai[scf > 1] = np.nan
    print(f.split("/")[-1], "mean LAI:", np.nanmean(lai))

Practical tips¶

• LAI and FPAR are stored as uint8. Multiply LAI by 0.1 and FPAR by 0.01, and always remove 249–255 first (255 fill, 254 inland water, 253 barren or sparse vegetation, 252 permanent snow and ice, 251 wetland, 250 urban, 249 unclassified). Averaging them as they are produces absurd LAI values near 25.
• Bits 5–7 of FparLai_QC (SCF_QC) give the algorithm path. 0 (main algorithm) and 1 (main algorithm, saturated) are reliable; 2 and 3 are the empirical NDVI backup; 4 means not produced. Backup retrievals are common during the cloudy monsoon season over Korea, so analyse by quality class.
• Reflectance saturates over dense forest, so LAI is underestimated and highly uncertain there (QC 1, saturation). Take care when comparing peak LAI in tropical or temperate broadleaf forest.
• Bits are read from the right (bit 0). As in the User Guide example, 64 is binary 1000000, so bits 5–7 equal 2 (backup algorithm).
• The tiles use a sinusoidal projection, so pixel edges do not line up with a latitude–longitude grid. Use AppEEARS or GDAL for point extraction or reprojection. Korea spans several tiles, including h28v05 and h29v05.
• According to the MODIS quality team (LDOPE), Terra and Aqua orbit drift has noticeably changed MODIS pixel size, scan gaps and diurnal sampling since late 2024. When interpreting trends over recent years, compare against VIIRS LAI for the same period.
• For long-term analysis use the combined MCD15A2H (from 2002-07). If you need 2000–2002 you have to use Terra-only MOD15A2H, so state that the sampling changes at that point.

Notices and change history¶

References for your paper¶

Dataset citation¶

MCD15A2H V061 (Terra+Aqua, 8-day) — DOI 10.5067/MODIS/MCD15A2H.061

@misc{mcd15a2h_v061,
  author    = {Myneni, Ranga and Knyazikhin, Yuri and Park, Taejin},
  title     = {{MODIS/Terra+Aqua} Leaf Area Index/{FPAR} 8-Day {L4} Global 500m {SIN} Grid {V061}},
  year      = {2021},
  publisher = {NASA Land Processes Distributed Active Archive Center},
  doi       = {10.5067/MODIS/MCD15A2H.061}
}

MCD15A3H V061 (Terra+Aqua, 4-day) — DOI 10.5067/MODIS/MCD15A3H.061

@misc{mcd15a3h_v061,
  author    = {Myneni, Ranga and Knyazikhin, Yuri and Park, Taejin},
  title     = {{MODIS/Terra+Aqua} Leaf Area Index/{FPAR} 4-Day {L4} Global 500m {SIN} Grid {V061}},
  year      = {2021},
  publisher = {NASA Land Processes Distributed Active Archive Center},
  doi       = {10.5067/MODIS/MCD15A3H.061}
}

MOD15A2H V061 (Terra, 8-day) — DOI 10.5067/MODIS/MOD15A2H.061

@misc{mod15a2h_v061,
  author    = {Myneni, Ranga and Knyazikhin, Yuri and Park, Taejin},
  title     = {{MODIS/Terra} Leaf Area Index/{FPAR} 8-Day {L4} Global 500m {SIN} Grid {V061}},
  year      = {2021},
  publisher = {NASA Land Processes Distributed Active Archive Center},
  doi       = {10.5067/MODIS/MOD15A2H.061}
}

Reference papers¶

First year of global MODIS LAI/FPAR (algorithm and early evaluation) — DOI 10.1016/S0034-4257(02)00074-3

@article{myneni2002modislai,
  author  = {Myneni, R. B. and Hoffman, S. and Knyazikhin, Y. and Privette, J. L. and Glassy, J. and Tian, Y. and Wang, Y. and Song, X. and Zhang, Y. and Smith, G. R. and Lotsch, A. and Friedl, M. and Morisette, J. T. and Votava, P. and Nemani, R. R. and Running, S. W.},
  title   = {Global products of vegetation leaf area and fraction absorbed {PAR} from year one of {MODIS} data},
  journal = {Remote Sensing of Environment},
  volume  = {83},
  number  = {1-2},
  pages   = {214--231},
  year    = {2002},
  doi     = {10.1016/S0034-4257(02)00074-3}
}

Collection 6 evaluation, Part 1 (consistency and improvements) — DOI 10.3390/rs8050359

@article{yan2016modisc6,
  author  = {Yan, Kai and Park, Taejin and Yan, Guangjian and Chen, Chi and Yang, Bin and Liu, Zhao and Nemani, Ramakrishna R. and Knyazikhin, Yuri and Myneni, Ranga B.},
  title   = {Evaluation of {MODIS} {LAI/FPAR} Product Collection 6. Part 1: Consistency and Improvements},
  journal = {Remote Sensing},
  volume  = {8},
  number  = {5},
  pages   = {359},
  year    = {2016},
  doi     = {10.3390/rs8050359}
}

Acknowledgements¶

Following the EOSDIS Data Use and Citation Guidance, LP DAAC asks you to cite the DOI of each product you actually used (8-day or 4-day; Terra, Aqua or combined). The DOI of Aqua-only MYD15A2H V061 is 10.5067/MODIS/MYD15A2H.061. For the algorithm, cite Myneni et al. (2002) together with the C6 evaluation paper.

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¶

• MOD15 C6.1 User Guide (PDF)
• MOD15 Algorithm Theoretical Basis Document (ATBD, PDF)
• Earthdata catalog — MCD15A2H V061
• Earthdata catalog — MCD15A3H V061
• MODIS/VIIRS land quality known issues (LDOPE)
• MODIS Land Team news (orbit drift, C7 plans)
• LAADS DAAC — MODIS to VIIRS Transition