Download MODIS Data from NASA

Searches NASA's CMR catalog, downloads MODIS HDF files via httr2, extracts the requested band, mosaics tiles, crops to the area of interest, scales values, and saves one GeoTIFF per time step. Output files are named modis_{band}_{YYYY}_{MM}.tif and work directly with process_raster_collection().

Usage

download_modis(
  shapefile,
  start,
  end,
  product = "MOD13A3",
  band = "1 km monthly EVI",
  out_dir = ".",
  username = NULL,
  password = NULL,
  version = "061",
  scale_factor = "auto",
  keep_hdf = FALSE,
  overwrite = FALSE
)

Arguments

shapefile

An sf object defining the area of interest. Reprojected to WGS84 (EPSG:4326) for the NASA query.

start

Character. Start date in any common format (e.g. "2023-01-01").

end

Character. End date in any common format.

product

Character. MODIS product short name (default "MOD13A3" for monthly vegetation indices at 1 km).

band

Character. HDF subdataset name to extract (default "1 km monthly EVI"). Use terra::describe(hdf_file) on a downloaded HDF to see available subdatasets.

out_dir

Character. Output directory for GeoTIFFs (default "."). Created if it does not exist.

username

Character or NULL. NASA Earthdata username. Falls back to Sys.getenv("EARTHDATA_USERNAME").

password

Character or NULL. NASA Earthdata password. Falls back to Sys.getenv("EARTHDATA_PASSWORD").

version

Character. MODIS collection version (default "061").

scale_factor

"auto", numeric, or NULL. When "auto" (default), the scale factor is looked up from a built-in table of known MODIS bands (e.g. 0.0001 for EVI/NDVI, 0.02 for LST). Pass a number to override, or NULL to skip scaling.

keep_hdf

Logical. Keep raw HDF files after processing? Default FALSE (deletes them). When TRUE, saved to {out_dir}/hdf/.

overwrite

Logical. Re-download and reprocess existing files? Default FALSE.

Value

Invisible character vector of output GeoTIFF paths.

Details

Use modis_options() to browse available products.

Examples

if (FALSE) { # \dontrun{
boundary <- sf::st_read("guinea_adm0.geojson")

# download monthly EVI for 2023
paths <- download_modis(
  shapefile = boundary,
  start     = "2023-01-01",
  end       = "2023-12-31",
  out_dir   = "data/evi",
  username  = "my_user",
  password  = "my_pass"
)

# then use the existing raster processing pipeline
result <- process_raster_collection(
  directory    = "data/evi",
  shapefile    = sf::st_read("guinea_adm3.geojson"),
  id_cols      = c("adm0", "adm1", "adm2", "adm3"),
  aggregations = "mean"
)

# download NDVI instead
download_modis(
  shapefile = boundary,
  start     = "2023-01-01",
  end       = "2023-12-31",
  band      = "1 km monthly NDVI",
  out_dir   = "data/ndvi"
)

# keep raw HDF files for inspection
download_modis(
  shapefile = boundary,
  start     = "2023-01-01",
  end       = "2023-12-31",
  out_dir   = "data/evi",
  keep_hdf  = TRUE
)
} # }