如何在R语言中根据经纬度自动获取MODIS瓦片？

Hey Geoff, great question—automating MODIS tile lookup without a GUI is totally doable, and I’ve got two reliable approaches for you:

1. Stick with the MODIS package (no GUI required) #

You might not have noticed, but the getTile() function actually has an interactive parameter that lets you skip the GUI entirely. Just pass your latitude/longitude values directly and set interactive = FALSE:

library(MODIS)

# Example coordinates (replace with your own)
target_lat <- 40.7128
target_lon <- -74.0060

# Fetch tile info without GUI
tile_result <- getTile(lat = target_lat, lon = target_lon, interactive = FALSE)

# Extract h/v tile numbers
tile_h <- tile_result@tileH
tile_v <- tile_result@tileV

cat("Your MODIS tile: h", tile_h, "v", tile_v, sep = "")

This returns a MODISextent object, and you can pull the horizontal (tileH) and vertical (tileV) tile IDs from it directly. Super straightforward if you’re already using the MODIS package.

2. Calculate tiles manually (no package dependencies) #

If you want to avoid relying on the MODIS package altogether, you can compute the tile numbers using the official MODIS grid rules. The sinusoidal grid splits the globe into 36 horizontal tiles (h00–h35) and 18 vertical tiles (v00–v17), and the math is simple:

get_modis_tile <- function(lat, lon) {
  # Convert negative longitudes to 0-360 range
  lon_360 <- ifelse(lon < 0, lon + 360, lon)
  
  # Calculate horizontal tile (each covers 10 degrees of longitude)
  tile_h <- floor(lon_360 / 10)
  
  # Calculate vertical tile (each covers 10 degrees of latitude, starting from north)
  tile_v <- floor((90 - lat) / 10)
  
  # Format with leading zeros for consistency (e.g., h02 instead of h2)
  tile_h <- sprintf("h%02d", tile_h)
  tile_v <- sprintf("v%02d", tile_v)
  
  return(paste(tile_h, tile_v))
}

# Test with example coordinates
get_modis_tile(lat = 40.7128, lon = -74.0060)
# Output: "h29 v05"

This works for most standard MODIS products. Just note that if you’re working with a non-standard tile system, you’ll need to adjust the formula, but this covers 99% of use cases.

内容的提问来源于stack exchange，提问作者Geoff