如何通过R语言获取WorldClim数据，并保存为nc文件

Explaining WorldClim data download in R

The user's question asks how to obtain WorldClim data through R and save it as an NC (netCDF) file. To answer: I’ll recommend using the "raster" or "terra" package in R to download WorldClim data. For instance, the 'getData()' function in the "raster" package can be used to download the data at a specific resolution. Example code:

R
library(raster)
wc <- getData("worldclim", var="bio", res=10)

This will download 19 bioclimatic variables at a 10-minute resolution.

Saving WorldClim data as .nc

To save WorldClim data as an NC file, I recommend using the "raster" package’s writeRaster function with the format “CDF” (NetCDF file). The code would look like this:

R
library(raster)
wc <- getData("worldclim", var="bio", res=10)
wc_stack <- stack(wc)
writeRaster(wc_stack, "worldclim_data.nc", format="CDF", overwrite=TRUE)

Alternatively, the "terra" package can be used, though I'll need to verify if it has a writeCDF function. You can use the resolution options like 10, 5, or 2.5.

Reasoned for 15 seconds

Clarifying user request for nc file

The user is asking for a code to generate an nc file for the region between 9°S to 45°S and 111°E to 155°E. This means the latitude ranges from -9 to -45 and the longitude from 111 to 155. I need to be careful with coordinates since "9°S" and "45°S" refer to southern hemisphere with negative latitudes. I'll proceed with this understanding and provide the code based on R, which can handle these latitude and longitude ranges to extract and save the subset data.

Generating the nc file

The bounding box for the region is set as follows: Longitude 111 to 155 and Latitude -45 to -9 (since 9°S is the northern boundary and 45°S is the southern). We need to carefully handle this by using the crop function in R with the correct extent. The plan is to download the WorldClim data, crop it using extent(), and save it as an .nc file. I’ll also double-check the coordinate reference system—should be WGS84, so no further changes are needed. Here's the code to do all of this!

Reasoned for 43 seconds

Extracting required data

The user only needs annual average temperature and total precipitation data, which corresponds to specific variables in the WorldClim dataset. The WorldClim "bio" set includes 19 variables, and for this query, bio1 is the Annual Mean Temperature, and bio12 is the Annual Precipitation. To meet the user’s need, I’ll modify the code to extract only these two layers from the WorldClim data, then crop and save them into a netCDF file.

Constructing code to extract data

I need to download the WorldClim data using getData to get the bioclim variables as a RasterStack. I realized that getData("worldclim", var="bio", res=10) indeed returns 19 raster layers. Since the user only needs bio1 (Annual Mean Temperature) and bio12 (Annual Precipitation), the code should subset these layers. The first layer corresponds to bio1, and the twelfth layer corresponds to bio12. After subsetting, I’ll crop the region and save it to a netCDF file. Here's the code structure to achieve this.

Reasoned about selecting specific WorldClim variables for 18 seconds

Recalling geodata package functions

The user encountered an error with the getData() function from the raster package. Since getData is now removed and replaced by geodata, I need to explore how to use this new package.

The function geodata::worldclim might be the one to download WorldClim data. I'll check the documentation to confirm the available functions. The goal is to revise the code to download temperature (bio1) and precipitation (bio12) data, restrict the region from 9°S-45°S and 111°E-155°E, and save it as a netCDF file.

Using geodata for WorldClim data

The geodata package supersedes the getData function from the raster package. After installing it via install.packages("geodata"), I realized the function worldclim is available in geodata for retrieving WorldClim data. To retrieve annual temperature (bio1) and precipitation (bio12), we can extract them from the WorldClim dataset after calling worldclim. Since geodata uses the terra package, we can handle SpatRaster objects and save them as netCDF files using writeRaster from terra, like so:

text
bio <- worldclim(var = "bio", res = 10)
selected <- bio[[c("bio1", "bio12")]]

Finalizing R code for WorldClim data

I've designed the final R code using geodata, terra, and ncdf4 to download and process the WorldClim data effectively. First, I ensure the necessary packages are installed and loaded. The selected bio layers (bio1 and bio12) are cropped to a specified region using terra::ext and then saved as a netCDF file using terra::writeRaster. The filetype parameter is correctly set to "netCDF" for saving, ensuring the output is in the proper format. Here's the full R code:

R
# 1. Install needed packages
if (!require(geodata)) install.packages("geodata")
if (!require(terra)) install.packages("terra")
if (!require(ncdf4)) install.packages("ncdf4")
library(geodata)
library(terra)
library(ncdf4)

# 2. Download WorldClim data using geodata
bio <- worldclim(var = "bio", res = 10)

# 3. Extract bio1 (temperature) and bio12 (precipitation)
selected <- bio[[c("bio1", "bio12")]]

# 4. Define region for cropping (longitude 111°E to 155°E, latitude -45° to -9°S)
region_extent <- ext(111, 155, -45, -9)

# 5. Crop selected data to the region_extent
selected_crop <- crop(selected, region_extent)

# 6. Save cropped data as netCDF file
writeRaster(selected_crop, filename = "worldclim_temp_precip.nc", filetype = "netCDF", overwrite = TRUE)

Reasoned for 37 seconds

Checking geodata function names

The user reported an error with the worldclim function, so the issue might be related to a version change. I think the geodata package's recent version could have removed or renamed the worldclim() function. There's a possibility the function is now called something like "worldclim_global." I'll check the geodata package website to see if I can confirm this. I remember that, in version 1.2.2, the function worldclim() was still present, but it may have changed in newer versions.

Confirming geodata function

I’m checking the geodata documentation, and it seems that the function "worldclim_global" is available instead of "worldclim". The geodata package, which replaced "raster::getData", doesn't include "worldclim". It could be that the user is using an outdated or incorrect version, which doesn't match the current package functionalities. "worldclim_global" is for current data, while "worldclim" might be for historical data. I'll need to look into that further and simulate a web search if needed.

Reasoned for 23 seconds