從較大的柵格堆棧中創建柵格堆棧的許多子集

Question

這是一個困難的時間概念化。我目前有一個netcdf文件，其中包含1979-2016年的每月氣候平均值。我想從這個較大的光柵堆棧中單獨製作一個光柵堆棧（即從1個光柵堆棧到38）。我不能爲我的生活弄清楚這一點。任何建議都會很棒！數據可以在這裏找到：http://nimbus.cos.uidaho.edu/abatz/DATA/vpd19792016.nc

我迄今爲止開始如下：

library(raster) 
    library(lubridate) 
    library(rgdal) 
    library(tidyverse) 
    library(tools) 

    file = "../data/raw/climate/aet_19792016.nc" 
    file_split <- file %>% 
    basename %>% 
    strsplit(split = "_") %>% 
    unlist 
    var <- file_split[1] 
    year <- substr(file_split[2], start = 1, stop = 4) 
    endyear <- substr(file_split[2], start = 5, stop = 8) 

    start_date <- as.Date(paste(year, "01", "01", sep = "-")) 
    end_date <- as.Date(paste(ifelse(year == endyear, year, endyear), "12", "31", sep = "-")) 
    date_seq <- seq(start_date, end_date, by = "1 month") 
    month_seq <- month(date_seq) 

    nc <- nc_open(file) 
    nc_att <- attributes(nc$var)$names 
    ncvar <- ncvar_get(nc, nc_att) 
    tvar <- aperm(ncvar, c(3,2,1)) 
    proj <- "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0 " 

    raster <- brick(tvar, crs= proj) 
    extent(raster) <- c(-124.793, -67.043, 25.04186, 49.41686) 
    names(raster) <- paste(var, year(date_seq), 
            unique(month(date_seq, label = TRUE)), 
            sep = "_") 
    unstack(raster) 

在此先感謝您的任何建議！

Answer 1

首先，你必須正確地命名基於Years和months光柵堆棧：

library(ncdf4) 
library(raster) 

nc <- nc_open(file) 
nc_att <- attributes(nc$var)$names 
ncvar <- ncvar_get(nc, nc_att) 
tvar <- aperm(ncvar, c(3,2,1)) 
proj <- "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0 " 
r <- brick(tvar, crs= proj) 
extent(r) <- c(-124.793, -67.043, 25.04186, 49.41686) 
# name raster brick based on year and month 
dates <- format(seq(as.Date(paste(1979,'/1/1',sep='')), as.Date(paste(2016,'/12/31', sep="")), by='month'), '%Y%m') 
names(r) <- dates 
r_sub <- subset(r, grep("X1982", names(r))) # subset based on year