Question

我目前有一个气温平均值的rasterstack，我想通过平均值进行趋势或线性回归，这样我最终可能会找到线性回归的斜率。我的简化脚本如下：

# Temp Annual Averages
# (will be) all data

library(raster)
library(rasterVis)

path<-"/net/nfs/merrimack/raid/Northeast_US_Downscaling_cmip5/"
vars = c("tasmin","tasmax")
mods = c("ACCESS1-0", "ACCESS1-3",
     "bcc-csm1-1-m", "bcc-csm1-1")
      #, "CanESM2", "CCSM4", "CESM1-BGC", "CESM1-CAM5", "CMCC-CM",
      #"CMCC-CMS", "CNRM-CM5", "CSIRO-Mk3-6-0", "FGOALS-g2", "GFDL-CM3", 
      #"GFDL-ESM2G", "GFDL-ESM2M", "HadGEM2-AO", "HadGEM2-CC", "HadGEM2-ES",
      #"inmcm4", "IPSL-CM5A-LR", "IPSL-CM5A-MR", "MIROC5", "MIROC-ESM-CHEM", 
      #"MIROC-ESM", "MPI-ESM-LR", "MPI-ESM-MR", "MRI-CGCM3", "NorESM1-M")
scns = c("historical")

#vars (v) = variables, mods (m) = models, scns (s) = scenarios
for (iv in 1:2){
  for (im in 1:4){
    for (is in 1:1){
      for(iy in 1980:1983){
        loop = paste("variable = ", iv,";  model = ",im,";  scenario = ",is,";  
year = ",iy, sep=" ")
          print(loop)
          #Tells us clearly which model, variable, scenario, and year 
          # being looped through each time
        full<-paste(path, vars[iv], "_day_", mods[im], "_", scns[is], 
"_r1i1p1_", iy, "0101-", iy, "1231.16th.nc", sep="")
        # this will print out 
            #/net/nfs/merrimack/raid/Northeast_US_Downscaling_cmip5/NameOfFiles.nc

        # this line will print the full file name
        # This creates character string with name of file we want
        print(full)

        # 1. use the brick function to read the full netCDF file.
        # note: the varname argument is not necessary, but if a file has 
       # multiple variables, brick will read the first one by default.
       # brick function part of the raster package, brick is giant 3 
       # dimensional matrix. full references the full file pack
         air_t<-brick(full, modname=mod[im])
        print(dim(air_t))
   #     print(head(air_t))
        #Use the calc function to get an average for each grid cell over the 
         #entire year
         #calc -- calculate something on brick, fun can equal mean, max, or 
         #min (can define own function here-has to be a function of a single vector)
         # na.rm = T to remove and ignore NA values
        annual_ave_t<-calc(air_t, fun = mean, na.rm = T)
        print(dim(annual_ave_t))
        if(iy == 1980){
          annual_ave_stack = annual_ave_t
        }else{
          annual_ave_stack<-stack(annual_ave_stack, annual_ave_t)
        }  # end of if/else 
      }   # end of year loop
      #if 2006, this is first average, else (otherwise) layer the average onto 
      #all other averages.
      #can create empty stack and define each number to each layer of the stack

      # use calc function to get a trend (linear) 
      # from the annual_ave_stack 
      time <- 1:nlayers(annual_ave_stack)
      print(length(time))

#FIND LINEAR REGRESSION THROUGH RASTERSTACK OF AVERAGES


 #Plot the average annual air temp. Layer 1 shows y-intercept, Layer 2 shows slope
  levelplot(annual_ave_stack, margin = F, package = "raster")

    }   # end of scenario loop
  }   # end of model loop
}  # end of variable loop

希望这是有道理的。我希望这一系列的脚本能够进入所有大写评论的位置。

Answer 1

您最好提供可重复的示例;但您可以使用

计算trend

lm_fun = function(x) { if (is.na(x[1])){ NA } else { m = lm(x ~ time); summary(m)$coefficients[2] }} # slope

annual_ave_stack.slope= calc(annual_ave_stack, lm_fun)

通过rasterstack平均值计算线性回归和斜率

1 个答案: