如何使用多边形shapefile按区域提取NetCDF数据帧

Question

I'am trying to extract the variable "swh_ku" from multiple NetCDF files together with their corresponding latitude and longitude values into csv files using a polygon shapefile or it's extent. I'm working with Jason-1 global altimetry swath data but I only need the data for the domain represented by the shapefile. I just need help with some lines of code that would complete the working code bellow so I can extract only the data for the region I'm interested in.

I've tried several software applications such as QGIS, ESA SNAP, Broadview Radar Altimetry Toolbox (BRAT) with no success unfortunately because I couldn't find a way automate the extraction process for the hundreds of NetCDF files. So I resorted to code with which I'm fairly new but managed to get it working after reading other posts. I've tried opening the files as raster or brick to use the #extract or #mask functions because they seem more straightforward but I couldn't manage to work them out.

Link to data: https://drive.google.com/drive/folders/1d_XVYFe__-ynxbJNUwlyl74SPJi8GybR?usp=sharing

library(ncdf4)
library(rgdal)
library(raster)
my_read_function <- function(ncname) {
  setwd("D:/Jason-1/cycle_030")
  bs_shp=readOGR("D:/Black_Sea.shp")
  e<-extent(bs_shp)
  ncfname = ncname
  names(ncin[['var']])
  dname = "swh_ku"
  ncin = nc_open(ncfname)
  print(ncin)
  vars<-(names(ncin[['var']]))
  vars
  lon <- ncvar_get(ncin, "lon")
  nlon <- dim(lon)
  head(lon)
  lat <- ncvar_get(ncin, "lat", verbose = F)
  nlat <- dim(lat)
  head(lat)
  print(c(nlon, nlat))
  sm_array <- ncvar_get(ncin,dname)
  dlname <- ncatt_get(ncin,dname,"long_name")
  dunits <- ncatt_get(ncin,dname,"units")
  fillvalue <- ncatt_get(ncin,dname,"_FillValue")
  dim(sm_array)
  ls()
  sm.slice <- sm_array[]
  sm.vec <- as.vector(sm.slice)
  length(sm.vec)
  lonlat <- expand.grid(lon, lat)
  sm.df01 <- data.frame(cbind(lonlat, sm.vec))
  names(sm.df01) <- c("lon", "lat", paste(dname, sep = "_"))
  head(na.omit(sm.df01), 20)
  csvfile <- paste0(ncname,".csv")  
  write.table(na.omit(sm.df01), csvfile, row.names = FALSE, sep = ",")
}
my_files <- list.files("D:/Jason-1/cycle_030/")
lapply(my_files, my_read_function)  

Answer 1

Looks like your data is not gridded.

library(ncdf4)
library(raster)

bs <- shapefile("Black_Sea.shp")
# simplify so that the data will look better later
bs <- as(bs, "SpatialPolygons")

f <- list.files("cycle_022", pattern="nc$", full=TRUE)

Loop would start here

ncfname = f[1]
dname = "swh_ku"
ncin = nc_open(ncfname)
lon <- ncvar_get(ncin, "lon")
lat <- ncvar_get(ncin, "lat", verbose = F)
sm_array <- ncvar_get(ncin, dname)
xyz <- na.omit(cbind(lon, lat, sm_array))
p <- SpatialPoints(xyz[,1:2], proj4string=crs(bs))
p <- SpatialPointsDataFrame(p, data.frame(xyz))
x <- intersect(p, bs) 

x has the points that intersect with the Black Sea

plot(bs)
points(x)
head(x@data)