How to Extract Point Data from Multiple Subsetted Netcdf Files for Sea Surface Density using terra::rast() and terra::extract functions in R

Question

Description of Data

I downloaded MULTIOBS_GLO_PHY_S_SURFACE_MYNRT_015_013 data for daily sea surface density from 2012 to 2024 from The Copernicus Website. A person can only download 2 years of data at a time; therefore, I have five .nc files ranging over 12 years.

Copernicus Website

https://data.marine.copernicus.eu/product/MULTIOBS_GLO_PHY_S_SURFACE_MYNRT_015_013/download

Desired Output

My aim is to extract the point data from the netcdf files from 2012 to 2024 for the specific survey dates that we did during fieldwork and insert the values as a column in my dataframe.

Issue

I get this error message when I run the line SSSs to select the variable sea surface density from the netcdf files

Error: Not compatible with requested type: [type=character; target=double].

I have successfully extracted chlorophyll-a and sea surface temperature point data from Lm3 AQUA MODIS files using exactly the same method and I don't understand what I'm doing wrong with sea surface salinity.

I can't share the dataframe due to ownership issues

Many thanks if anyone can help

Code

Read in all 5 netcdf files for sea surface density

#SS1: January 1st 2012 to December 31st 2014

sss1<-terra::rast("~/Documents/cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729236771868.nc")
plot(sss1)
sss1<-subset(sss1, 1:1096)

#SS2: January 1st 2015 to December 31st 2017

ss2<-terra::rast("~/Documents/cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729236939325.nc")

#Subset to extract sea surface density only 
ss2<-subset(ss2, 1:1096)

#SS3: January 1st 2018 to December 31st 2020

ss3<-terra::rast("~/Documents/cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729237129644.nc")

#Subset to extract sea surface density only 
ss3<-subset(ss3, 1:1096)

#SS4: January 1st 2021 to December 16th 2023

ss4<-terra::rast("~/Documents/cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729252813869.nc")

#Subset to extract sea surface density only 
ss4<-subset(ss3, 1:1080)


#SS5: December 17th 2023 to October 12th 2024

ss5<-terra::rast("~/Documents/cmems_obs-mob_glo_phy-sss_nrt_multi_P1D_1729253007917.nc")

#Subset to extract sea surface density only 
ss5<-subset(ss3, 1:301)

Concatenate all the subsetted netcdf files from Jan 2012 to Oct 12th 2024 together

total_sss<-c(sss1, ss2, ss3, ss4, ss5)

Extract the Point Data

#Select the sea surface density variable from the netcdf files 
SSSs <- terra::rast(total_sss, "dos")

#Make the dataframe a spatial object of class = "sf" with a CRS of 4326
Ds_Points <- st_as_sf(x=MyDf, 
                      coords = c("Longitude_E_DD", "Latitude_N_DD"), 
                      crs = 4326)
    
#Extract the SSS values from the netcdf files 
SSSs_Data <- terra::extract(SSSs, Ds_Points)

#Add the SSS results to the dataframe with the function cbind 
MyDf <- cbind(MyDf, SSS = SSSs_Data$dos)

This is the information stored in 'total_sss'

class       : SpatRaster 
dimensions  : 189, 154, 4669  (nrow, ncol, nlyr)
resolution  : 0.125, 0.125  (x, y)
extent      : 28.75, 48, 8.625, 32.25  (xmin, xmax, ymin, ymax)
coord. ref. : lon/lat WGS 84 
sources     : cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729236771868.nc:dos  (1096 layers) 
              cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729236939325.nc:dos  (1096 layers) 
              cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729237129644.nc:dos  (1096 layers) 
              ... and 2 more source(s)
varnames    : dos (Sea surface density) 
              dos (Sea surface density) 
              dos (Sea surface density) 
              ...
names       : dos_d~=-0_1, dos_d~=-0_2, dos_d~=-0_3, dos_d~=-0_4, dos_d~=-0_5, dos_d~=-0_6, ... 
unit        :       kg/m3,       kg/m3,       kg/m3,       kg/m3,       kg/m3,       kg/m3, ... 
time        : 2012-01-01 to 2020-12-31 UTC

This is the information stored in SS1

class       : SpatRaster 
dimensions  : 189, 154, 1096  (nrow, ncol, nlyr)
resolution  : 0.125, 0.125  (x, y)
extent      : 28.75, 48, 8.625, 32.25  (xmin, xmax, ymin, ymax)
coord. ref. : lon/lat WGS 84 
source      : cmems_obs-mob_glo_phy-sss_my_multi_P1D_1729236771868.nc:dos 
varname     : dos (Sea surface density) 
names       : dos_d~=-0_1, dos_d~=-0_2, dos_d~=-0_3, dos_d~=-0_4, dos_d~=-0_5, dos_d~=-0_6, ... 
unit        :       kg/m3,       kg/m3,       kg/m3,       kg/m3,       kg/m3,       kg/m3, ... 
time        : 2012-01-01 to 2014-12-31 UTC

It seems in order to reproduce the error, we would have to register with the website. Is it possible that the data type in one or more of your files is being read in as character instead of numeric? — Allan Cameron
– Allan Cameron, Commented Oct 27, 2024 at 17:00
@AllanCameron Is there a way of storing files somewhere where a Stackoverflow user can click on a link to find the files? If I could do this then I'd link my files in this question. If you know how to do this, that would be amazing — Alice Hobbs
– Alice Hobbs, Commented Oct 27, 2024 at 17:04
This is what ChatGPT said: The error message you're encountering suggests that you're trying to convert or assign a character type to a variable that expects a double (numeric) type. This often happens in programming languages like R or Python when there is a type mismatch. — Alice Hobbs
– Alice Hobbs, Commented Oct 27, 2024 at 17:05
Is there a way to convert the values to numeric? I know with ordinary data but these are SpatRasters — Alice Hobbs
– Alice Hobbs, Commented Oct 27, 2024 at 17:06

Grzegorz Sapijaszko · Accepted Answer · 2024-10-27 17:46:31Z

1

Replace:

SSSs <- terra::rast(total_sss, "dos")

with

SSSs <- terra::subset(total_sss, 1:xx)

where xx is number of layers for dos varname or with

SSSs <- total_sss[[1:xx]]

or

SSSs <- total_sss["dos_d"]

where dos_d is partial mach of names. Please have a look on terra:subset() function. And Robert's answer: How to properly subset data using their variable names in terra?

BTW, to temporary share files you can use google drive/github.

edited Oct 27, 2024 at 17:46

answered Oct 27, 2024 at 17:10

Grzegorz Sapijaszko

3,7242 gold badges9 silver badges21 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Alice Hobbs Over a year ago

Thank you so much for your answer, I really appreciated that. I am gong to look into using GIthub to store temporary files for Stackoverflow questions so they are reproducible.

Grzegorz Sapijaszko Over a year ago

Glad it helped :). BTW, when the conference poster will be ready, I would like to see it somehow :)

Alice Hobbs Over a year ago

I'm still doing my analysis but I'll need to start designing it this week. I don't know how we can connect as I don't want to put my email address on a public forum. I'm very happy to show you my poster. Any suggestions with how we can connect

Collectives™ on Stack Overflow

How to Extract Point Data from Multiple Subsetted Netcdf Files for Sea Surface Density using terra::rast() and terra::extract functions in R

1 Answer 1

3 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related