I am using NASA's Black Marble monthly nighttime light (NTL) product, VNP46A3. The data are in .h5 format and apart from the radiance product, they come with quality assurance (QA) data. I am using the All_Angle_Composite_Snow_Free radiance NTL image for my analysis. Using the below code I extracted the radiance image as well as the QA rasters:
library(terra)
wd <- "path/"
s <- sds(paste0(wd, "VNP46A3.A2018182.h06v05.001.2021125183820.h5"))
# extract single image as spatraster from h5
r <- s[5] # s[] are the radiance and QA rasters
writeRaster(r, paste0(wd, "NTL.tif"))
The next step is the preprocessing of the NTL images, i.e., to eliminate bad quality pixels. I am using the ifel function from the terra package in R to achieve this. There are 2 QA rasters:
- Quality is a binary raster and the good quality pixels have the value of 0
- Num is the number of observations being used to create the monthly raster and should have a value greater than 0
I want to create ifel statement(s) such as:
if the Quality = 0 and Num > 0 then keep the NTL raster values else set the NTL raster values to NULL.
I have tried to create 2 separate ifel statements and then combine the results but I am getting an error:
NTL <- rast(paste0(wd, "NTL.tif"))
Num <- rast(paste0(wd, "Num.tif"))
Quality <- rast(paste0(wd, "Quality.tif"))
r <- ifel(Quality = 0, NTL, NULL)
# Error in .local(test, ...) : argument "no" is missing, with no default
r1 <- ifel(Num > 0, NTL, NULL)
# Error in .local(test, ...) : argument "no" is missing, with no default