geom_tile failed when mappng soil_moisture in R

Question

I have two soil moisture hdf images. The first one is 3Km resolution and the second one is 36km. Using same code, the second one can map in the following code as below: soil moisture in 36Km resolution

The first one doesn't show the soil moisture information: soil moisture in 3km resolution

The code I'm using is as below

install.packages("hdf5")
BiocManager::install("hdf5")

devtools::install_github("hhoeflin/hdf5r")

library(devtools)
library(BiocManager)
library(rhdf5)
library(tidyr)
library(ggplot2)

mydata <- h5read("/Users/ss/Downloads/SMAP_L2_SM_A_01725_D_20150529T123629_R13080_001.h5",
                 "Soil_Moisture_Retrieval_Data")
str(mydata)

latitude<-mydata$latitude
longitude<-mydata$longitude
soil_moisture<-mydata$soil_moisture
soil_moisture[soil_moisture==-9999]<-NA
soil_moisture[soil_moisture>0.5]<-NA
soil_moisture[soil_moisture<0.02]<-NA

data<-cbind(latitude,longitude,soil_moisture)
data1<-data[complete.cases(data),]

soil_moisture = soil_moisture*4

data1<-as.data.frame(data1)
xlmax<-max(data1$longitude)
xlmin<-min(data1$longitude)
ylmax<-max(data1$latitude)
ylmin<-min(data1$latitude)


dataplot<-ggplot(data1)+
          geom_tile(aes(x=longitude,y=latitude,fill=soil_moisture))+
          xlab("Longitude (deg)") + # x-axis label
          ylab("Latitude (deg)") +
          geom_path(data = map_data("world"),
                    aes(x = long, y = lat, group = group))+
          scale_fill_distiller(palette = "YlOrRd",limits=c(0.02,0.5),name="SM") +
          coord_fixed(xlim =c(floor(xlmin),ceiling(xlmax)),ylim=c(floor(ylmin),ceiling(ylmax)))
print(dataplot)
# ggsave("myplot.png",width=8,height=8,unit="cm",dpi=300)

Are there definitely data at the 3 km res? Sometimes with complicated datasets the problem is the data, not the code. Are your data and the worldmap using the same CRS? — mlcyo
I used crs(mydata) crs(map_data("world")) to see their coordinates, and it returns na. Is this a problem? — guo_lll

guo_lll guo_lll · Accepted Answer · 2020-10-14T09:49:40

to anyone who might encounter the same problem, you can use

dataplot<-ggplot(data1)+
          geom_point(aes(x=longitude,y=latitude,fill=soil_moisture))+
          xlab("Longitude (deg)") + # x-axis label
          ylab("Latitude (deg)") +
          geom_path(data = map_data("world"),
                    aes(x = long, y = lat, group = group))+
          scale_fill_distiller(palette = "YlOrRd",limits=c(0.02,0.5),name="SM") +
          coord_fixed(xlim =c(floor(xlmin),ceiling(xlmax)),ylim=c(floor(ylmin),ceiling(ylmax)))
print(dataplot)

to plot the soil moisture map of 3km resolution.

replace geom_tile with geom_point.

geom_tile failed when mappng soil_moisture in R

1 Answers