Analysing remotely sensed vegetation indices and their response to drought.
This repository contains R code. For a demo knit the RMarkdown file vignette_rsvi.Rmd.
Site selection was done based on:
- Site where fLUE method worked and where clear soil moisture droughts were identified. I.e., site belongs to clusters cDD, cGR, or cLS as described in Stocker et al. (2018) New Phytologist.
- Site was among the subset of homogenous sites, selected by Manuela Balzarolo. All sites with a homogenous surrounding are listed in file
data/FLUXNET-2015_Tier1/meta/fluxnet_quality_check_homogenous_OK_PRI_mbalzarolo.csv.
Reproduce it as follows:
Get sites where fLUE data is available
read_csv("~/data/flue/flue_stocker18nphyt.csv")
df_flue_sites <- select(df_flue, site, cluster) %>% unique()Add meta information for sites
library(rsofun)
df_flue_sites <- df_flue_sites %>% rename(sitename=site) %>% left_join( rsofun::metainfo_Tier1_sites_kgclimate_fluxnet2015, by="sitename" )Subset sites belonging to cluster cGR, cDD, or cLS
df_sub <- df_flue_sites %>% filter(cluster %in% c("cGR", "cDD", "cLS"))
write_csv(df_sub, path = "~/data/flue/metainfo_sites_clusters_cGR_cDD_cLS_flue_fluxnet2015.csv")Get list of sites with homogenous surrounding and subset fLUE sites base on that list
df_homo <- read_csv("~/data/FLUXNET-2015_Tier1/meta/fluxnet_quality_check_homogenous_OK_PRI_mbalzarolo.csv")
df_sub_homo <- filter( df_sub, sitename %in% df_homo$sitename )
write_csv(df_sub_homo, path = "./data/sites.csv")