In stineb/ingestr: Environmental point data ingest

library(ingestr)
library(ggplot2)
# library(readr)
# library(dplyr)
# library(lubridate)
# library(rsofun)

Single site

Prepare access to Google Earth Engine

The library gee_subset by Koen Hufkens can be downloaded from this link and used to extract data directly from Google Earth Engine. Note that this requires the following programmes to be available:

• git: You can use Homebrew to installing git by entering in your terminal: brew install git.
• python

Then, carry out the follwing steps:

• In your terminal, change to where you want to have the repository. In this example, we're cloning it into our home directory: ```{sh, eval = FALSE} cd ~ git clone https://github.com/khufkens/google_earth_engine_subsets.git

To get access to using the Google Earth Engine API (required to use the `gee_subset` library), carry out the following steps in your terminal. This follows steps described [here](https://github.com/google/earthengine-api/issues/27).

1. Install google API Python client
```{sh, eval = FALSE}
sudo pip install --upgrade google-api-python-client

I had an error and first had to do this here following this link: ```{sh, eval = FALSE} sudo pip install --ignore-installed six

2. Install pyCrypto
```{sh, eval = FALSE}
sudo pip install pyCrypto --upgrade

1. Install Python GEE API ```{sh, eval = FALSE} sudo pip install earthengine-api

4. Run authentification for GEE
```{sh, eval = FALSE}
earthengine authenticate

1. Finally, try if it works. This shouldn't return an error: ```{sh, eval = FALSE} python -c "import ee; ee.Initialize()"

### Download data

To facilitate the selection of data products and bands to be downloaded, you may use the function `get_settings_gee()` which defines defaults for different data bundles (`c("modis_fpar", "modis_evi", "modis_lai", "modis_gpp")` are available). We use `"modis_evi"`, downloading the MODIS/006/MOD13Q1, band EVI data.

The following example is for downloading MODIS EVI data.
```r
settings_gee <- get_settings_gee(
  bundle            = "modis_evi",
  python_path       = system("which python", intern = TRUE),
  gee_path          = "~/google_earth_engine_subsets/gee_subset/",
  data_path         = "~/data/gee_subsets/",
  method_interpol   = "linear",
  keep              = TRUE,
  overwrite_raw     = FALSE,
  overwrite_interpol= TRUE
  )

This can now be used to download the data to the directory specified by argument data_path of function get_settings_gee().

df_gee_modis_fpar <- ingest_bysite(
  sitename  = "CH-Lae",
  source    = "gee",
  year_start= 2010,
  year_end  = 2012,
  lon       = 8.365,
  lat       = 47.4781,
  settings  = settings_gee,
  verbose   = FALSE
  )

Plot this data.

plot_fapar_ingestr_bysite(df_gee_modis_fpar, settings_gee)

Multiple sites

Using the same settings as specified above, we can download MODIS FPAR data for multiple sites at once from GEE:

settings_gee <- get_settings_gee(
  bundle            = "modis_evi",
  python_path       = system("which python", intern = TRUE),
  gee_path          = "~/google_earth_engine_subsets/gee_subset/",
  data_path         = "~/data/gee_subsets/",
  method_interpol   = "linear",
  keep              = TRUE,
  overwrite_raw     = FALSE,
  overwrite_interpol= TRUE
  )

df_gee_modis_evi <- ingest(
  siteinfo= ingestr::siteinfo %>% 
    dplyr::filter(!(sitename %in% c("AU-GWW", "AU-Lox", "AU-Rob", "AU-TTE", "CN-Dan"))),
  source  = "gee",
  settings= settings_gee,
  verbose = FALSE
  ) 

df <- df_gee_modis_evi %>% 
  unnest(data) %>% 
  dplyr::select(sitename, date, evi = modisvar_interpol)

## quick check
df %>% 
  dplyr::filter(sitename == "AR-SLu") %>% 
  ggplot(aes(x = date, y = evi)) +
  geom_line()

write_csv(
  df, 
  path = "~/data/fluxnet_subsets/EVI_MOD13Q1_gee_subset.csv"
  )

Collect all plots.

list_gg <- plot_fapar_ingestr(df_gee_modis_fpar, settings_gee)
#purrr::map(list_gg, ~print(.))

stineb/ingestr documentation built on July 21, 2024, 6:39 p.m.