1_fetch_spatial.R

source("1_fetch/src/Download_nhd.R")
source('1_fetch/src/download_states_shp.R')
source('1_fetch/src/fetch_nhdplus_data.R')

p1_sp_targets_list <- list(
  
  # Lake locations (lat/lon) Fetch #
  
  ## Reading and cleaning list of saline lakes
  tar_target(
    p1_lakes_sf,
    {read_csv('1_fetch/in/saline_lakes.csv', col_types = 'ccnn') %>% 
        st_as_sf(coords = c('Lon','Lat'), crs = p0_selected_crs) %>% 
        rename(point_geometry = geometry, lake = `Lake Ecosystem`, state = State) %>% 
        mutate(state_abbr = case_when(state == 'California' ~ 'CA',
                                      state == 'Nevada' ~ 'NV',
                                      state == 'Utah' ~ 'UT',
                                      state == 'Oregon' ~ 'OR',
                                      state == 'California/Oregon' ~ 'CA', # Put lake crossing OR CA border as CA for now.  
                                      TRUE ~ 'NA'),
               lake = str_to_title(lake),
               lake_w_state = paste(lake, state_abbr, sep = ','),
               lake_name_shrt = trimws(str_replace(lake, pattern = 'Lake', replacement = "")))
    }
  ),
  
  ## Lake sf dataset from sharepoint - project team provided
  ## This target is the same as p2_saline_lakes_sf but includes the manually drawn polygon Carson Sink.
  ## This target has been created to: 
  ## 1/ have a second multipolygon dataset for lakes
  ## 2/ append Carson Sink to p2_saline_lakes_sf
  ## This should be manually downloaded to local 1_fetch/in/sharepoint folder 
  tar_target(
    p1_saline_lakes_bnds_sf,
    st_read('1_fetch/in/SalineLakeBnds/SalineLakeBnds.shp') %>% 
      st_transform(crs=st_crs(p1_lakes_sf)) %>% 
      ## Formatting for easier rbind with p2_saline_lakes_sf
      rename(GNIS_Name = Name) %>% 
      mutate(lake_w_state = paste0(GNIS_Name,',',State)) %>% 
      select(lake_w_state, GNIS_Name, geometry)
  ),
  
  # States Shp Fetch - used for Lakes Querying #
  
  ## Download states shp
  tar_target(
    p1_download_states_shp,
    download_states_shp(url = pO_states_dwnld_url, 
                        out_path = '1_fetch/out/states_shp'),
    format = 'file'
  ),
  
  tar_target(
    p1_states_sf,
    st_read(file.path(p1_download_states_shp,'statesp010g.shp'), quiet = TRUE) %>%
      filter(STATE_ABBR %in% c('CA',"NV",'UT','OR')) %>% 
      st_transform(crs = st_crs(p1_lakes_sf)) %>% 
      select(NAME,STATE_ABBR, geometry)
  ),
  
  
  # nhdhr download and fetch #
  
  ## 1st fetch of huc08 of lakes to be ableo all relevant high res nhd data (water bodies, huc6, huc8, huc10 areas) for focal lakes 
  tar_target(
    p1_huc08_full_basin_sf,
    get_huc8(AOI = p1_lakes_sf$point_geometry)
  ),
  
  ## Split huc ids to 04 to pull nhdhr - download_nhdhr() requires that huc id param be huc4.
  tar_target(
    p1_huc04_for_download,
    substr(p1_huc08_full_basin_sf$huc8, start = 1, stop = 4) %>%
        unique() %>%
        ## adding 1601 which is a HU4 that contains watersheds relevant to Great Salt Lake 
        append('1601')
  ),
  
  ## Download high res nhd data to get lake water bodies 
  tar_target(
    p1_download_nhdhr_lakes_path,
    download_nhdhr_data(nhdhr_gdb_path = '1_fetch/out/nhdhr',
                        huc04_list = p1_huc04_for_download),
    format = 'file'
  ),
  
  ## Fetch waterbodies, huc6, huc8, huc10 from hr and place in local gpkg
  tar_target(p1_nhd_gpkg, 
             get_downloaded_nhd_data(gdb_path = p1_download_nhdhr_lakes_path,
                                     out_gpkg_path = '1_fetch/out/nhd_WB_HUC6_HU8_HU10.gpkg',
                                     layer = c('NHDWaterbody','WBDHU6', 'WBDHU8', 'WBDHU10')),
             format = 'file'
  ),
  
  # # OPTIONAL - if you already have the gpkg, and do not want to build,
  # # place it in the correct folder and comment out target above : 
  # # In sharepoint, gpkg is located here: https://doimspp.sharepoint.com/sites/IIDDStaff/Shared%20Documents/Forms/AllItems.aspx?id=%2Fsites%2FIIDDStaff%2FShared%20Documents%2FFunction%20%2D%20Data%20Pipelines%2FData%5FScience%5FPractitioners%5FMtgs%2Ftasks%2Fsaline%5Flakes%2FData&viewid=acf8af64%2Daaf3%2D4f23%2D8b74%2D2859a89813c5
  # tar_target(p1_nhd_gpkg, 
  #            '1_fetch/out/nhd_WB_HUC6_HU8_HU10.gpkg'),
  
  
  # Lakes Fetch #
  
  ## Read in all nhd hr waterbodies in our entire basin
  ## This is then processed to our focal lakes in 2_process.R
  tar_target(p1_nhdhr_lakes,
             sf::st_read(p1_nhd_gpkg,
                         layer = 'NHDWaterbody',
                         ## filtering to larger lakes because nhdhr has a lot of very small lake polygons
                         query = 'SELECT * FROM NHDWaterbody WHERE Shape_Area > 7e-08',
                         quiet = TRUE)
  ),
  
  # HUC area Processing #
  ## Note - using p2_saline_lakes_sf - created in 2_process.R - to scope HUCs
  
  ## Reading in huc6 entire basin 
  tar_target(
    p1_basin_huc6_sf,
    st_read(p1_nhd_gpkg, layer = 'WBDHU6', quiet = TRUE) %>% 
      st_transform(crs = st_crs(p2_saline_lakes_sf))
  ),
  
  ## Reading in all huc8 in entire basin
  tar_target(
    p1_basin_huc8_sf,
    st_read(p1_nhd_gpkg, layer = 'WBDHU8', quiet = TRUE) %>% 
      st_transform(crs = st_crs(p2_saline_lakes_sf))
  ),
  
  ## Reading in all huc10 in entire basin
  tar_target(
    p1_basin_huc10_sf,
    st_read(p1_nhd_gpkg, layer = 'WBDHU10', quiet = TRUE) %>% 
      st_transform(crs = st_crs(p2_saline_lakes_sf))
  ),
  
  ## Fetch watershed boundary areas filtered to our lakes - huc8 - HR
  ### note possible duplicate polygons since some individual saline lakes have same huc08 
  tar_target(
    p1_lakes_huc6_sf,
    p1_basin_huc6_sf %>% 
      ## filter HUC6 to only the Huc6 surrounding saline lakes
      st_join(p2_saline_lakes_sf, left = FALSE) %>%
      filter(!is.na(GNIS_Name)) %>% 
      distinct()
  ),
  
  ## Fetch watershed boundary areas filtered to our lakes - huc8 - HR
  ### note possible duplicate polygons since some individual saline lakes have same huc08 
  tar_target(
    p1_lakes_huc8_sf,
    p1_basin_huc8_sf %>% 
      ## filter HUC8 to only the HUC8 with within selected HUC6
      st_join(x = ., y = p1_lakes_huc6_sf[,c('HUC6','lake_w_state')],
              join = st_within, left = FALSE) %>%
      filter(!is.na(HUC6)) %>% 
      distinct()
  ),
  
  ## Fetch watershed boundary areas - huc10  
  ### note possible duplicate polygons since some individual saline lakes have same huc10 
  ### This target is very slow to build! 
  tar_target(
    p1_lakes_huc10_sf,
    p1_basin_huc10_sf %>%
      ## Filtering HUC10 to within selected HUC8s
      st_join(x = .,
              y = p1_lakes_huc8_sf[,c('HUC6', 'HUC8', 'lake_w_state')],
              join = st_within, left = FALSE) %>%
      filter(!is.na(HUC8)) %>% 
      distinct()
  ),
  
  # Flowlines Fetch #
  
  ## Grab vector of our huc08s in order to run branching for nhd flowlines fetch  
  tar_target(
    p1_huc8_vec, 
    {unique(p1_lakes_huc8_sf$HUC8)}
  ),
  
             
  # Fetch nhdplus flowlines for each selected huc8 region separately through dynamic branching - note difference between branches 
  tar_target(
    p1_lake_flowlines_huc8_sf_lst,
    {get_nhdplus(AOI = {p1_lakes_huc8_sf %>% filter(HUC8 %in% p1_huc8_vec)},
                 realization = 'flowline') %>%
        ## fixing col that are automatically transforming to char
        mutate(across(c(surfarea, lakefract, rareahload), ~as.numeric(.x)),
               HUC8 = p1_huc8_vec) 
    }, 
    pattern = map(p1_huc8_vec),
    iteration = 'list'
  ),
  
  tar_target(
    p1_lake_flowlines_huc8_sf,
    bind_rows(p1_lake_flowlines_huc8_sf_lst)
  ),
  
  # # NWIS site fetch from nhdplus 
  # # Fetch NWIS sites along tributaries and in our huc08 regions. 
  # # for comparison purposes
  # ## Will require further filtering (e.g. ftype == ST, along flowlines only)
  tar_target(
    p1_nwis_sites_from_nhdplus_lst,
    {tryCatch(expr = get_huc8(id = p1_huc8_vec) %>% get_nwis(AOI = .) %>%
                ## making as dataframe to load with tar_load()
                as.data.frame() %>%
                mutate(HUC8 = p1_huc8_vec),
              error = function(e){
                return(warning(e$message))
              },
              warning = function(w){
                return(message(paste(w$message, 'huc8:', p1_huc8_vec)))
              }
    )},
    pattern = map(p1_huc8_vec),
    iteration = 'list'
  ),
  
  tar_target(
    p1_nwis_sites_from_nhdplus,
    bind_rows(p1_nwis_sites_from_nhdplus_lst)
  ),
  # 
  # ## Pulling site no from gauge sites to then query nwis and WQP with data retrieval
  tar_target(
    p1_site_ids_from_nhdplus,
    {p1_nwis_sites_from_nhdplus %>% pull(site_no) %>% unique()}
  )

)