add green data

.gitignore

@@ -10,3 +10,4 @@ inst/doc
 /doc/
 /Meta/
 /.env
+/data-raw/

inst/codec_catalog/make_all_codec_dpkg.R

@@ -14,7 +14,7 @@ dpkgs <-
     get_codec_dpkg("traffic-v0.1.2"),
     get_codec_dpkg("drivetime-v0.2.2"),
     get_codec_dpkg("environmental_justice_index-v0.1.0"),
-    get_codec_dpkg("landcover-v0.1.0"),
+    get_codec_dpkg("green-v0.1.0"),
     get_codec_dpkg("parcel-v0.1.1"),
     get_codec_dpkg("property_code_enforcements-v0.2.0"),
     get_codec_dpkg("xx_address-v0.2.0")

inst/codec_data/green/README.md

@@ -0,0 +1,29 @@
+# Green
+
+[![latest github release for green dpkg](https://img.shields.io/github/v/release/geomarker-io/codec?sort=date&filter=green-*&display_name=tag&label=%5B%E2%98%B0%5D&labelColor=%238CB4C3&color=%23396175)](https://github.com/geomarker-io/codec/releases?q=green&expanded=false)
+
+## About
+
+Annual measures of greenness and built environment. The codec dpkg includes the most recently available vintage for each data product (2023 greenspace and impervious, 2021 tree canopy, 2024 EVI, and 2020 city treecanopy)
+
+## Data
+
+### greenspace 
+
+The National Landcover Database (NLCD) classifies land at a 30 m x 30 m resolution into one of [16 categories](https://www.mrlc.gov/data/type/land-cover). Here, a grid cell is considered greenspace if its NLCD land cover classification is in any category other than water, ice/snow, developed medium intensity, developed high intensity, or rock/sand/clay. Grid cells are aggregated to census tracts by calculating the percentage of overlapping "green" grid cells in each tract.
+
+### tree canopy and impervious surface
+
+Percentage [tree canopy](https://www.mrlc.gov/data/type/tree-canopy) and [impervious surface](https://www.mrlc.gov/data/type/fractional-impervious-surface) coverage at 30 m x 30 m resolution were obtained from NLCD and aggregated to census tract by calculating the mean tree canopy percentage for grid cells overlapping each tract.
+
+### EVI
+
+The Enhanced Vegetation Index (EVI) is a measure of greenness that ranges from -0.2 to 1, with higher values corresponding to more vegetation. A cloud-free composite EVI raster at a resolution of 250 × 250 m was created by assembling individual images collected via remote sensing between June 9 and June 24, 2024. Tract averages were calculated using the mean EVI of overlapping grid cells. The EVI raster file (`MOD13Q1.A2024161.h11v05.061.2024181211403.hdf`) was downloaded from https://search.earthdata.nasa.gov/search on 2025-03-12, scaled by 0.0001, and rounded to 3 decimal points 
+
+### park service area coverage and park greenspace coverage
+
+The percentage of area covered by at least one park service area (defined as a 10-minute walk or 0.25 mile buffer around each park) and the percentage of area covered by park greenspace was calculated for each tract. Data was obtained from correspondence with Cincinnati Parks in August 2024.
+
+### city tree canopy 
+
+2020 LIDAR measured tree canopy area by 2010 census block groups summed by 2010 census tracts then interpolated to 2020 census tract using area weighting. City tree canopy is only available for census tracts within the City of Cincinnati.

inst/codec_data/green/dataverse.R

@@ -0,0 +1,45 @@
+#' get_dv_url
+#' get the download URL for a file hosted on a dataverse instance
+#' @param persistent_id the dataset's unique, persistent identifier
+#' @param filename the name of the file in the dataset
+#' @param version the dataset's version
+#' @param server_url the dataverse instance's URL
+#' @export
+get_dv_url <- function(persistent_id, filename = NULL, version = "latest", server_url = "https://dataverse.harvard.edu") {
+
+  stopifnot(is.character(persistent_id))
+  stopifnot(is.character(version))
+  stopifnot(is.character(server_url))
+  if (! substr(persistent_id, 1, 4) == "doi:") stop("`persistent_id` must begin with 'doi:'", call. = FALSE)
+  if (version == "latest") version <- ":latest"
+
+  req <-
+    httr2::request(server_url) |>
+    httr2::req_user_agent("pcog (https://github.com/geomarker-io/pcog)") |>
+    httr2::req_url_path_append("api", "datasets", ":persistentId", "versions", version) |>
+    httr2::req_url_query("persistentId" = persistent_id) |>
+    httr2::req_error(
+      is_error = function(resp) httr2::resp_status(resp) != 200,
+      body = function(resp) glue::glue("version {version} of {persistent_id} not found at {server_url}")
+    )
+  resp <- httr2::req_perform(req)
+
+  the_files <-
+    httr2::resp_body_json(resp)$data$files |>
+    vapply(\(.) .$dataFile[["id"]], integer(1))
+
+  names(the_files) <-
+    httr2::resp_body_json(resp)$data$files |>
+    vapply(\(.) .$dataFile[["filename"]], character(1))
+
+  if (length(filename) == 1 && filename %in% names(the_files)) {
+    file_id <- the_files[[filename]]
+  } else {
+    message("available files for ", persistent_id, " include: \n  ", paste(names(the_files), collapse = "\n  "))
+    if (length(filename) == 0) stop("no filename requested", call. = FALSE)
+    stop("filename ", filename, " not found.", call. = FALSE)
+  }
+
+  cog_url <- glue::glue("https://dataverse.harvard.edu/api/access/datafile/{file_id}")
+  return(as.character(cog_url))
+}

inst/codec_data/green/green.R

@@ -0,0 +1,177 @@
+if (tryCatch(read.dcf("DESCRIPTION")[1, "Package"] == "codec", finally = FALSE)) {
+  devtools::load_all()
+} else {
+  library(codec)
+}
+message("Using CoDEC, version ", packageVersion("codec"))
+source("inst/codec_data/green/dataverse.R")
+
+nlcd_legend <-
+  tibble::tribble(
+    ~value, ~landcover_class, ~landcover, ~green,
+    11, "water", "water", FALSE,
+    12, "water", "ice/snow", FALSE,
+    21, "developed", "developed open", TRUE,
+    22, "developed", "developed low intensity", TRUE,
+    23, "developed", "developed medium intensity", FALSE,
+    24, "developed", "developed high intensity", FALSE,
+    31, "barren", "rock/sand/clay", FALSE,
+    41, "forest", "deciduous forest", TRUE,
+    42, "forest", "evergreen forest", TRUE,
+    43, "forest", "mixed forest", TRUE,
+    51, "shrubland", "dwarf scrub", TRUE,
+    52, "shrubland", "shrub/scrub", TRUE,
+    71, "herbaceous", "grassland", TRUE,
+    72, "herbaceous", "sedge", TRUE,
+    73, "herbaceous", "lichens", TRUE,
+    74, "herbaceous", "moss", TRUE,
+    81, "cultivated", "pasture/hay", TRUE,
+    82, "cultivated", "cultivated crops", TRUE,
+    90, "wetlands", "woody wetlands", TRUE,
+    95, "wetlands", "emergent herbaceous wetlands", TRUE
+  )
+
+green_rast <-
+  dpkg::stow("https://s3-us-west-2.amazonaws.com/mrlc/Annual_NLCD_LndCov_2023_CU_C1V0.tif") |>
+  terra::rast()
+
+hc <- 
+  codec::cincy_county_geo() |> 
+  sf::st_as_sfc() |> 
+  terra::vect() |> 
+  terra::project(green_rast)
+
+tract <- 
+  codec::cincy_census_geo("tract", "2020") |> 
+  terra::vect() |> 
+  terra::project(green_rast)
+
+out <- tibble::tibble(census_tract_id_2020 = tract$geoid)
+
+# land cover - greenspace
+out$greenspace_2023 <-
+  green_rast |>
+  terra::crop(hc) |>
+  terra::extract(tract) |>
+  dplyr::rename(value = 2) |>
+  dplyr::left_join(nlcd_legend, by = "value") |>
+  dplyr::group_by(ID) |>
+  dplyr::summarize(greenspace = round(sum(green) / dplyr::n() * 100)) |>
+  dplyr::pull(greenspace)
+
+# tree canopy
+out$treecanopy_2021 <- 
+  dpkg::stow("https://s3-us-west-2.amazonaws.com/mrlc/nlcd_tcc_CONUS_2021_v2021-4.zip") |>
+  unzip(files = "nlcd_tcc_conus_2021_v2021-4.tif", exdir = "/Users/RASV5G/Library/Application Support/org.R-project.R/R/stow/") |>
+  terra::rast() |>
+  terra::crop(hc) |>
+  terra::extract(tract, fun = mean, ID = FALSE) |>
+  dplyr::pull(1)
+
+# impervious
+out$impervious_2023 <- 
+  get_dv_url(
+      persistent_id = "doi:10.7910/DVN/KXETFC",
+      filename = glue::glue("Annual_NLCD_FctImp_2023_CU_C1V0_COG.tif"),
+      version = "latest"
+    ) |>
+  terra::rast(vsi = TRUE) |>
+  terra::extract(tract, fun = mean, ID = FALSE) |>
+  dplyr::pull(1)
+
+# EVI 
+out$evi_2024 <- 
+  terra::sds("data-raw/MOD13Q1.A2024161.h11v05.061.2024181211403.hdf")[2] |> 
+  terra::project(terra::crs(hc)) |>
+  terra::crop(hc) |>
+  terra::extract(tract, fun = mean) |>
+  dplyr::mutate(evi = round(`250m 16 days EVI`*0.0001, 3)) |>
+  dplyr::pull(evi)
+
+# parks
+sf::st_layers(dsn = "data-raw/TPL.gdb/")
+tract_sf <- codec::cincy_census_geo("tract", "2020") |> sf::st_transform(5072)
+
+parks_sa <- 
+  sf::st_read("data-raw/TPL.gdb/", layer = "ParkServiceAreas", quiet = TRUE) |>
+  sf::st_transform(sf::st_crs(tract_sf)) 
+
+parks_green <- 
+    sf::st_read("data-raw/TPL.gdb/", layer = "Greenspace", quiet = TRUE) |>
+    sf::st_transform(sf::st_crs(tract_sf)) 
+
+get_area_pct <- function(tract, parks) {
+  area_denom <- sf::st_area(tract) |> as.numeric()
+
+  intersection <- sf::st_intersection(tract, parks)
+
+  if(nrow(intersection) < 1) {
+    return(0)
+  } else {
+    area_numer <- 
+      intersection|>
+      sf::st_union() |>
+      sf::st_area() |>
+      as.numeric()
+
+    return(round(area_numer / area_denom * 100))
+  }
+}
+
+out$park_service_area <- 
+  purrr::map_dbl(
+    1:nrow(tract_sf),
+    \(x) get_area_pct(tract_sf[x,], parks_sa)
+  )
+
+out$park_greenspace <- 
+    purrr::map_dbl(
+      1:nrow(tract_sf),
+      \(x) get_area_pct(tract_sf[x,], parks_green)
+    )
+
+# city tree canopy
+city_canopy <- 
+  sf::st_read("data-raw//Cincinnati Canopy Inforamtion 2/CincinnatiCanopyInformation.gdb", layer = "Block_Group_Canopy_Change", quiet = TRUE) |> 
+  sf::st_drop_geometry() |>
+  dplyr::select(
+    block_group_id_2010 = ID, 
+    TreeCanopy_2020_Area
+  ) |>
+  dplyr::mutate(census_tract_id_2010 = stringr::str_sub(block_group_id_2010, 1, 11)) |>
+  dplyr::group_by(census_tract_id_2010) |>  #2010 tract/bgs 
+  dplyr::summarize(TreeCanopy_2020_Area = sum(TreeCanopy_2020_Area)) |>
+  dplyr::left_join(codec::cincy_census_geo("tract", "2013"), by = c("census_tract_id_2010" = "geoid")) |>
+  sf::st_as_sf() |>
+  sf::st_transform(3735) |>
+  dplyr::mutate(
+    tract_area = sf::st_area(s2_geography), 
+    city_treecanopy_2020 = as.numeric(round(TreeCanopy_2020_Area/tract_area*100))
+) |>
+  sf::st_drop_geometry() |>
+  dplyr::select(census_tract_id_2010, city_treecanopy_2020)
+
+out$city_treecanopy_2020 <- 
+  dplyr::left_join(codec::cincy_census_geo("tract", "2013"), city_canopy, by = c("geoid" = "census_tract_id_2010")) |>
+  dplyr::rename(
+    census_tract_id_2010 = geoid, 
+    geometry = s2_geography
+  ) |>
+  sf::st_transform(5072) |>
+  cincy::interpolate(to = cincy::tract_tigris_2020, weights = "area") |>
+  sf::st_drop_geometry() |>
+  dplyr::pull(city_treecanopy_2020)
+
+# dpkg
+out_dpkg <-
+  out |>
+  dplyr::mutate(year = 2023) |>
+  as_codec_dpkg(
+    name = "green",
+    version = "0.1.0",
+    title = "Greenspace and Built Environment",
+    homepage = "https://github.com/geomarker-io/codec",
+    description = paste(readLines(fs::path_package("codec", "codec_data", "green", "README.md")), collapse = "\n")
+  )
+
+dpkg::dpkg_gh_release(out_dpkg, draft = TRUE)

vignettes/articles/data.Rmd

@@ -18,4 +18,5 @@ Click the latest release badge for more information about the data package, incl
 | American Community Survey Measures |[![latest github release for acs_measures dpkg](https://img.shields.io/github/v/release/geomarker-io/codec?sort=date&filter=acs_measures-*&display_name=tag&label=%5B%E2%98%B0%5D&labelColor=%238CB4C3&color=%23396175)](https://github.com/geomarker-io/codec/releases?q=acs_measures&expanded=false)|
 | Average Annual Daily Truck and Total Traffic Counts |[![latest github release for traffic dpkg](https://img.shields.io/github/v/release/geomarker-io/codec?sort=date&filter=traffic-*&display_name=tag&label=%5B%E2%98%B0%5D&labelColor=%238CB4C3&color=%23396175)](https://github.com/geomarker-io/codec/releases?q=traffic&expanded=false)|
 | Voter Participation Rates |[![latest github release for voter_participation dpkg](https://img.shields.io/github/v/release/geomarker-io/codec?sort=date&filter=voter_participation-*&display_name=tag&label=%5B%E2%98%B0%5D&labelColor=%238CB4C3&color=%23396175)](https://github.com/geomarker-io/codec/releases?q=voter_participation_rates&expanded=false)|
+| Green |[![latest github release for green dpkg](https://img.shields.io/github/v/release/geomarker-io/codec?sort=date&filter=green-*&display_name=tag&label=%5B%E2%98%B0%5D&labelColor=%238CB4C3&color=%23396175)](https://github.com/geomarker-io/codec/releases?q=green&expanded=false)|