Calculating a goal model with only national-level data

[jules32]

Here is a question that often comes up: what if there is a goal where you only have data at the at the national level (example: region_id 0), not for each region (example: region_id 1 to 8)? While there are ways to try to allocate the data to smaller spatial scales, they have a lot of assumptions and are dependent on a lot of things that aren't very transparent.

Instead, it's possible to calculate the status for region_id 0 and then allocate the status score to the other regions (you can also do this with trend if you need to). This will give the same status score to each region, but pressures and resilience layers can still act differently on each region.

Here is some code to illustrate (and also now in the toolbox-demo)

LSP_ex <- function(layers){
  ### an example when there are only data available for the entire assessment area, not for each region.

  ### our assessment area has 8 regions...

  ## total offshore/inland areas
  inland <- layers$data$rgn_area_inland1km %>%
    select(region_id = rgn_id, area_inland1km = area); inland
  offshore <- layers$data$rgn_area_offshore3nm %>%
    select(region_id = rgn_id, area_offshore3nm = area); offshore

  ## total protected
  inland_prot <- layers$data$lsp_prot_area_inland1km %>%
    select(region_id = rgn_id, cmpa = a_prot_1km, year); inland_prot
  offshore_prot <- layers$data$lsp_prot_area_offshore3nm %>%
    select(region_id = rgn_id, cp = a_prot_3nm, year); offshore_prot


  ### ...but let's pretend we only had data for the entire assessment area and not those regions:

  inland   <- data_frame(region_id = 0, area_in = 175147)
  offshore <- data_frame(region_id = 0, area_off = 621887.8)
  inland_prot   <- data_frame(region_id = 0, cmpa = 15369.5)
  offshore_prot <- data_frame(region_id = 0, cp = 30635.50)

  ## combine to one dataframe
  lsp_data <- inland %>%
    left_join(inland_prot) %>%
    left_join(offshore) %>%
    left_join(offshore_prot)

  ### you can calculate the status for just this overall assessment area, and then apply the status score to all the regions so that the Toolbox can calculate the pressures/resilience for those regions.

  ## model
  status_0 <- lsp_data %>%
    dplyr::mutate(score = area_in/cp + area_off/cmpa)

  ## apply to all regions
  status <- data_frame(region_id = 1:8,
                       score  = status_0$score) %>%
    mutate(goal = "LSP",
           dimension = "status") %>%
    arrange(goal, dimension, region_id, score)

}