MATLAB 'structure' files containing archived exposure-age data 
- Extracted from ICE-D on 04/24/2024 
- Preprocessed according to the procedure outlined in the accompanying manuscript,
  including screening out ages prior to the Last Glacial Maximum, removing sites
  with no data or in the Antarctic Peninsula or sub-Antarctic islands, removes 
  unpublished C14 data 
- Preprocessing also lumps sites based on an input model grid resolution (2km nested   
  domain grids, or 40km continent-wide grid resolution). Lumped sites have all original
  short_names and long_names (as well as age measurements) appended together in the 
  lumped site structure fields.

At each site, some fields are preserved from the ICE-D dataset. For example, site information: 
  short_name, name, mean_lat, mean_lon; and sample information: lat_DD and lon_DD
  (latitude and longitude in decimal degrees), elv_m (sample elevation, meters), elv_above_ice
  (sample elevation above the modern ice surface, meters), what (text description of the sample 
  type), nuclide (cosmogenic nuclide measured), t (resulting exposure age measurement, years), 
  dtint (measurement error, years).
Our preprocessing steps additionally populate new fields: the Monte-Carlo iteration 
  analysis to identify the youngest age-elevation-bounding samples filtered the original 
  ICE-D fields to isolate the youngest age-elevation-bounding samples: elv_young, elvice_young, 
  t_young, lat_young, lon_young, nuclide_young, dt_young
The uncertainty used for model-data comparison is based on the age range across the Monte Carlo
  ‘cloud’ at any given elevation (the 1σ range), or the minimum geologic uncertainty, whichever 
  is larger, as described in the main text. For each site where we applied the age-elevation 
  bounding analysis (i.e., sites with > 3 samples spanning >100 m of elevation change), we add
  fields dt_mcn_0m, dt_mcn_5m, dt_mcn_20m (with a 0m, 5m, or 20m vertical elevation window, 
  respectively). Note that dt_young (age uncertainty) is simply the measurement error of the 
  youngest age-elevation bounding samples; we used dt_mcn_20m (i.e., a 20m vertical elevation 
  window) for the final model scores presented in the manuscript.  
Our preprocessing also adds a 'region' number field into the data structure:
  1 - Weddell Sea Embayment
  2 - Amundsen Sea Embayment
  3 - Ross Sea Embayment (later subdivided into 7,8)
  4 - Wilkes Land
  5 - Aurora margin
  6 - East Antarctic margin
  7 - Transantarctic Mountains region
  8 - Marie Byrd Land / Eastern Ross Sea region
We add a 'domain' field to contain domain number, corresponding to the nested model domain covering 
  each site, if applicable (note that we only proceed with model-data comparison for sites with > 3 
  samples spanning >100 m of elevation change, so if the site does not meet those criteria, it may 
  not have a model domain number associated with it)
Site weight field 'Sw' (0-1) is computed based on the number of regional sites within a search
  distance, as described in main text.  



Usage: for example, to plot the youngest age-elevation bounding samples at HUGBLUFF, 
  assuming you want to use the 2km nested model domain for model-data comparison:

  this_site = all_sites_data_2km.HUGBLUFF;
  errorbar(this_site.t_young,this_site.elvice_young,this_site.dt_young,'o','horizontal');
  this_site.region % 7 (Transantarctic Mountain region)

