The goal of icebreaker is to practice using geotargets to answer questions and use real data.
In this case, thanks to the gracrious assstance of @mdsumner, we explore how sea ice near Casey Station in Antarctica changes over time.
We used geotargets and dynamic branching to cleverly read in many rasters and not overload local memory.
As a first exploration, we produced this gif:
library(targets)
tar_source()
source("packages.R")
#>
#> Attaching package: 'arrow'
#> The following object is masked from 'package:utils':
#>
#> timestamp
#> terra 1.8.70
tar_load(ice_summary)
tar_load(plotted_ice_summaries)
date_summary <- summary(ice_summary$time)
date_summary
#> Min. 1st Qu. Median
#> "2024-08-21 00:00:00" "2024-11-20 00:00:00" "2025-02-23 00:00:00"
#> Mean 3rd Qu. Max.
#> "2025-02-21 14:04:16" "2025-05-25 00:00:00" "2025-08-24 00:00:00"We can get the aggregate of ice thickness by microwave - we can see the dates from 2024-08-21 to 2025-08-24.
ice_summary
#> # A tibble: 365 × 6
#> time min q1 median q3 max
#> <dttm> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 2024-08-21 00:00:00 1 64 89 98 100
#> 2 2024-08-22 00:00:00 1 79 94 99 100
#> 3 2024-08-23 00:00:00 1 84 95 99 100
#> 4 2024-08-24 00:00:00 1 93 98 100 100
#> 5 2024-08-25 00:00:00 1 77 94 99 100
#> 6 2024-08-26 00:00:00 1 68 90 99 100
#> 7 2024-08-27 00:00:00 1 55 82 94 100
#> 8 2024-08-28 00:00:00 1 75 93 98 100
#> 9 2024-08-29 00:00:00 1 78 94 99 100
#> 10 2024-08-30 00:00:00 1 57 85 95 100
#> # ℹ 355 more rowsWe can also plot the median sea ice thickness
plotted_ice_summariesAnd even have this as a POLAR CO-ORDINATE system!
plotted_ice_summaries + coord_polar()