Problem
cd_compare() requires the user to specify both time windows explicitly. There's no sensible default that helps a first-time user understand what climate departure looks like for their AOI.
Status (post-v0.2.1)
The vignettes have settled on a de-facto framing — recent 2015-2025 vs WMO-style 1951-1980 baseline, paired with both 75-yr and 45-yr Theil-Sen trends. Both peace-fwcp and kootenay-lake use this consistently. The methodology lit review from #54 (archived at planning/archive/2026-05-issue-53-snow-lit-review/findings.md) backs the choice — Mann-Kendall + Theil-Sen with no prewhitening per Yue & Wang 2002 is correct for our 76-yr series with strong trends.
Remaining scope
Wire those windows in as cd_compare() defaults so a first-time user gets a sensible answer without having to supply both windows. Document the rationale (cumulative-impact framing — "2 °C warmer" punchline rather than per-year rate) in the function's roxygen plus a brief note in the vignettes.
Tightly coupled with #43 (window-vs-window p-value); both should land together.
Out of scope
Problem
cd_compare()requires the user to specify both time windows explicitly. There's no sensible default that helps a first-time user understand what climate departure looks like for their AOI.Status (post-v0.2.1)
The vignettes have settled on a de-facto framing — recent 2015-2025 vs WMO-style 1951-1980 baseline, paired with both 75-yr and 45-yr Theil-Sen trends. Both
peace-fwcpandkootenay-lakeuse this consistently. The methodology lit review from #54 (archived atplanning/archive/2026-05-issue-53-snow-lit-review/findings.md) backs the choice — Mann-Kendall + Theil-Sen with no prewhitening per Yue & Wang 2002 is correct for our 76-yr series with strong trends.Remaining scope
Wire those windows in as
cd_compare()defaults so a first-time user gets a sensible answer without having to supply both windows. Document the rationale (cumulative-impact framing — "2 °C warmer" punchline rather than per-year rate) in the function's roxygen plus a brief note in the vignettes.Tightly coupled with #43 (window-vs-window p-value); both should land together.
Out of scope