Sail window/Lake Dunmore, Vermont
Off season.
A yes-or-no read on when the conditions actually allow sailing this week, built from the Open-Meteo forecast and four plain gates.
NWS is higher today: peak 10 kt against Open-Meteo's 7 kt.
Lake Dunmore is out of season from November 2 through April 30. The forecast below is shown for reference only.
Next seven days, hour by hour
Each bar is one hour of sustained wind on a fixed 0 to 25 knot scale. The shaded band is the 5 to 20 knot target. Green bars clear every gate and are sailable... everything else fell short somewhere. Arrows point the way the wind is blowing. A small sky icon under each hour reads cloud cover, from a warm sun to a grey cloud, and the thin bar under each hour is the chance of rain, going solid once it crosses the 30% gate. Hover an hour for the full detail. Scroll sideways for the whole week.
Sailable windows
No stretch of 3 or more hours cleared every gate this week. That is a real answer, not a gap... the conditions simply do not line up.
Adds a calendar that keeps itself current: windows appear and disappear as the forecast changes, so it never goes stale. The events show as free, never busy.
If that opens a browser instead of your calendar, your system has webcal: pointed at the wrong app. Paste this into your calendar's subscribe or add by URL field instead... it does the same thing, and it is also what Google Calendar wants:https://sail-window.vercel.app/dunmore/calendar.ics
Near misses
Stretches of 3 or more hours that failed exactly one gate. Never sailable, shown so an empty week is distinguishable from a broken page.
Nothing came close.
How this decides
A weather forecast tells you what the weather will be. That is not the question. The question is when this week you can actually go sailing, which has a yes-or-no answer. Every hour is judged on its own against four gates, and it either clears all four or it does not. No scoring, no weighting, no partial credit.
Computed for the spot's own latitude and longitude, so it tracks the season instead of assuming a fixed window. Anything after dark is discarded before the other rules run.
Below 5 knots a keelboat is drifting with opinions. Above 20 it stops being relaxing. This is the band, not an aspiration.
A separate field from sustained wind, and the two diverge constantly in valley terrain. Ten sustained gusting twenty-two is a different day from a steady fourteen, and an average cannot tell them apart. So every window names its strongest gust, not just its sustained range... a day can sit comfortably inside the wind band and still be a handful.
The forecast gives a probability, not a yes or no, so the rule needs a line. The 30% mark is the conventional "slight chance" cutoff... it keeps usable days without pretending rain is impossible.
From hours to windows
Consecutive passing hours merge into one window, and any window shorter than 3 hours is thrown away... roughly what it takes to get out, sail properly, and come back without watching the clock. A lone good hour looks encouraging on a chart and is useless in practice. Windows are listed soonest first, not ranked... they have all cleared every gate, so the only thing that separates them is when they happen.
Near misses, and why they exist
Pass-or-fail rules throw away the most useful thing the forecast knows: how close a failure was. An hour just below the wind floor fails exactly like an hour of dead calm. So any stretch of 3 or more hours that failed exactly one gate is reported separately, with the gate named and the margin shown. It is never presented as sailable. It exists because a page that is empty for a week is otherwise indistinguishable from a page that is broken.
What it deliberately does not do
- It does not pretend to know the wind on the water. The underlying model grid is several miles across, coarser on longer forecasts, and terrain finer than that is smoothed away. No public forecast resolves a single lake or bay at that scale, and claiming otherwise would be the easiest lie in the whole system.
- It ignores cloud cover. Overcast and 12 knots is a good sailing day. Treating sunshine as a requirement would delete a large share of perfectly good afternoons for what is really a comfort preference.
- It runs May through November only. Outside the season a fine forecast can still clear every gate on a day the water is frozen or otherwise out of use.