Eclipse 2028 practical guide
Eclipse Weather Decision Timeline
This page sets out when to decide: five checkpoints from a month out to the morning of the eclipse. For what the long-term records say, see the weather prospects page; for how to read forecasts, the eclipse weather guide.
Thirty days out: remove impossible options
Review access, accommodation, fuel range, winter road exposure and geometry. Do not treat monthly rainfall or cloudy-day means as a forecast for 22 July. Eliminate places the group cannot reach lawfully and safely even if historical climate appears attractive.
Seven days out: watch agreement
Compare successive official forecasts and note whether the broad pattern persists. Record cloud layers, precipitation, wind, frost, warnings and roads separately. A single icon hides timing and vertical cloud structure. Keep primary and fallback sites within realistic fatigue limits.
Seventy-two to twenty-four hours: commit resources
Select a primary site, realistic fallback and no-move option. Confirm access, toilets, parking, horizon and communications. Write the last departure time for the fallback. Brief each driver that a weather option expires when it cannot be reached with margin.
Eclipse morning: observe the trend
Use current official forecasts, radar where precipitation matters, satellite imagery, warnings and direct sky observations. Distinguish low cloud, high cloud and a clear gap moving toward or away from the Sun. Never trespass, stop on a carriageway or use unsuitable roads to chase blue sky.
After the deadline
Set up, rehearse filters and accept the chosen site. Totality behind cloud can still produce darkness and environmental changes. A safe, organised group at an imperfect location is a better outcome than missing the eclipse in traffic.
Keep a written decision log: evidence checked, time, choice and person responsible. It prevents the group from reopening every decision when anxiety rises.
Read cloud evidence at the right scale
Climatology is useful when choosing between regions months ahead; it cannot distinguish the exact cloud layer over one observing site on eclipse afternoon. In the final week, compare the official forecast location, cloud base and cover, precipitation, wind, visibility and warnings. A low total-cloud percentage is not automatically a clear Sun if a broken low layer sits on the relevant horizon.
Use satellite imagery to understand motion, not to create false precision. Note the image time, direction of movement and whether the gap is approaching before maximum rather than merely appearing near the town. A forecast disagreement is a reason to keep a safe primary site and fallback, not a reason to drive farther without a lawful destination.
Write the no-move rule before anxiety arrives
Choose a final relocation time based on the slowest realistic route, parking, walking, setup and contact sequence. After that time, stay unless the current site becomes unsafe. Tell every driver that watching the sky, stopping on a road shoulder, or making an improvised turn is not a weather strategy. Record the decision and its evidence so the group can concentrate on the event.
When official forecasts disagree
Different forecasting agencies use different models, and it is normal for two official sources to diverge on cloud cover for the same afternoon, especially more than three days out. Do not average the disagreement into a false middle number. Instead, note what both agree on - a frontal system's rough timing, a wind change, a pressure trend - and treat the specific cloud percentage as the least reliable part of either forecast. Agreement between sources should raise confidence in a plan; disagreement should trigger the fallback discussion, not a coin toss.
A cloud-cover percentage also is not the same as a clear view of the Sun. A forecast of 30% cloud can still mean a single low band sitting exactly on the relevant horizon at the relevant minute, while 70% cloud elsewhere in the sky leaves the Sun untouched. Read cloud type and altitude alongside the headline percentage where the forecast provides it, and treat a bare percentage from a phone app as a starting point, not a verdict.
If two official sources genuinely conflict on the day itself and there is no time left to wait for agreement, favour the source with the shorter forecast horizon and the finer geographic resolution over a broader regional outlook, and weight direct sky observation from the exact site above either.
Frequently asked questions
When should I stop watching forecasts and just commit to a site?
Around 72 to 24 hours out, once you can name a primary site, a realistic fallback, and a no-move option. After that, keep monitoring but treat the decision as made unless safety requires a change.
Is a clear-sky climatology percentage a promise of a clear eclipse?
No. It's a long-run historical average for a region, not a forecast for 22 July 2028. Use it only to rule out clearly unfavourable regions months ahead; use current forecasts for the final call.
Found an error? See the correction policy.
Put the timeline into practice
Check current named-station cloud statistics, confirm your exact coordinates against the path, and build the 72-hour decision checklist before eclipse week.