HOW THE COMPARISON WORKS
The reasoning.
Not a black box.
Every weather concern should have a visible reason. Every missing input should have a visible consequence.
1. A deliberately bounded route
Choose a curated corridor, direction and a departure time within the next 72 hours. The initial release does not compute arbitrary address-to-address routes or optimize a detour. Distances and durations are rounded planning baselines; the checkpoint positions are selected samples, not a continuously sampled road geometry.
2. Match forecasts to estimated arrival
We add each checkpoint’s baseline travel minutes and its share of the total stop allowance to the departure time. The arrival is matched to the NWS hourly period whose start is at or before the arrival and whose end is after it. A missing future period stays unavailable. Temperature is shown in Fahrenheit and wind in miles per hour.
3. Check official active alerts
We request current actual NWS alerts for each checkpoint and match their stated onset/effective and end/expiry times to the arrival. Test and cancelled alerts are excluded. The absence of a matched alert only describes alerts currently issued; new alerts may be issued before a later trip.
4. Apply an explainable weather-concern category
The highest applicable trigger at any checkpoint determines the route category. We do not average away a severe checkpoint. The four categories are fewer, some, elevated and significant weather concerns. They are product heuristics, not NWS warning categories, accident probabilities or safety certifications.
| Input | How it changes the category |
|---|---|
| Forecast rain or air temperature ≤35°F | At least some concerns. Cold air does not establish icy pavement. |
| Snow, thunderstorms, fog, dust or smoke mentioned | At least elevated concerns. |
| Freezing rain, sleet, ice pellets or blizzard mentioned | Significant concerns. |
| Passenger-car sustained wind | Some at 20 mph, elevated at 30 mph, significant at 40 mph. |
| High-profile sustained wind | Some at 15 mph, elevated at 25 mph, significant at 35 mph. Not official vehicle limits. |
| Matched official alert | Elevated; significant when the official severity is Severe or Extreme. |
For a wind range, we use its upper bound. Missing gusts, measured visibility, pavement temperature and true road-relative crosswind are not invented. Precipitation probability is displayed when present but is not treated as observed rainfall or flooding.
5. Compare like with like
We compare your departure with +3, +6 and +9 hours, using the same checkpoints and assumptions. All required forecast fields and alert responses must be available for all windows before we identify a lower category. Ties favor the earlier departure; no improvement is reported unless the category is strictly lower. This is a weather comparison, not a reopening prediction or a GO instruction.
6. Keep road reports separate
Selected Caltrans records are filtered by interstate number and a geographic bounding box. They can include both directions and do not supply a full route-wide closure feed. Record timestamps older than 30 minutes are flagged. Camera metadata freshness does not verify image capture time; camera snapshots are not used to calculate the weather category.
Freshness and resilience
Successful NWS point-to-grid mappings are cached for up to 24 hours, hourly forecasts for 10 minutes and alerts for 5 minutes. Chain-control feeds are cached for 2 minutes and camera metadata for 5 minutes. Forecast issue times older than 24 hours are not used for the comparison. Retrieval and provider timestamps remain separate. Provider failures are never converted into empty success responses.
Use it as a planning aid
Follow official road restrictions, warnings and on-scene instructions. Recheck before departure. Allow for forecast and travel-time uncertainty. Do not use the tool for emergency routing or while driving.
Try a route comparison