Atlantic hurricane peak — August 20 through October 10, and the five-day forecast cone that decides everything

The peak window

Per NOAA Climate Prediction Center 30-year averages, the Atlantic hurricane season's peak — when 80% of annual tropical cyclone activity occurs — runs from August 20 through October 10. The single statistical peak day is September 10.

For coastal Atlantic and Gulf households, this is the seven-week window when household-level decisions matter most. Most years end without a direct hit. The years that don't are concentrated in this window.

Reading the five-day forecast cone correctly

The National Hurricane Center five-day forecast cone is the single most important household decision tool. Most households read it wrong. Two things matter:

1. The cone shows likely path of the storm CENTER, not the storm's full extent.

A storm with a 200-mile diameter has effects that extend well beyond the cone. The cone is the cone of probability for the center; the storm's outer bands extend much farther.

The cone is also a probability cone, NOT a deterministic forecast. The actual storm path is typically inside the cone 60-70% of the time at the 5-day forecast. By day 2-3, accuracy improves but still not deterministic.

2. The error decreases dramatically with shorter lead times.

NHC verification data shows median forecast errors: - 5-day forecast: ~175 miles - 4-day: ~130 miles - 3-day: ~95 miles - 2-day: ~65 miles - 1-day: ~35 miles - 12-hour: ~20 miles

This means the cone at 5 days out is a planning tool. The cone at 2 days out is a decision tool. The cone at 12 hours is execution.

Household decision thresholds

When a storm appears in the NHC five-day outlook and your area is in or near the cone:

• Day 5 (120 hours out): Acknowledge the threat. Update your plan. Mental rehearsal. No physical action required yet, but if you have not pre-packed your evacuation bag, do it now. Verify gas tank above half full.
• Day 3 (72 hours out): Decision point. If a Category 3+ storm and your evacuation zone is mandatory in such a storm, leaving now means open roads. Leaving in 24 hours often means traffic jams. Top off gas, charge all batteries, verify documents in evacuation bag are current.
• Day 2 (48 hours out): Last comfortable evacuation window. Roads will be progressively more congested. Hotels in safe regions begin to book up. Decide stay-or-go.
• Day 1 (24 hours out): If you are staying, shutter time. Window protection, sandbags at low entrances, fill bathtub, charge everything, verify shelter location, brief family. If you are leaving and have not yet — leave now, not later.
• 12 hours out: Last call. Many municipalities close traffic windows. If you are still home, you are staying.

Three household-level actions specific to peak season

Beyond the standard hurricane preparedness items (water, food, batteries, medications), three actions specific to peak season are worth noting:

• Verify your insurance situation now. NFIP flood insurance has a 30-day waiting period from purchase to coverage. If you do not have flood insurance and are in a Special Flood Hazard Area, the time to buy it for THIS season is gone (purchases now don't cover anything before September 20). Standard homeowner's insurance does not cover flood damage. Check your policy.
• Identify two evacuation destinations. Primary (where you'd actually go), backup (in case primary becomes inaccessible). Both should be 100+ miles inland from your home. Pre-identify relatives or affordable hotels.
• Charge all medical equipment batteries weekly. CPAP, oxygen concentrators, glucose meters, dialysis equipment — these have battery backups but only if you've charged them. Weekly check is the discipline.

The "shelter in place" decision

Not every household evacuates for every storm. The decision to stay or go depends on:

• Your evacuation zone (A and B = leave; C and beyond = often shelter is reasonable)
• Storm category (Cat 1-2 = often shelter is reasonable for inland homes; Cat 4-5 = leave regardless of zone)
• Specific path (a direct hit at Cat 1 may produce worse damage than a glancing Cat 3)
• Your home's construction (hurricane-rated windows, elevated foundation, sturdy roof = better odds of sheltering)
• Special considerations (children, elderly, pets, medical equipment dependence)

The default assumption "we'll just shelter" is the wrong default for evacuation-zone-A households facing a Cat 3+ storm. The default assumption "we always evacuate" is the wrong default for inland households facing a Cat 1.

The right default is "we've made a decision rule in advance, and we follow it when the forecast triggers it."

What this is not

This is not a prediction that any specific household will face a catastrophic storm this season. The vast majority of coastal residents will not face a direct hit. The point of the cone-reading discipline above is to convert a high-uncertainty decision (will I leave?) into a low-uncertainty decision (when the forecast cone covers my area at 3 days out, I leave; otherwise I stay).

The cost of running this discipline is one focused planning conversation with your household before September. The cost of NOT running it, in a year when a storm directly threatens your home, is mid-storm panic.

One thing this week: look up your specific evacuation zone if you don't already know it. Five minutes on your state emergency management website. Most coastal residents never do it.

— Systems Fail Lab