Meridian is a rule-based reading of published circadian research. It does not copy a commercial jet-lag product, and it does not run a proprietary model. The numbers below are the ones the planner actually uses.
How a plan is built
Markers. Core body temperature minimum (CBTmin) is estimated at 2.5 hours before your usual wake time. Dim-light melatonin onset (DLMO) is estimated at 2 hours before your usual bedtime. Eastman and Burgess (2009) place the temperature minimum about 3 to 4.5 hours before wake for an 8-hour sleep, and sleep onset about 2 to 2.5 hours after DLMO. These are field estimates, not a lab measurement of your phase.
Direction. If the destination clock is less than 12 hours ahead, the plan advances (eastward). If it is less than 12 hours behind, the plan delays (westward). At about 12 hours or more, it compares how long each direction would take and picks the faster one. Advances use a smaller daily cap, so they usually lose that comparison.
Daily caps. Gentle is 0.75 h advance / 1 h delay. Balanced is 1 h / 1.5 h. Aggressive is 1.25 h / 2 h. Balanced is near the field rates Eastman and Burgess cite (about 57 min/day advancing and 92 min/day delaying). Aggressive approaches their bright-light working estimates, with the advance kept under the 1.5–2 h/day sleep shift they found ran ahead of the clock.
Light. To advance, seek bright light in the 4 hours after CBTmin and avoid it in the 4 hours before. To delay, do the reverse. A softer “some light” block sits beside the bright block so the exposure stays continuous. Seek-light is never placed inside scheduled sleep. When the delay lobe falls entirely during sleep, bright light moves to the last 2.5 waking hours before bed — still before CBTmin, and long rather than pulsed.
Sleep, naps, melatonin, caffeine. Sleep shifts with the markers, up to the daily cap. A short optional nap can appear on the flight day and the arrival day. For advances, 0.5 mg melatonin is placed 3 hours before that day’s DLMO, the middle of the 2–4 hour advance peak in Burgess et al. (2010). A time 5 hours before DLMO is the older working estimate for a 3 mg dose, which this app does not use. Morning 0.5 mg is suggested only for delays of 6 hours or more, with a caution. Caffeine is allowed from wake until 8 hours before bedtime.
Days. Shifts under 3 hours get 2 prep days. Larger shifts get 3. Those days are shown in origin time. From the arrival date on, days are shown in destination time until the shift is finished.
Papers
Khalsa et al., 2003
Khalsa SBS, Jewett ME, Cajochen C, Czeisler CA.
A phase response curve to single bright light pulses in human subjects.
Light before the core body temperature minimum delays the clock; light after it advances the clock. The plan uses about four hours on the sensitive side of that minimum and does not ask you to seek light while you are scheduled to sleep.
Shift the sleep schedule with the clock instead of jumping onto destination time overnight. Field estimates in the paper are about 57 minutes a day advancing and 92 minutes a day delaying; with timed bright light their working estimates were about 1.5 hours a day advancing and 2 hours a day delaying. Pre-flight days, morning light to advance, evening light and dim mornings to delay, and short naps come from this practical framework.
Burgess et al., 2010
Burgess HJ, Revell VL, Molina TA, Eastman CI.
Human phase response curves to three days of daily melatonin: 0.5 mg versus 3.0 mg.
For a 0.5 mg dose, the largest advances were when melatonin was taken about 2–4 hours before the dim-light melatonin onset (the fitted peak was about 3.3 hours before). The delay region was less distinct and peaked soon after wake time. The plan times 0.5 mg three hours before the estimated DLMO for advances, and only suggests a morning dose for delays of 6 hours or more.
Serkh & Forger, 2014
Serkh K, Forger DB.
Optimal schedules of light exposure for rapidly correcting circadian misalignment.
Published under CC BY. In their model, sustained light re-entrains the clock faster than short pulses, and the optimal pattern is one long light block and one dark block each day. The plan keeps bright light and the softer light beside it as one continuous stretch.
Drake et al., 2013
Drake C, Roehrs T, Shambroom J, Roth T.
Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed.
A 400 mg dose of caffeine still disrupted sleep when taken 6 hours before bed. The plan stops caffeine 8 hours before the target bedtime, a buffer past that result, and allows it earlier in the day for alertness.
A 5-hour delay of meals shifted the rhythm of blood glucose and did not shift melatonin. Meal timing is treated here as a supporting habit for peripheral clocks, not as a way to move the main body clock. Feast-and-fast schedules such as the Argonne diet are not used; evidence for them is limited.
Serkh & Forger license
“Optimal schedules of light exposure for rapidly correcting circadian misalignment” is published in PLOS Computational Biology under the Creative Commons Attribution license. Meridian uses the paper’s finding that sustained light beats short pulses. It does not reproduce the paper’s figures, tables, or model code.
What this app does not do
It does not measure your melatonin or temperature. It does not know if you are a morning person or a night owl, and those differences change how fast a clock will move. It plans one destination, with an optional single connection. It ignores seasons of daylight except through the civil time zone, including daylight-saving rules, on each date. A 23- or 25-hour daylight-saving day is still drawn as a 24-hour clock.
Eastman and Burgess often discuss delaying, rather than advancing, once an eastward trip is about 8 time zones. Meridian only runs that comparison once the clock change reaches about 12 hours, which is the rule this version is built to.