A hot, windy, uphill mile is not vaguely harder. It is harder by an amount, and the amount is made of separate charges that can be listed one at a time. This page lists them.
Because a single adjusted number is unfalsifiable, and a list of charges is not.
Most conditions calculators give you one output: run 8:04 instead of 7:15. You cannot argue with that. You cannot tell whether it is mostly the heat or mostly the hill, whether the model thinks the wind matters more than you do, or which line to be sceptical about. It asks for trust and offers nothing to check.
An itemised bill is checkable. If the heat line says twenty-eight seconds and your own experience says fifteen, that is a specific disagreement about a specific number, and it is one the app can eventually settle with your own runs. Meridian would rather be wrong in a way you can point at.
They multiply. They do not add.
Each factor is a fractional cost against the pace it is applied to, and they are applied in sequence rather than summed. Heat costing 6% and a climb costing 5% is not 11% — it is 1.06 × 1.05, or 11.3%. The difference is small at these sizes and stops being small in a marathon on a hot, hilly course, which is exactly the situation where getting it wrong ruins a race.
That is also why the numbers in the list do not add up to the total perfectly. Each line shows what that factor costs on its own, measured by removing it and seeing what changes — the same way you would find out what one item on a bill was worth. Charges that interact cannot each be given exclusive credit for the overlap.
Dry bulb plus dew point, as a combined sum, past a threshold. Quadratic rather than linear, because the published gradients steepen as conditions worsen. Scaled by time on feet, not distance: the same air costs a four-hour marathoner far more than it costs a two-hour one. Reduced by how acclimated your home climate has made you.
Ely et al. (2007); Vihma (2010); Tyler et al. (2016)Aerodynamic drag scales with the square of your speed relative to the air, so a headwind costs more than the equivalent tailwind returns. On an out-and-back you do not break even. Averaged over every direction unless a heading is known.
Pugh (1971); Davies (1980)Minetti’s fifth-order polynomial for the metabolic cost of running on a slope, integrated over the course rather than applied to the net elevation change. A rolling course with zero net gain still costs something, and treating it as flat is the most common way to get a hilly race badly wrong.
Minetti et al. (2002), Journal of Applied PhysiologyAerobic capacity falls from sea level upward, not from some threshold several hundred metres up. The decline is close to linear and the effect on pace is smaller than the effect on capacity, because running economy improves slightly in thinner air. Offset by time spent living high.
Wehrlin & Hallén (2006), European Journal of Applied PhysiologyA flat multiplier, and the least defensible line on the bill. Track is slightly faster than road, trail slower, sand much slower, and the figures are coaching consensus rather than anything measured in a lab. It is on the receipt as its own line so you can see how much of the total rests on the weakest evidence.
Coaching consensus — the one line without a paper behind itSleep, stress, fuelling, how the last week went, and whether you are getting ill.
All of these are real and several of them are larger than anything on this page. Meridian leaves them off because it cannot measure them honestly, and a bill with an invented line on it is worse than a short bill. What is here is the part of a hard day that is attributable to the air, the ground and the terrain — the part that is the same for everyone standing on that start line, and the part you can plan around in advance.
TestFlight is opening to a small group of runners across a range of climates.