Why Evapotranspiration Is the Foundation of Modern Irrigation: the FAO-56 Paper Explained

The number almost nobody checks before watering

Anyone who deals with irrigation — professionals and ordinary homeowners alike — tends to fall back on rules of thumb: 'twice a week', 'early morning', 'a finger in the soil'. Those are useful heuristics, but none of them answers the underlying question: how much water did the garden actually lose in the last 24 hours? That question has a precise name — evapotranspiration — and a reference document that has been the standard used by agronomists, smart controller manufacturers and weather services worldwide for 27 years: the FAO Irrigation and Drainage Paper No. 56.

It's not a minor document: it's among the most cited publications in FAO's history, with over 11,500 citations in scientific articles. And yet outside agronomy circles almost nobody has actually read it — most online content that name-drops 'Penman-Monteith FAO-56' does so as a buzzword, without explaining what the document actually says or why it became the standard. This is an attempt to close that gap, in language readable by someone who tends a garden, not someone who studies one.

The paper at a glance

The question: why a universal method was needed

Before 1998, anyone estimating a crop's or a lawn's water needs had roughly a dozen different methods to choose from — Blaney-Criddle, Hargreaves, Thornthwaite, Class A pan evaporation — each calibrated to a specific climate and often disagreeing with each other by 30-50% on the very same site. An agronomist in California and one in southern Italy could apply different formulas and get non-comparable results, even starting from the same raw weather data.

FAO had already tried to solve this in 1977 with the earlier Paper No. 24, but subsequent tests against lysimeter networks (tanks that directly measure water lost from soil, the most reliable instrument for validating any formula) showed that attempt at standardisation also lost accuracy outside the climate it had been calibrated on. The working group led by Richard Allen started over with a specific goal: a single, physically grounded equation, valid everywhere in the world, that required only easily measurable weather data.

What they did: from the original Penman to the FAO-56 version

The starting point is the Penman-Monteith equation, which combines the energy balance (how much solar energy reaches the surface) and aerodynamic transport (how effectively wind and humidity carry water vapour away) into a single physical formula. Unlike empirical methods, it isn't calibrated to a specific climate: it derives from thermodynamics and atmospheric physics, so in theory it works the same way everywhere.

FAO-56's specific contribution was standardising the parameters of a 'reference crop' — a hypothetical 12 cm-tall grass surface, with fixed surface resistance and albedo — so the result (called ET0, reference evapotranspiration) is the same regardless of who calculates it, as long as the same weather inputs are used: maximum and minimum temperature, relative humidity, wind speed at 2 metres, and solar radiation (or sunshine hours, from which it's estimated).

The working group then validated the resulting equation against available lysimeter networks across very different climates — from semi-arid Idaho to the Mediterranean — systematically comparing results against every previous method on the same datasets.

What they found: why Penman-Monteith won

The paper's central finding is that FAO-56 Penman-Monteith showed the best agreement with direct lysimeter measurements across nearly every climate tested — while simpler methods (which use only temperature, like Hargreaves or Thornthwaite) proved reliable only in climates similar to the one they were originally calibrated on, with significant systematic errors elsewhere: some overestimate in humid climates, others underestimate in windy or arid ones.

That's where the practical two-step formula now running under the hood of every modern smart controller (Rachio, Hunter Hydrawise, Orbit B-hyve) comes from: first calculate ET0 (the standardised reference lawn's water need) from weather data, then multiply by a crop coefficient Kc specific to the actual plant — a mature lawn typically has a Kc between 0.6 and 0.8 depending on the season, an established Mediterranean hedge often less. ETc (the real water need) = Kc × ET0. It's the same principle already covered, from the smart-controller side, in /enIrrigation automations: schedules, weather and sensors in practice.

A concrete figure to picture this: on a summer day in the 90s°F with moderate wind and clear skies, ET0 typically runs 6-8 mm; on a cloudy late-spring day it can drop to 2-3 mm. With the same Kc of 0.7, the lawn's real water need on those two days differs by a factor of three — a difference no fixed 'twice a week' rule can capture.

What it actually means for your garden

The most direct practical consequence is that a fixed-schedule timer is, by design, almost always wrong: it delivers the same amount of water regardless of whether yesterday was 95°F, dry and windy or 72°F and overcast. ET-based (weather-based) controllers solve exactly this, pulling local or nearby-station weather data and recalculating ET0 daily to automatically adjust cycle length — a documented water saving of 20-40% versus a fixed programme, with equal lawn health.

For anyone without (or unwilling to buy) a smart controller, the principle is still useful manually: on hot, windy, dry days the lawn loses more water and needs longer or more frequent watering; on cool, humid days the need drops sharply, and sticking to the same schedule just wastes water and encourages root rot.

How this connects to SprinklerMap: what it confirms, and what it does NOT do

Worth being precise here: SprinklerMap's design tool does not calculate evapotranspiration — that's not what it's built for, and we don't want to imply otherwise. The tool handles the hydraulic side: sprinkler placement, coverage, available pressure per zone, material lists.

The connection to FAO-56 is conceptual but real: the paper itself, in its chapters on practical application, makes clear that an accurate ET calculation is pointless if water isn't distributed uniformly across the ground — a poorly placed sprinkler that soaks one spot twice as much as another defeats any precision in the theoretical water-need calculation. Correct hydraulic design (the work SprinklerMap does) is the precondition for any ET calculation, with any method, to be worth applying. For the soil-based run-time calculation side, the practical guide is /enHow to set irrigation run times by soil type.

Frequently asked questions about evapotranspiration and FAO-56

Are ET0 and ETc the same thing? No: ET0 is the water need of a standardised reference lawn, the same everywhere for the same weather; ETc is the actual plant's real water need, obtained by multiplying ET0 by the crop coefficient Kc.

Do I need a weather station in my garden to use these calculations? No: most smart controllers pull data from nearby public weather stations or services like Open-Meteo; a personal station (like an Ecowitt) improves accuracy but isn't essential.

Is the FAO-56 method still considered valid in 2026? Yes, it remains the internationally recommended reference method; later revisions and refinements exist (we'll cover them in the next article in this series), but the 1998 conceptual framework hasn't been superseded.

Why not just use temperature, which is easier to measure? Because temperature alone ignores wind and humidity, two variables that can double or halve real evapotranspiration at the same temperature — exactly the limitation the paper documents for simpler methods.

The short version

FAO-56 isn't an academic footnote: it's the reason 'how much to water' finally has a calculable answer instead of a rule of thumb. Penman-Monteith beat earlier methods because it's physically grounded, not calibrated to a single climate — and every smart controller today promising to 'save water by adapting to the weather' is, under the hood, applying exactly this 1998 paper. Understanding the principle makes it easier to interpret whatever number your smart controller's app shows you.