When to Irrigate: A Practical Guide to Scheduling That Saves Water and Money

Most farmers I talk to worry about how to irrigate: drip tape vs. sprinklers, pump sizing, filtration. Almost nobody asks the question that matters more: when should you actually turn the water on?

That’s a mistake. A perfectly designed drip system running on a bad schedule wastes as much water as a leaky hose. And water isn’t free. In parts of California, agricultural water runs $200-400 per acre-foot. In southern Spain, it’s climbed past €0.30 per cubic meter. In parts of India, farmers pay with diesel to run pumps. Every unnecessary irrigation cycle burns money twice.

Getting irrigation scheduling right isn’t complicated. It doesn’t require expensive sensors or a degree in agronomy. What it needs is paying attention to a few signals your soil and crops are already sending you.

Why most irrigation schedules are wrong

The default approach on most farms is “irrigate every X days.” Every 3 days. Every 5 days. Twice a week. The problem with calendar-based scheduling is obvious once you think about it: plants don’t read calendars.

A tomato plant on a 35°C day with low humidity loses 6-8 mm of water through transpiration. The same plant on a 22°C cloudy day loses maybe 2 mm. If you’re irrigating the same amount both days, you’re either underwatering when it’s hot or dumping water into deep drainage when it’s cool.

Researchers at UC Davis tracked this across 42 farms in the Central Valley. The farms using fixed schedules averaged 27% more water than those using demand-based scheduling, with no yield difference. That’s not a rounding error. That’s real money left in the field.

The feel test actually works

Before we get to sensors and spreadsheets, let’s talk about the oldest method in farming: sticking your hand in the dirt.

The “feel and appearance” method has been formalized by the USDA. You grab a handful of soil from root depth (15-30 cm for most vegetables, deeper for trees) and squeeze it into a ball. Different soil textures behave differently when they’re at the right moisture level:

Sandy soil at field capacity: forms a weak ball, leaves slight moisture on your palm. When it’s time to irrigate: falls apart completely, no moisture stain.
Loam at field capacity: forms a ball, doesn’t crumble, leaves heavy moisture stain. Time to irrigate: barely forms a ball, feels dry, crumbles easily.
Clay at field capacity: forms a ball, ribbons out when squeezed, leaves very thick moisture stain. Time to irrigate: holds shape but cracks appear, feels dry to touch.

I’ve watched farmers in Israel use this method for decades on high-value crops. It’s not primitive. It’s just not automated. The catch is you need to actually do it: walk the field, dig down to root depth, check. On a 2-hectare vegetable operation, that’s 10 minutes of work. On 50 hectares of field corn, you’ll want something faster.

Evapotranspiration: the number that runs everything

If you want to get more precise, the concept to understand is evapotranspiration (ET for short). It’s the sum of water evaporated from the soil surface plus what the plant transpires through its leaves. Think of it as the plant’s daily water bill.

ET data is free. The FAO publishes reference ET (ET₀) for thousands of weather stations worldwide. Your local agricultural extension service almost certainly publishes daily ET values. In the US, CIMIS (California Irrigation Management Information System) gives you station-level data updated hourly. In Australia, the Bureau of Meteorology does the same through its Water and the Land service.

You take the reference ET and multiply it by a crop coefficient (Kc) to get the actual crop water use. A young tomato crop has a Kc around 0.4. A mature fruiting tomato crop is closer to 1.1. FAO Irrigation and Drainage Paper 56 lists Kc values for basically every crop you’d grow.

The math is simple: if yesterday’s ET₀ was 6 mm and your crop’s Kc is 0.8, the crop used 4.8 mm of water. To convert to volume: 1 mm over 1 hectare = 10,000 liters. So that’s 48,000 liters per hectare your crop consumed yesterday. That’s how much you need to put back.

When ET gets tricky

The limitation of ET-based scheduling is that it assumes water is always available. In reality, your soil might have 50 mm of plant-available water in the root zone. If the crop uses 5 mm per day, you’ve got 10 days before stress sets in. On a sandy soil, that number drops to maybe 4 or 5 days.

This is where the concept of “management allowed depletion” (MAD) comes in. For most vegetables, you don’t want to deplete beyond 30-40% of available water. Below that, yield starts dropping before the plant even shows visible stress. For deep-rooted crops like alfalfa or mature orchards, you can push to 50-60%.

I’ve seen lettuce growers in Arizona hit a 15% yield improvement just by switching from a fixed 3-day schedule to ET-based scheduling with a 30% MAD trigger. That’s not incremental. That’s the kind of improvement you’d install a whole new system to get.

What about soil moisture sensors?

Sensors are great if you can afford them. A tensiometer costs $30-80 and measures soil water tension directly. When the reading hits 30-50 centibars for most vegetables, it’s time to irrigate. Capacitance sensors run $100-300 per unit and give you volumetric water content as a percentage.

But here’s the thing: sensors tell you the current moisture level. They don’t tell you how much to apply. You still need to know your soil’s field capacity and the crop’s rooting depth. A sensor reading of 25% volumetric water content means completely different things in sand versus clay.

I’d argue that for farms under 20 hectares, the feel test plus free ET data gets you 90% of the benefit of sensors at zero incremental cost. Put the money you’d spend on sensors into better filtration or pressure regulation instead. Those solve problems that scheduling alone can’t fix.

When to irrigate during the day matters too

Time of day affects how much water actually reaches the roots. Early morning (4-7 AM) is ideal: wind speeds are low so drift is minimal, evaporation is minimal, and the crop is about to enter its peak transpiration period. Late evening is second best, though it increases disease pressure in humid climates.

Midday irrigation in hot, windy conditions can lose 15-25% to evaporation before water even hits the soil, and that’s with drip. With sprinklers, the number can hit 35%.

One exception worth knowing: in extremely hot conditions above 40°C, a brief midday sprinkle on sandy soils can cool the root zone enough to prevent heat stress. Israeli researchers documented this in bell pepper fields in the Negev. A 3-minute midday pulse reduced canopy temperature by 4°C and improved fruit set by 12%. But this is a narrow use case. For almost everyone else, early morning wins.

The simplest approach that actually works

If you take nothing else from this article, here’s a workflow that costs nothing and works on any farm:

Find your local ET data source. Your extension agent knows it. Bookmark it.
Look up the crop coefficient for what you’re growing at its current growth stage. FAO 56 has the table.
Multiply ET₀ by Kc. That’s your daily water use.
Check your soil by feel once a week at root depth. If it’s drying faster than the ET math predicts, your Kc might be off or your soil’s holding less water than you assumed.
Irrigate early morning. Apply what the ET math says the crop used since last irrigation, adjusted for any rain.

That’s it. No subscription fees. No sensors to calibrate. No PhD required. Just some basic arithmetic and the willingness to walk your field once a week.

The farms that get this right aren’t necessarily the ones with the most expensive equipment. They’re the ones that treat irrigation as a decision, not a routine.