A cold drink on a hot day does not require the ice to have been made that day. Persian ice houses, called yakhchals, separated the two jobs: obtain ice when the weather allowed it, then protect a large stock until people wanted it in summer.
The dome is the part that catches the eye, but the system has three jobs:
- Supply: bring in water or naturally formed ice and snow where available.
- Production: expose shallow water to suitable winter-night conditions.
- Storage: move the ice into a protected stock and slow the heat reaching it.
Those jobs explain why a freezing basin needs exposure to its surroundings while the ice store needs protection from them.
Start with winter, even in the desert
Dry country can have cold winters. The climate data used in a 2017 study of the Meybod ice house include very hot summer weather and winter temperatures below freezing. “Desert” does not mean that every night is warm.
Some places could bring in natural ice or snow from nearby mountains. Others produced ice in shallow outdoor basins. Workers supplied water, allowed it to freeze during suitable nights, and transferred the ice into storage. Local conditions determined which supply made sense.
A shallow basin has a useful advantage over a deep pond: less water must be cooled for each patch of surface exposed to the night. A tray and a bucket can hold water, but they give the cold surroundings different amounts of work per unit of exposed area.
The wall and the open sky had different jobs
A tall wall shaded the freezing area from winter sunshine. In their architectural study of Iranian ice repositories, Bahareh Hosseini and Ali Namazian describe shade walls, water channels and several forms of storage. Not every installation was a domed building.
The wall helped prevent sunlight from undoing the cooling. Overhead, however, an open view was useful. The basin could lose energy by emitting thermal radiation toward the night sky.
Those requirements are compatible: shade the water from a low sun while leaving much of the sky visible. Putting a roof over the entire freezing surface would change its exchange of radiation. The protected store and the exposed production basin served different purposes.
The selected illustration shows the direction of an energy loss. Its arrow is a schematic explanation, not a measured heat flow or a claim that the moon supplies refrigeration.
Water can be colder than the reported air temperature
Objects emit infrared radiation and receive radiation from their surroundings. When more energy leaves a surface than arrives, it cools. A clear sky can allow substantial outward heat loss; clouds return more radiation toward the ground and reduce that opportunity.
The National Weather Service’s frost explanation makes an everyday version of this distinction. Frost can appear on a surface while the official air-temperature reading remains above freezing. The temperature of a windshield or the ground is not necessarily the temperature measured in a sheltered instrument several feet above it.
Water still has to reach its freezing point to become ice. Radiative cooling explains how its surface can become colder than the surrounding air; it does not let warm liquid skip the freezing process. Nor does one frosty windshield prove that every pond will freeze. Clouds, wind, humidity, water depth and heat arriving from the ground all affect the balance.
What the engineering model establishes
Hareth Pochee, John Gunstone and Oliver Wilton simulated an ice-making pond using a typical meteorological year for Yazd. In their model, radiation toward the sky supplied the largest cooling contribution during the coldest week, with evaporation also significant. Evaporation takes energy away as some water becomes vapour.
That is a result for a specified model and climate dataset. It supports a physical explanation; it is not a record of every historical winter or an ice-output guarantee. The study separately modelled storage losses, with assumptions about soil properties, insulation and the quantity of ice initially stored.
This distinction matters whenever a modern calculation is used to explain an old building. A plausible mechanism and a predicted amount are different claims. Changing the weather, construction or operation can change the prediction.
Keeping the ice takes more than a cool cellar
Stored ice still receives heat. An underground pit shelters it from sunlight, while an enclosing structure and insulation slow the transfer. Hosseini and Namazian describe straw between ice layers and provision for meltwater beneath the main pit. Drainage deals with melting; it does not manufacture additional cold.
Thick walls also respond slowly to changing outdoor conditions. The Department of Energy’s account of thermal mass explains why substantial walls can damp temperature swings, and why the daily heating-and-cooling cycle matters. Thermal mass and insulation are related design considerations with different roles: storing heat is different from resisting its flow.
Neither effect makes an empty cellar a freezer. A useful ice store starts with ice and limits the heat reaching it. Construction slows the loss of a stock that was already frozen. It does not replace the winter production or imported supply that filled the store.
Melting is part of the storage process
Ice absorbs energy as it melts. Some of the stock can disappear while the remaining mixture stays near its melting temperature. The familiar cooler gradually filling with water illustrates the principle: incoming heat is being absorbed, even though the ice supply is shrinking.
The Department of Energy’s thermal-storage factsheet describes ice as a latent-heat storage medium. Modern systems likewise separate the time cooling is produced from the time it is used. That comparison concerns the storage principle, not identical equipment or operating conditions.
There is also a historical limit to the familiar story. The architectural literature acknowledges uncertainty about early development; surviving domes do not justify a precise invention date for every form of ice storage. The method depended on weather or imported ice, water, maintenance and labour. “No electricity” never meant that the summer supply filled itself.
For a related distinction, Egyptian buildings could delay incoming heat without producing ice. In a yakhchal, winter supplied the frozen stock. Summer engineering was largely the work of making it last.
