OPOpenPithMake curiosity visibleGeography lab / Water cycle

About this lab

Four lessons, one account: water moves between ocean, air and land, every share of it is written in the books, and what goes in must equal what comes out.

The water cycle is not a ring of arrows but a conserved account. Every water body has a stock and every pathway has a flux. Vapour stays in the air for only about 9.5 days on average, while the water in the ocean, counted against evaporation, turns over only every three thousand years or so.

The picture is a schematic section from ocean to snow-capped mountains, not any real place. Height and horizontal distance are drawn 1:1, not exaggerated; the topsoil and aquifer in the section at the front of the model are enlarged about 80 times just to be visible, while the soil column of the infiltration lesson is at true scale. Every particle is emitted by the model’s account: the number of particles is in a fixed ratio to the numbers in the account, it is not decoration. Particle counts do not represent stocks.

Four results

Four results

  • Cycle paths: about 486 thousand km³ of water evaporates each year, 413 from the ocean and 73 from land; the same amount falls, 373 on the ocean and 113 on land. Of the 413 evaporated from the ocean, 373 falls back into the sea (sea loop) and 40 is carried onto land and returns by rivers (sea–land loop); the 73 evapotranspired from land condenses over land and falls again (land loop). The three paths share one atmosphere; they are not three sealed containers.
  • Infiltration and recharge: infiltration depends on how well the soil conducts water, not on how many pores it has. Clay’s total porosity, 0.475, is larger than sand’s 0.437, yet its saturated conductivity is only 0.3 mm/h against 117.8 mm/h for sand. The wetting front must reach the bottom of the soil layer before water recharges groundwater; the wetter the soil beforehand, the less extra water it can hold and the earlier recharge starts.
  • Flood peak: in the same 10 mm of rain, the bare-soil tray releases little and late, while the tray that is 80% sealed releases early and all at once. The largest 10-second outflow and the total runoff are different things: the sponge lowers the peak but the total hardly changes; once the sponge is full the peak returns to the sealed tray’s level.
  • Water budget: P + inflow from outside = ET + outflow + ΔS. With P = 200, ET = 120 and Q = 50 mm, ΔS = +30 mm. A reservoir only changes the monthly split and leaves the annual total alone; diversion and pumping are extra outflows that turn ΔS into 5 mm and 0. A multi-year mean ΔS near zero does not mean the store is constant in every month.

Model and sources

Model and sources

  • Global stocks come from the USGS Water Science School (Shiklomanov 1993); fluxes come from Figure 1 of Huntington 2010, which reproduces Trenberth et al. 2007 (cited second-hand from Huntington 2010; the Trenberth paper itself was not opened). The proportions agree with Dai and Trenberth 2025: about 65% of land precipitation returns to the air by evapotranspiration, about 35% runs off to the sea, and about 10% of ocean evaporation is carried onto land by wind. The two sources estimate stocks differently, for example groundwater at 23400 thousand km³ in the first and 15300 thousand km³ in the second: published estimates simply differ, and particle counts in the picture do not represent stocks.
  • Infiltration uses the Green–Ampt model: cumulative infiltration F satisfies F = K·t + ψΔθ·ln(1 + F ÷ ψΔθ), and water ponds only once rain intensity exceeds the conductivity K. Soil parameters for sand, loam and clay come from Table 2 of Rawls et al. 1983; the prior wetness (20% dry, 80% wet of the effective pore space) and the 500 mm soil depth are teaching values. Ponded water runs off at once; depression storage and redistribution of soil water after the rain stops are not included.
  • Urbanisation and flood peaks is an original teaching model: trays of 20 cm × 30 cm, each sprinkled with 10 mm (600 mL) over 60 seconds and recorded in 10-second bins up to 300 seconds. Routing uses a linear reservoir plus a pure delay and is not calibrated to any real basin; only differences between groups are compared. The sponge sits before the outlet and its capacity can be changed. Group D is pre-wetted bare soil, kept apart and never averaged with the other three.
  • The monthly split of the water budget is illustrative, not the climate of any place: January to December only serve to show change within the year. Diverting 25 mm, pumping 30 mm (all consumed) and planting trees that add 15 mm of evapotranspiration are teaching values; reservoir evaporation and ecological base flow are not counted.
Back to geography ↑