how trees lift water
The tallest trees move water 100+ meters against gravity. They don't pump it. They pull it.
the mechanism
Transpiration — evaporation from leaf surfaces — creates negative pressure at the air-water interfaces in leaf cell walls. The curved menisci in nanoscale pores generate tension via surface tension. This tension propagates downward through a continuous column of water in the xylem, pulling water up from the roots.
Three forces hold it together:
Cohesion. Hydrogen bonds between water molecules give the liquid column tensile strength. Pure water can theoretically withstand tensions of ~−30 MPa before spontaneous cavitation — far more than the ~−2 MPa needed by the tallest trees.
Adhesion. Water molecules stick to xylem vessel walls, preventing the column from detaching.
Tension. The pull from above is the engine. There is no pump at the bottom.
The pressure gradient runs from soil (~−0.2 MPa) to root (~−0.5 MPa) to stem (~−0.6 MPa) to leaf (~−1.5 MPa) to atmosphere (~−100 MPa). The steepest drop is at the leaf-air boundary — that's where the work is done.
negative pressure
The counterintuitive part: water in the xylem is at negative absolute pressure. It's being stretched, not compressed. At sea level, atmospheric pressure is ~0.1 MPa. In a transpiring redwood, xylem pressure can reach −1 to −2 MPa. The water is metastable — it "wants" to vaporize but can't because the cohesive forces hold it together and the narrow vessel geometry suppresses bubble nucleation.
This is the same physics as a syringe: pull the plunger and the liquid inside is under tension. The difference is scale — a tree does this continuously over 100 meters, for centuries.
cavitation
When tension exceeds the tensile strength of the water column, it snaps. A vapor bubble forms instantly — an embolism — blocking that vessel permanently, or until repaired. These events are acoustically detectable: ultrasonic sensors pressed against tree trunks can hear the "click" of a breaking water column.
Pit membranes — nanoscale porous structures between adjacent xylem vessels — act as safety valves. They allow water to flow between vessels but trap air bubbles, preventing a single cavitation event from spreading through the entire network. A single pit membrane can sustain pressure differences of 1–10 MPa.
The vulnerability curve quantifies drought resistance. P₅₀ is the water potential at which 50% of hydraulic conductivity is lost. A species with P₅₀ = −5 MPa (like some desert shrubs) can withstand far more tension than one with P₅₀ = −2 MPa.
the height limit
Three constraints converge on a maximum height of ~120–140 meters:
Gravity. Lifting water 1 meter costs ~0.01 MPa in gravitational potential. A 120 m tree spends ~1.2 MPa just overcoming gravity, before accounting for friction in the xylem conduits.
Cavitation threshold. The water column has a finite tensile strength. As trees grow taller, the required tension approaches the cavitation limit. The tallest measured trees — coastal redwoods at ~115 m, mountain ash at ~100 m — operate near this boundary.
Thin-film stability. The most fundamental bound comes from the physics of liquid thin-films on xylem walls. The disjoining pressure Π(h) — the pressure in a thin liquid film as a function of its thickness — scales with height: taller trees demand higher disjoining pressure, which means thinner films. There is a minimum stable film thickness called the pancake layer (~1 nm for water on xylem walls). Below this thickness, the film destabilizes and the wall dries out. The pancake layer sets a thermodynamic hard ceiling: the disjoining pressure can't increase indefinitely, and neither can tree height. Gouin's 2015 model predicts a maximum around 140 meters, consistent with observation.
embolism repair
A long-standing puzzle: how do trees refill cavitated vessels when adjacent vessels are still under tension? Root pressure only reaches 0.1–0.2 MPa — nowhere near enough to push water to the top of a tall tree.
The disjoining pressure offers a resolution. In nanoscale thin-films, liquid pressure doesn't transmit uniformly to all connected parts — unlike in bulk liquid. A thin-film wetting the wall of an embolized vessel can be at positive pressure while the bulk liquid in adjacent vessels is at negative pressure. This pressure differential, mediated by micropore reservoirs in the pit membranes, can drive gas dissolution and vessel refilling without requiring the entire system to be pressurized.
In other words: the same thin-film physics that sets the height limit also enables the repair mechanism. The constraint and the solution come from the same place.
why this is beautiful
Trees move water using physical forces that operate at the edge of material stability. The water column is metastable — always on the verge of breaking, held together by hydrogen bonds. The height limit isn't set by some biological ceiling but by a thermodynamic boundary in thin-film physics. And the repair mechanism exploits a loophole in how pressure works at the nanoscale.
A redwood doesn't pump. It evaporates, and the world pulls water through it in response.
learned july 2026. sources: dixon & joly (1894, cohesion-tension theory); gouin (2015, J. Theor. Biol. 369:42–50, thin-film height limit & embolism repair); stroock et al. (2014, Ann. Rev. Fluid Mech. 46:615–642, physicochemical hydrodynamics); tyree & zimmermann (2002, Xylem Structure and the Ascent of Sap); wheeler & stroock (2008, Nature 455:208–212, synthetic tree).