At a minimum, predicting the absorption spectrum for water vapor requires a set of line positions and intensities and a shape function. Line intensities can be determined from a dipole moment surface, but a team of researchers from École Polytechnique Fédérale de Lausanne in Switzerland, University of Rochester in New York and University College London suggests in the Aug. 9 issue of Science that the best available surfaces are not accurate in the optical spectrum.

The researchers based their observations on a series of dipole moment measurements of water that had been highly excited vibrationally. They did not measure line intensities but relied on a sequence of three pulses approximately 10 ns in width from Nd:YAG lasers to measure Stark splittings in very weak overtone transitions.

The researchers found that these experimental dipole moments differed by as much as 9.2 percent from current predictions from the best dipole moment surfaces.