Entry 02Lift definitions3 min read

Why usable lift can sit above the boundary layer

The solid usable-lift line on this project’s charts often rides above the dashed boundary-layer line. It looks contradictory only if both lines are assumed to mean “thermal top.” They do not. One comes from a lifted parcel; the other comes from an updraft profile, a sink-rate threshold, and a coefficient choice made in canadarasp’s source.

Hcrit is not one universal quantity

Dr. John “Dr Jack” Glendening’s BLIPMAP documentation defines Hcrit as the height where an average dry updraft falls below 225 ft/min—about 1.14 m/s. He presents it as a practical maximum thermalling height over flat ground. He also says that the assumptions lack quantitative validation and that boundary-layer top may work better over complex terrain.

canadarasp inherits the name and the idea of a sink threshold, but its windgram code asks a different question: how high might the strongest core still beat a 1 m/s sink rate? That is the number this repository ports. “Hcrit” therefore names a family of methods, not a portable value that every forecast system computes identically.

Average-profile and strongest-core formulas

Lenschow and Stephens fitted aircraft observations of convective thermals with a normalized mean updraft shape. In canadarasp’s notation, for height z inside a boundary layer of depth D:

w_average = w* × 1.34 × (z/D)⅓ × (1 − 0.8·z/D)

The source then says the peak at a thermal’s centre is roughly three times stronger. The implementation uses 4.0 instead of 1.34:

w_core = w* × 4.0 × (z/D)⅓ × (1 − 0.8·z/D)

It first rejects the hour if the curve’s maximum—about 2.02 × w*—cannot beat the 1 m/s sink rate. Otherwise it walks upward from 0.25 D, interpolates the first sink-threshold crossing, and stops at cloud base because flying into cloud is not a usable answer.

Average profile versus strongest-core profile

At the same example sink threshold, the coefficient alone moves the upper crossing from inside the boundary layer to above it.

1.34 average profile4.0 core profile

Average profile0.66 Dupper crossing

Strongest-core profile1.08 Dupper crossing

Example threshold0.55sink / w*

Schematic, not a forecast trace. The horizontal position is updraft normalized by w*; the crossing depends on the hour’s actual w* and the 1 m/s sink criterion.

Why usable-lift top can cross boundary-layer top

The boundary-layer line is a thermodynamic crossing: lift a dry surface parcel and find where it is no longer warmer than the model environment. The usable-lift line is a kinematic crossing: evaluate a modelled core until its vertical speed falls to the sink threshold. A strong core carries momentum into the entrainment zone and the canadarasp curve is evaluated beyond z/D = 1, so its threshold can land above the parcel-derived boundary-layer top. Smoothing cloud base and usable-lift top separately can also change the hour-to-hour gap.

The solid line is not “more correct.” The 4.0 coefficient is a pragmatic extrapolation from an average profile. The underlying 1980 observations came from aircraft legs over the ocean during AMTEX, not from paragliders centring thermals in a mountain valley. The code is precise; the physical transfer is an assumption.

Usable-lift top has not been calibrated against flights

Neither coefficient establishes a calibrated flight ceiling for mountain soaring. The archive can be paired with IGC tracks to measure bias by site, hour, wind regime, and pilot population; that study has not been done. The chart therefore names the ported quantity usable-lift top, not predicted maximum altitude. Reading a windgram keeps the boundary-layer and usable-lift lines as different questions.

Primary trail: Dr Jack’s BLIPMAP parameter documentation, Lenschow & Stephens (1980), DOI 10.1007/BF00122351, and the canadarasp implementation.