Yonsei University scheme

YSU is the standard non-local PBL scheme in WRF and probably the most widely used PBL option in regional modelling. Its defining feature is treating convective mixing as a property of the whole boundary layer rather than of local gradients.

Two elements do the work. A counter-gradient term allows heat transport up-gradient, which sounds wrong and is exactly right for a convective boundary layer: thermals carry warm surface air to the inversion regardless of the local gradient they pass through, and a purely gradient-driven scheme cannot express that. And an explicit entrainment layer at the boundary-layer top handles the mixing of warm free-tropospheric air downward, which is what actually controls how fast the boundary layer grows.

The entrainment treatment is the reason YSU tends to produce deeper, drier, warmer boundary layers than local schemes such as MYNN. That difference matters for low cloud: a boundary layer that grows faster and entrains more dry air is less likely to sustain stratocumulus.

Options worth knowing

• topo_wind = 1 applies a topographic correction to surface winds, representing extra drag from sub-grid topography and enhanced flow over hilltops (Jiménez and Dudhia, JAMC, 2012). It works with the PBL scheme only. In complex terrain the uncorrected surface wind bias is substantial, and this matters more than its obscurity suggests.
• topo_wind = 2 is a simpler correction based on terrain variance.
• ysu_topdown_pblmix = 1 enables top-down mixing driven by radiative cooling at cloud top. Relevant for stratocumulus-topped boundary layers, where cloud-top cooling — not surface heating — drives the turbulence. For maritime subtropical work this is the physically appropriate setting and is not on by default.

Pairing and limits

YSU expects the MM5 similarity surface-layer scheme; the two were developed together and unsupported combinations produce quietly wrong fluxes rather than errors.

Its weakness is stable conditions, where the non-local justification does not apply and the scheme reduces to a local formulation without being designed as one. For nocturnal boundary layers a scheme with a proper TKE budget is generally better founded.

See also: PBL schemes generally, MYNN for the local alternative, and the tropical WRF setup.