Falklands Lee Wave Regime Toolv0.4

Mount Pleasant Airfield · flow regime diagram of Sheridan & Vosper (2006), with terrain sampled from a 90 m DEM

Sounding inputs

zᵢ is the inversion height above mean sea level (for example from an upwind NWP ascent over the sea). H is terrain height above the reference elevation set under “Terrain height H” (default 0 m = mean sea level). Type a value or use the slider; typed values may go beyond the slider range. Wind direction is the direction the wind blows from; the slider covers 270° through 360° to 090°.

Uncertainty (point cloud)

Normal distributions; Δθ, zᵢ and U are truncated so they cannot go below zero. H is drawn from the individual ray maxima (or from the override ± below).

Terrain height H

Flow regime diagram

–

    Terrain sampled upwind — drag on the map, or click it then use ← →, to set the wind direction

     
    ray crest below inversion ray crest above inversion terrain above the inversion (zᵢ above mean sea level)

    What would change the answer

    Smallest single-input change (others held) at which the central estimate moves to a different regime.

      Trail of saved ascents

      No ascents saved. Use “Add to trail” to compare successive soundings on the diagram.
      Caveats and limitations — read before relying on the output
      1. Origin of the diagram. Fig. 4 of Sheridan & Vosper (2006) summarises idealised 2-D simulations (Met Office BLASIUS model) of flow over a single cosine ridge 2.5 km wide and 400 m high, in a two-layer atmosphere: a neutral layer capped by a sharp inversion, with N = 0.01 s⁻¹ and uniform wind above. This tool fixes N at 0.01 s⁻¹ (there is no N input): if the stability above your inversion is very different (check with N² = (g/θ)·dθ/dz on the ascent), the regime boundaries may shift. All runs used U = 8 m s⁻¹; Fᵢ and H/zᵢ were varied through Δθ and zᵢ. Applying it at other wind speeds relies on the non-dimensional scaling holding.
      2. Regimes are nearest-neighbour. The diagram is a grid of 81 discrete simulations (spacing about 0.1 in Fᵢ and 0.1–0.2 in H/zᵢ), not drawn boundaries. This tool assigns each point the regime of the nearest simulated point, so boundaries sit halfway between points by construction. The true boundaries are uncertain by roughly that much. The points were digitised from the published figure; check them against the paper if a result matters.
      3. Limited 3-D validation. The paper tests the 2-D diagram against three 3-D simulations of northerly flow over East Falkland (8 m s⁻¹; Fᵢ = 0.6 and 0.4 with zᵢ = 800 m, and Fᵢ = 1 with zᵢ = 333 m) and refers to field observations (Mobbs et al. 2005). This tool has not been verified against operational outcomes such as closures or pilot reports.
      4. H is not unique. In the paper, the eastern ridge behaves as H ≈ 400 m, the western part of the ridge as a higher H (above 600 m), and the southern ridge is already modified by upstream orography. Gaps, channelling and wakes are 3-D effects the 2-D diagram cannot represent. The automatic H here is the highest terrain along each ray within the sector and fetch you set; a single isolated peak can dominate it. Check the map and profile, and override H where you disagree.
      5. DEM. Copernicus GLO-90 is a 90 m surface model: it smooths summits (so H tends to be slightly underestimated) and heights are relative to the EGM2008 geoid. zᵢ is entered as height above mean sea level and H is measured above the reference elevation (default 0 m, mean sea level), so both share one datum; this suits an inversion height taken from an upwind NWP ascent over the sea. The MPA elevation (71 m) is used only for labelling; the DEM reads about 66 m there.
      6. Input quality. zᵢ, Δθ and U come from one ascent (often an upwind NWP profile over the sea, which may differ from conditions over the island). Choosing the inversion base is subjective, the balloon drifts downwind, the atmosphere evolves between ascents, and real profiles have shear and veer with height whereas the paper assumes uniform wind (its U is the upstream geostrophic wind above the boundary layer). State which level you used for U.
      7. Applicability. The method needs a strong low-level inversion and has been examined for northerly flow at MPA. If the inversion lies below the hill crest (H/zᵢ above about 1.1) it is outside the diagram: the paper notes that flow then tends to go around the ridge and its low-inversion case (zᵢ = 333 m) is not plotted on the diagram.
      8. Flow type, not intensity. The result indicates the type of flow expected downstream, not gust speed, turbulence severity or the likelihood of a closure. Rotors in the simulations are quasi-steady recirculations; in reality they are intermittent and unsteady (the field campaign saw short ~1 h episodes of flow separation).
      9. Probabilities are input-uncertainty only. The percentages show how the chosen error bands propagate through the diagram. They exclude diagram resolution and model error, are not calibrated forecast probabilities, and change if you change the error settings.
      10. Not an operational product. Use alongside NWP, observations and forecaster judgement. Version 0.4 is a prototype.