Weather Analysis and Forecasting
Weather analysis and forecasting examines station models, upper-air charts, thickness, thermodynamic diagrams, and the use of NWP guidance. The exam tests whether you can turn observations into a short-range forecast and verify it honestly.
Core rules
- Surface station models: temperature and dewpoint in °C, sea-level pressure in hPa with the leading 9 or 10 omitted, wind direction as the direction the wind is coming from, and cloud cover in eighths.
- Thickness is proportional to mean layer virtual temperature: Z1000-500 ≈ 5400 m corresponds to a mean temperature near -7°C. Thickness patterns locate baroclinic zones better than surface isobars alone.
- On a skew-T, log-p diagram, parcel path versus environment gives CAPE/CIN and the level of free convection; wind barbs reveal shear.
- Forecast verification uses bias, MAE, POD, FAR, and CSI. Ensemble spread is not the same as forecast error.
Worked example
A 1000-500 hPa thickness is 5400 m. Use the hypsometric equation: Tmean = gΔz/(Rd ln(p1/p2)) = (9.81 × 5400)/(287 × ln2). ln2 = 0.693, so the denominator is 198.9; the numerator is 52,974; Tmean ≈ 266 K, or about -7°C. This mean virtual temperature is consistent with a winter snow threat in Korea when the 540-dam thickness line is near the station, though surface temperature and wet-bulb still decide precipitation type.
Traps
Reading station pressure as sea-level pressure; using wind direction as the direction the wind is blowing toward; treating thickness as surface temperature; ignoring model spin-up and terrain resolution; and reporting a good POD without mentioning a high false-alarm rate.