ATPL Meteorology common mistakes
ATPL Meteorology · Understand the traps and how to avoid them.
Reading the QNH from a METAR and using it directly to compute a true altitude or terrain clearance without applying the temperature correction, treating pressure altimeter error and temperature altimeter error as the same thing.
Candidates learn 'QNH gives elevation' as a single rule and never separate the pressure-error correction (already handled by setting QNH) from the residual temperature error, which persists even with a correct QNH and grows with height above the station and with ISA deviation.
What to do instead
Adopt a fixed two-step routine: step 1, set QNH so the altimeter reads elevation on the ground (pressure error removed); step 2, always ask 'is the mean temperature of the column colder or warmer than ISA?' and apply the rule that cold air puts the true altitude below the indicated altitude. Practise with a worked example: airfield elevation 1000 ft, QNH correct, OAT such that the column is 20 degrees C colder than ISA, giving roughly 4 percent per 10 degrees of error, so true altitude is about 80 ft lower than indicated.
Classifying a stability problem by looking only at the environmental lapse rate against 1.98 degrees C per 1000 ft, without checking whether the lifted parcel is saturated and therefore whether to compare against the dry or saturated adiabatic lapse rate.
The 1.98 figure is the most memorised number in the syllabus, so it becomes the default comparator; candidates forget that once the parcel reaches the condensation level the correct comparator is the much smaller saturated rate, which flips the answer from stable to conditionally unstable.
What to do instead
Force yourself to write three lines before answering: (a) environmental lapse rate, (b) is the parcel saturated at the level in question, (c) therefore DALR or SALR. Then state the rule explicitly: stable if ELR < SALR, conditionally unstable if SALR < ELR < DALR, unstable if ELR > DALR. Also note that conditional instability is the normal state of the troposphere, so 'conditionally unstable' is often the correct answer when the layer is unsaturated and the ELR sits between the two rates.
Decoding a TAF by reading the change groups as a chronological sequence of weather rather than as conditional or temporary amendments to the prevailing conditions.
TAFs are read left to right like a narrative, and abbreviations such as TEMPO, BECMG, PROB30 and FM are treated as 'then this happens' rather than 'this may temporarily interrupt the prevailing conditions' or 'a gradual transition occurs within this window'.
What to do instead
Annotate every TAF in a fixed order: prevailing wind, visibility, weather, cloud, then each change group labelled with its type and time window. Memorise the semantics: FM starts a new prevailing state at a precise time; BECMG describes a gradual change occurring at an unspecified moment within the window; TEMPO describes fluctuations lasting less than an hour each and, in total, less than half the period; PROB gives a probability of the condition occurring. Then answer questions about 'the worst case during your flight window' by combining prevailing plus TEMPO, not by picking whichever group appears last.
Treating the standard atmosphere as a description of the real atmosphere and applying ISA values (15 degrees C, 1013.25 hPa, 1.98 degrees C per 1000 ft) as if they were the actual conditions at a given place and time.
The ISA is taught first and heavily drilled, so its constants become a mental default; candidates then answer 'what is the temperature at 20 000 ft' with the ISA value when the question supplies actual surface data, or fail to notice the question is asking for a deviation.
What to do instead
Before any calculation, underline whether the question says 'standard atmosphere' or gives actual surface conditions. If actual data are given, compute the ISA temperature at the level first, then subtract to get the ISA deviation, and carry that deviation through. Keep a one-line definition on your scratch sheet: ISA is a reference model with defined sea-level values and defined lapse rates, not a forecast.
Assuming that because the wind is reported as calm or light at the surface, mechanical turbulence, low-level wind shear and standing-wave activity can be disregarded.
Surface wind is the only wind most candidates routinely read, and the relationship between surface wind, gradient wind and the free-air wind above the friction layer is under-practised; the effect of stable air over hills producing mountain waves and rotors is also easily forgotten.
What to do instead
Build a habit of comparing surface wind with the forecast wind at 2000 ft and 5000 ft (or the significant-weather chart's upper winds) and flagging any marked change in speed or direction as shear. Remember the friction-layer rule that surface wind is backed and reduced relative to the gradient wind over land, so a strong gradient wind aloft can coexist with a modest surface report. For hills, ask whether the air is stable and the wind at ridge level is strong; if both are true, expect wave activity and rotor turbulence in the lee.
Diagnosing icing risk from the presence of cloud or from the freezing level alone, without considering the liquid water content, the temperature band and the type of cloud.
Candidates reduce icing to 'below freezing plus cloud equals icing', which over-predicts in cirrus and under-predicts the severity difference between stratiform and cumuliform cloud and the importance of the temperature range around 0 to minus 15 degrees C.
What to do instead
Use a three-part checklist: temperature at the level (most severe between about 0 and minus 15 degrees C, with supercooled large droplets possible nearer 0), cloud type and depth (deep stratiform gives widespread rime, cumuliform gives clear ice with high liquid water content), and whether the air is stable (stable stratiform cloud is more likely to hold supercooled droplets). Also distinguish freezing rain, which implies a warm layer aloft and severe clear icing, from ordinary in-cloud icing.
Reading a synoptic or significant-weather chart by locating the fronts and pressure centres but not deducing the wind, visibility, cloud and flight conditions in the sector between them.
Chart questions are treated as identification tasks ('where is the cold front?') rather than as interpretation tasks ('what will you fly through after it passes?'), so the candidate never links front position to the actual weather sequence and wind veer.
What to do instead
For every front on a chart, write four things: direction of movement, the wind before and after passage, the visibility and precipitation pattern, and the cloud sequence. Attach the standard model to each front type (warm front: cloud base lowering, visibility falling, continuous rain, wind veering; cold front: narrow band of heavy showers, squall line possible, rapid clearance and veer with a sharp temperature drop). For lows, note that the strongest gradient and worst weather are usually near the centre and along the fronts, and that the warm sector gives low cloud, drizzle and poor visibility.
Confusing the several pressure settings and what each one makes the altimeter read, mixing up QFE, QNH, QFF and the standard pressure setting when computing height, altitude or flight level.
The four settings are learned as a list of definitions rather than as four distinct uses, and the same altimeter is described in three different ways depending on which is set, so candidates swap them under time pressure.
What to do instead
Fix the mapping in one line each: QFE set gives height above the reference datum (usually the aerodrome); QNH set gives altitude above mean sea level; 1013.25 set gives flight level (pressure altitude); QFF is the mean sea level pressure reduced using actual temperature and is a meteorological quantity, not normally set on the altimeter. Then note the transition: below the transition level you fly altitudes on QNH, above it you fly flight levels on standard pressure, and the transition layer absorbs the difference.
Treating fog as a single phenomenon and predicting it from humidity alone, without distinguishing radiation fog, advection fog, frontal fog and orographic fog by their formation mechanisms and required conditions.
Fog is memorised as 'visibility below 1000 m' with a list of types, but the diagnostic conditions (clear sky and light wind for radiation fog, warm moist air over a cold surface for advection fog, mixing or frontal lifting for frontal fog) are not tied to the forecast reasoning.
What to do instead
For each fog type, memorise the trigger conditions as a short recipe: radiation fog needs clear skies, light wind, long night, moist boundary layer and often an anticyclone; advection fog needs a warm moist air mass moving over a colder surface, with moderate wind, and can occur at any time; frontal fog occurs in the precipitation zone of a front; orographic fog forms as moist air is forced upslope. Then, in a question, identify which ingredient the scenario supplies and match it to the type rather than guessing.
Answering thunderstorm questions by recalling that storms need moisture, instability and a lifting mechanism, but failing to apply the resulting hazards and the storm's life cycle to the flight-planning decision being asked about.
The three ingredients are learned as a definition to be recited, so the candidate cannot move from them to the practical consequences: downbursts, microbursts, hail, icing above the freezing level, lightning, and the fact that the mature stage is the most hazardous.
What to do instead
Convert the ingredients into a hazard map: unstable air plus moisture plus a trigger gives cumulonimbus, which brings a mature stage with the strongest up- and downdrafts, hail, severe icing and microburst risk. Then practise the decision rules: avoid the cell by a wide margin and never fly beneath it, expect wind shear and a sudden wind shift on the surface, and plan around the diurnal maximum in the afternoon over land. Link the life cycle stages to the hazards so a question about 'when is it most dangerous' is answered from the mechanism, not memory.