The Limits of WMO Normals in a Warming Climate
WMO climatological normals answer a specific question: what was typical at a reporting station during a standard 30-year period? The World Meteorological Organization defines them as reference statistics for assessing current or recent conditions and describing the conditions likely to be experienced at a location.
That usefulness also creates the main source of confusion. People often ask normals to answer questions about rapid warming, recent extremes, or present-day climate risk. Those are important questions, but normals were never designed to answer all of them on their own.
These are structural limits of climatological normals themselves, not bugs in the Climate Explorer interface.
The short answer is that normals remain useful, but only when the question matches the statistic. They are strong baselines for seasonal context and weak substitutes for time series, extremes, forecasts or spatial risk assessments.
Use the WMO Normals Explorer to inspect the station values. Use the station charts and CSV guide when you need a repeatable selection or download procedure; this article concentrates on interpretation.
A fixed baseline cannot move with the climate
The current WMO climatological standard normals refer to 1 January 1991 to 31 December 2020. That makes them far more relevant than older baselines, but they are still fixed averages over a closed 30-year window.
The most recent standard normal and the long-term climate-change reference period serve different purposes. WMO recommends 1991–2020 for many operational applications because it better represents recently experienced conditions. For long-term climate-change assessment, WMO retains 1961–1990 as a fixed reference period, while global warming relative to pre-industrial conditions is commonly assessed against 1850–1900. WMO Climatological Normals, WMO explanation of reference periods.
If a station has warmed notably after 2020, the normal will not capture that newer shift. In practice, a place can feel warmer than normal very often not because the normal is wrong, but because the climate has continued changing since the baseline ended.
This is exactly why the WMO updated the definition of climatological standard normals in 2015 to use the most recent 30-year period ending in a year ending with 0, rather than waiting for older non-overlapping standards. Even so, every normal is still a backward-looking reference.
Averages hide change within the 30-year window
Normals compress three decades into a single monthly or annual value. That is ideal for building a baseline, but it hides the direction and shape of change inside the period.
For example, a station could warm steadily from the early 1990s to the late 2010s and still end up represented by one mean value. Likewise, a rainfall regime could become more erratic while the final precipitation normal changes only modestly.
What normals do well is summarize typical conditions. What they do less well is reveal whether the climate was stable, drifting, or shifting rapidly during the baseline itself.
Standard normals do not describe extremes
Monthly and annual standard normals summarize average conditions or totals. They do not, by themselves, describe the frequency, duration, timing or intensity of extreme events.
A monthly precipitation normal can tell you how wet a place usually is. It cannot tell you whether rain now falls in fewer but more intense bursts. A temperature normal can tell you the average thermal background. It cannot tell you how often a place experiences severe heat waves, tropical nights, or record-breaking winter warm spells.
This is why normals are best treated as a first layer of interpretation, not the final answer for hazard, impact, or event analysis.
Climate indices answer a different question
When the real question is about extreme events, climatological normals are often best paired with climate indices derived from daily observations.
These indices are designed to capture the kinds of changes that matter for impacts and risk. Examples include the annual hottest day, tropical nights, warm day and warm night frequency, maximum 1-day precipitation, maximum 5-day precipitation, and consecutive dry days.
In practice, climate indices help answer questions that normals cannot answer well on their own:
- are heavy rainfall events becoming more intense?
- are hot nights becoming more frequent?
- are dry spells becoming longer or more persistent?
- are threshold-based extremes changing faster than the mean climate?
That is the key distinction: normals describe the baseline, while climate indices quantify selected characteristics of extremes, including their frequency, intensity, duration or persistence.
A station normal is not a regional climate
In the WMO 1991–2020 collection displayed by Climate Explorer, each normal is associated with a reporting station rather than an entire landscape.
Stations sit at specific elevations, exposures, and land-use settings. A downtown observatory, a coastal airport, and a nearby hillside site can all show meaningfully different normals even when they are part of the same urban area.
That makes normals highly informative, but still local. Problems appear when users treat a point normal as a perfect stand-in for a whole city, valley, or mountain region.
Global coordination does not remove coverage differences
The WMO collection is globally coordinated, but the records are submitted by national meteorological and hydrological services rather than recalculated centrally inside Climate Explorer.
In practice, that means some important differences remain:
- parameter availability varies by station
- station density differs sharply by country
- local station history still affects interpretation
- comparisons work best when you match the same variable, month, and station context
So the collection is standardized in purpose and structure, but not every station has the same observational depth or representativeness.
Station histories also matter. Relocations, instrument changes, observing-practice changes and changes in the surrounding environment can introduce discontinuities. National services may apply quality control, homogenization or adjusted-normal methods, so small differences should be interpreted alongside the available station metadata.
Match the dataset to the question
The best way to use WMO normals is to treat them as a baseline layer, then add the evidence required by the question:
- start with the WMO Normals Explorer if you want to inspect the actual station normals for 1991–2020
- use normals to understand the typical climate of a station
- compare monthly values when seasonality matters
- pair normals with long station series when you care about trends
- use climate indices from daily data when you care about threshold exceedance, persistence, or event severity
- switch to daily or hourly archives when you care about extremes or recent volatility
Inside Climate Explorer, pair the baseline with GHCNm for long-term monthly temperature change or with a national daily or hourly archive when event-level detail matters. The June 2026 European capitals comparison demonstrates a narrower use: comparing observed point-temperature peaks with a monthly normal while keeping the two measures distinct.
The practical rule
The main problem with WMO normals is not that they are flawed. It is that they are often asked to do more than they were designed to do.
Used correctly, they are one of the clearest ways to describe a station’s baseline climate. Used alone, they can understate rapid recent change, hide extremes, and encourage overconfident regional comparisons.
That is why the most reliable reading of a WMO normal is as a starting point: a stable climate reference that becomes much more powerful when combined with station history, trend analysis, and higher-frequency observations.
References
Frequently Asked Questions (FAQ)
Are WMO normals outdated?
No. The 1991–2020 period remains the most recent WMO climatological standard normal for many operational applications. It is a fixed baseline rather than a live summary of post-2020 conditions. For long-term climate-change assessment, WMO retains 1961–1990 as a fixed reference period.
Can WMO normals show climate change by themselves?
Not very well. A single set of normals summarizes a period, but it does not reveal how the station changed within that period. Trend datasets and historical station series are better for that job.
Why can two nearby stations have different WMO normals?
Because station normals are local. Elevation, exposure, distance from the coast, urbanization, and instrument setting all influence the climate measured at a station.
What should I use if I care about extremes or recent weather shifts?
Use normals for baseline context, then move to historical daily, hourly, or monthly series depending on the question. Climate indices such as hottest-day, heavy-rainfall, warm-night, or consecutive-dry-day metrics are often more useful for extreme-event analysis than normals alone. In Climate Explorer, start with the WMO Normals Explorer for the actual 1991-2020 normals, then use tools like GHCNm and the national station archives for trend and event-level analysis.
Data Annex
| Metadata field | Dataset and interpretation details |
|---|---|
| Data product | WMO Climatological Standard Normals displayed in the Climate Explorer WMO Normals section |
| Reference period | 1 January 1991–31 December 2020 |
| Spatial unit | A reporting weather station; a station normal is not a regional average |
| Data stewardship | Values are coordinated through WMO and submitted by national meteorological and hydrological services |
| Primary use | Describing a station’s reference climate for operational comparison |
| Important limitation | A fixed 30-year summary does not show the trend within the period, post-2020 change, or the frequency and duration of extremes. |
| Reader question | What normals provide | What should be added |
|---|---|---|
| What was typical during 1991–2020? | A station-level monthly or annual reference value | Station metadata when exposure, elevation, or location affects comparison |
| How has climate changed through time? | A baseline for comparison, but not the trend itself | A long, quality-controlled station series with a documented trend method |
| Are extremes becoming more frequent or intense? | Mean conditions or totals, depending on the variable | Daily or sub-daily data and climate-extreme indices |
| Does one value represent a city or region? | Conditions at the reporting station | Nearby stations, gridded data, and an assessment of spatial representativeness |
Data Sources
- WMO Climatological Standard Normals: definitions, standard periods and long-term reference-period guidance.
- NOAA NCEI WMO Climate Normals: archive structure, primary parameters, coverage and submission documentation.