- Write out 1991 to 2020 when you cite a 30-year normal, because that is the span NCEI averages.
- Each value in the average is the same calendar month from a different year, so a January normal uses 30 January values.
- NCEI also publishes a 15-year normal for 2006 to 2020 and a 20th Century Baseline for 1901 to 2000, and each has its own span.
- The monthly figures are built on gridded data at about 4 km resolution over the contiguous United States.
A 30-year normal averages 30 monthly values from 1991 to 2020. For example, the January maximum temperature normal is the average of the 30 January values between 1991 and 2020, so name that span whenever you quote one.
Net takeaways
What the 30-year normal measures
The title phrase describes a specific average, and NCEI defines it in one sentence: The 30-year normal is an average of months from 1991 to 2020. Two parts of that sentence carry the meaning. The first is the span, which is fixed at 1991 to 2020. The second is the unit being averaged, which is months. A month here means a calendar month such as January or July, so each normal is tied to one calendar month rather than to a loose period like a work week or a rainy spell.
That distinction matters when a plan names a season. The quote describes the normal as an average of months, and its example is a single month, so the monthly figure is the unit to cite. If a dispatch plan asks for a winter or spring figure, combining monthly values is a separate calculation. The quote does not describe a seasonal product, so anyone who builds a seasonal figure should label it as their own combination of monthly normals.
The word "normal" also needs care. In everyday speech it can mean typical or expected. In this usage it is a statistical average over a defined window of years. It describes what the average looked like across those 30 years. It is not a forecast for a particular day, and it is not a description of one year's weather. A dispatcher who reads a normal as the expected temperature for a specific job day will overstate what the number can support.
Because the span is fixed, the 30-year normal works as a stable reference. Two people who cite the 1991 to 2020 normal for the same month are pointing to the same underlying average, which makes the span the most important detail to record in a quote. A phrase like "the normal" with no years leaves the reader guessing which period was used.
Why NCEI keeps these records
The organization that publishes these averages has a statutory duty behind it. NCEI describes the duty this way: "NCEI is also responsible for fulfilling the congressional mandate ... to establish and record the climatic conditions of the United States." The NCEI page traces that mandate to a provision of the Organic Act of October 1, 1890, which it says established the Weather Bureau as a civilian agency.
| Item | Detail from NCEI |
|---|---|
| Congressional mandate | To establish and record the climatic conditions of the United States |
| Act the mandate stems from | Organic Act of October 1, 1890 |
| Agency the provision established | The Weather Bureau, as a civilian agency |
The normals are one product of this recordkeeping duty. The mandate is to establish and record climatic conditions, and a 30-year normal is one way of recording them: it condenses many years of observations into a fixed average for each month. That is why the normal has a documented origin and a stated span rather than being an informal rule of thumb. When you cite it, you are citing an output of an institutional record-keeping function, not a private estimate.
The 1890 origin named in the quote explains why NCEI holds the country's climate records at all. The normals rest on that archive, so the span you cite is part of a documented, long-running record. The act date is background for the reader. The figure you quote is the average for 1991 to 2020.
Worked example: the January maximum
The clearest way to see the rule is the example NCEI gives: the 30-year January maximum temperature normal is the average of the 30 January values between 1991 and 2020. Each of the 30 values is the January maximum temperature for one year in that window, so the normal is a single number built from thirty separate January records.
| Quantity | Value in NCEI's example |
|---|---|
| Period | 1991 to 2020 |
| Values averaged | 30 January values |
| Variable | Maximum temperature |
| Calendar month | January |
Walk through the arithmetic in plain steps. Take the January maximum from 1991, the January maximum from the next year, and so on through 2020. Add the 30 values together and divide by 30. The result is the January normal. The calendar month stays fixed while the year changes, so a February value never enters the January average, even though February falls in the same winter.
The example also shows what the name claims. The 30 counts the years averaged, and the variable is a single weather element, here the maximum temperature. A January minimum temperature normal would be a different average built from the same 30 Januaries. When a dispatch plan names a January figure, it should say which variable it means, because the same span can support several different normals.
The procedure is also repeatable. Anyone with the 30 January values can reproduce the normal and check it against the period you named. That is why the example works well in a quote: the rule is short enough to state in one sentence, and the arithmetic can be checked from the named years.
The gridded data behind the numbers
The monthly normals sit on a gridded dataset.
| Grid property | Stated value |
|---|---|
| Grid type | nClimGrid monthly grids |
| Spatial resolution | 1/24 degree latitude/longitude |
| Approximate cell size | About 4 km |
| Coverage | Contiguous United States |
| Input data | NCEI station temperature and precipitation values |
Two terms in that quote need plain definitions. Interpolation estimates values at places where no station sits, using values from nearby stations. In general usage, climatologically aided means the estimate also draws on the typical climate pattern of an area, not only on nearby readings. Together, these steps turn scattered station records into a continuous map.
The resolution affects how you describe a location. A normal from this grid is a value for a cell of about 4 km, not a reading at a particular airport or town. If a dispatch site sits inside a cell, the cell's value is the closest published figure for that site. If the plan needs a value at an exact coordinate, state in the plan that the figure comes from the gridded product, so the reader knows it is a cell average.
The coverage is the contiguous United States. In practice, that means the grid is defined for the contiguous United States, and the quoted description does not extend it to Alaska or Hawaii. A dispatch plan for a job in either state needs its own reference, and naming the grid as the source would be wrong.
Comparing the three baselines
NCEI publishes more than one reference period, and the differences matter when you name one. The 30-year normal runs from 1991 to 2020, as NCEI's definition states. The same page also lists 15-year normals for 2006 to 2020, which it describes as a more recent period of reference, and a 20th Century Baseline of averages from 1901 to 2000, which it produced for longer term climate change study comparisons.
| Baseline | Period | Purpose as stated by NCEI |
|---|---|---|
| 30-year normal | 1991 to 2020 | Average of months across the period |
| 15-year normal | 2006 to 2020 | A more recent period of reference |
| 20th Century Baseline | 1901 to 2000 | Longer term climate change study comparisons |
Read across the rows. The 15-year normal is the shortest window, and it ends in the same year as the 30-year normal, so it sits inside the later part of the 1991 to 2020 span. The 20th Century Baseline is the longest, and its span ends in 2000, so it overlaps only the first decade of the 30-year normal's window.
Each baseline answers a different question. The 30-year normal gives the average for the 1991 to 2020 period. The 15-year normal gives a window weighted toward recent years. The 1901 to 2000 baseline gives a long span for comparisons across a century. The column for purpose is the one that decides which row fits a given plan.
Which one should you pick?
The choice depends on what the figure is for, not on which one sounds most current. Each line below names a situation and the period that matches it.
If you are quoting a typical month for a plan, pick the 1991 to 2020 normal because NCEI defines the 30-year normal as an average of months from 1991 to 2020.
If you are comparing against a more recent reference, pick the 2006 to 2020 normal because NCEI describes it as a more recent period of reference.
If you are writing a long-term climate change comparison, pick the 1901 to 2000 baseline because NCEI produced it for longer term climate change study comparisons.
When a plan says only "the normal," write out 1991 to 2020 so the reader does not have to infer which product you meant. The 30-year normal is the named product, and its span is the one the rest of this post has been about.
The 15-year normal fits when recency is the reason for the choice. NCEI calls it a more recent period of reference, so it suits a comparison where the 1991 to 2020 span would bring in older years you want to leave out. Be clear in the plan that the window is shorter, because it averages fewer values for each month.
The 1901 to 2000 baseline answers a different kind of question. Use it only when the comparison itself spans a century. Do not place its figures in the same table as 1991 to 2020 values without labeling each column with its span.
When a 15-year normal fits
NCEI's page says that there are also 15-year normals for 2006–2020 that provide a more recent period of reference. The phrase "more recent period of reference" explains why this product exists. It gives a reference that leans toward the latest years in the record.
A quote that uses the 15-year normal needs its span as firmly as one that uses the 30-year normal. A reader who sees "the 15-year normal" with no years may assume it means the usual 30-year figure, and the two can differ. Write the span in the same sentence as the product name, for example: "the 15-year January maximum temperature normal, 2006 to 2020." That keeps the window visible to anyone who copies the line.
Keep the 15-year figure separate from the 30-year figure in any plan. The two periods overlap, since 2006 to 2020 falls inside 1991 to 2020, but they are different averages. If you place them side by side without labels, a reader may treat the gap between them as a change in the weather when it only reflects the choice of window.
Use the shorter window when recency is the reason you are citing the period at all. If the job only needs a typical month, the 30-year normal is the product NCEI defines as an average of months from 1991 to 2020.
When the 1901–2000 baseline fits
NCEI describes its third product as a 20th Century Baseline, a set of averages from 1901–2000. The page says it was produced "for longer term climate change study comparisons." The key word is comparison. The baseline lets one long period be set against another, and its own dates stop at 2000.
That end date is the detail most likely to be missed. Any year after 2000 falls outside the baseline's own span, so the baseline cannot describe those years by itself. A plan that compares recent conditions against it needs a separate reference for the years after 2000.
Label the product in the quote, not only the years. A figure tagged "1901 to 2000" and one tagged "1991 to 2020" can look like interchangeable entries in a list, but they come from different averages built for different purposes. Calling the source "the 20th Century Baseline" makes the choice visible to the reader.
Describe the baseline as what NCEI says it is: a set of averages across 1901 to 2000. Do not present it as a single season, a single month, or a forecast unless the source says so. The quote supports only the span and the comparison purpose, so a plan that stretches it further is claiming more than the source does.
Stating the period in a quote
The first defines the product: the 30-year normal is an average of months from 1991 to 2020. The second shows the unit in practice: the 30-year January maximum temperature normal is the average of the 30 January values between 1991 and 2020.
A quote built from these needs four pieces, which the table sets out in order:
| Piece | What it tells the reader | Wording from NCEI |
|---|---|---|
| Product | Which average | 30-year normal |
| Span | Which years | 1991 to 2020 |
| Variable | Which weather element | Maximum temperature |
| Calendar month | Which month's values | January |
Put all four in one line, such as "30-year January maximum temperature normal, 1991 to 2020." Each piece does a job. Without the variable, a reader cannot tell a maximum temperature normal from a minimum temperature normal built from the same 30 Januaries.
Avoid shortened forms. "The 30-year normal" drops the span, "January normal" drops the variable, and a phrase like "the 1991 normal" misdescribes a window that runs through 2020. Place the period directly after the figure so the two stay together if the sentence is quoted somewhere else.
Planning a dispatch with a named period
A dispatch plan that names a seasonal period needs more care than a plan that names a single month. The quoted NCEI definition describes the 30-year normal as an average of months from 1991 to 2020, and its example uses one month. A winter or spring figure built from several monthly normals is therefore a combination the plan makes itself, and the plan should label it that way.
List the fields the plan should carry for each named figure:
- The product name and span, such as the 30-year normal for 1991 to 2020, which NCEI's page defines.
- The calendar month or months used, so the reader can see which monthly values were combined.
- The variable, such as maximum temperature, since each variable is a separate average.
- A note on how the period was assembled, for example "average of the monthly normals for the named months" when the season is the plan's own combination.
If the plan uses the 15-year normal instead, write 2006 to 2020 in the same field and say that it is a shorter window. The NCEI page describes it as a more recent period of reference, so it is a different product from the 30-year figure. Do not swap one for the other in the same line of the plan without updating the span.
Read the plan back once after writing it. If each named figure carries its product, span, variable, and month, a reader can check it without asking which average was meant.
Your next step today: open your latest dispatch plan or quote that cites a climate figure, find every mention of "the normal," and rewrite each one with the product name, the full span, the variable and the calendar month in the same sentence.
Related on this site
For the plan-side paperwork, see Move a booked visit: 3 rules and 6 fields and Stairway and ladder access: 4 dispatch fields.
Related guides: 2 budget approaches to set before a crew's next year · Climate normals: 2 periods to compare for dispatch planning · Idle equipment costs: federal rule usually caps at 1 year.
Reader questions
Why does NCEI keep climate records at all? NCEI says it is responsible for fulfilling a congressional mandate "to establish and record the climatic conditions of the United States." The NCEI page traces that mandate to the Organic Act of October 1, 1890.
What does a 30-year normal average? It averages months from 1991 to 2020, according to NCEI's definition. The unit is the calendar month, not a single day.
How is a January normal built? It is the average of the 30 January values between 1991 and 2020. NCEI's example is the January maximum temperature, as shown on its page.
What grid are the monthly normals built on? NCEI describes nClimGrid monthly grids at a 1/24 degree latitude and longitude spatial resolution, about 4 km, covering the contiguous United States.
When would I use the 15-year normal? When recency is the reason for the reference. NCEI describes 15-year normals for 2006–2020 as a more recent period of reference.
What is the 20th Century Baseline for? NCEI produced it for longer term climate change study comparisons, as a set of averages from 1901–2000.
Does the gridded data cover Alaska or Hawaii? The quoted description covers the contiguous United States only, and it does not state coverage for Alaska or Hawaii. Check the NCEI page before assuming the grid applies there.