PSLE-SCI-REALITY-0551
Wait, what? A forecast of 25 mm is not a 25% chance of rain
A weather graphic shows a coloured patch over a region and one label: QPF: 25 mm. One learner reads the number as a probability: “There is a 25% chance of rain.” Another reads it as a rate: “Rain will be falling at 25 mm every hour.” A third treats it as an observation: “The rain gauge has already measured 25 mm.” The same number has produced three different stories because the readers have not yet identified what the number is designed to represent.
QPF means Quantitative Precipitation Forecast. In ordinary weather practice, it is a forecast amount of liquid-equivalent precipitation expected over a stated future period and spatial support. NOAA weather services describe QPF as the amount of liquid precipitation expected during defined periods; the exact product may be a point value, grid value, basin value or areal-average forecast. That makes the first learner habit simple but powerful: before using the number, identify its quantity, time window, space and status.
Quick Answer
No. “QPF = 25 mm” does not by itself mean a 25% probability of rain, a rain rate of 25 mm per hour, or 25 mm already observed. It means that the forecast product expects a precipitation accumulation of 25 mm for the stated forecast period and location or area definition. To judge the claim, check the valid period, whether the value is liquid equivalent, whether it is a point/grid/area quantity, when the forecast was issued, and whether you are looking at a deterministic QPF or a separate probabilistic product.
The Exact Learner Job This Volume Owns
This volume owns one real-world evidence-transfer job: how to read a QPF amount without turning an accumulation forecast into a probability, an instantaneous rate or an observation. It does not become the owner of cloud formation, the water cycle, rainfall mechanisms, graph reading, averages, units, uncertainty or weather-model physics. Those ideas already have their own learning homes. Here, we apply them to one communication object that a learner can meet on a weather map, app, news graphic or forecast table.
If you need the broader distinction between what was observed and what was inferred, route to How to Tell Observation, Inference, Prediction and Explanation Apart in PSLE Science. For a related weather quantity that is often mistaken for rainfall itself, see Reality Lab Vol.174 on precipitable water. For model-grid interpretation, see Reality Lab Vol.520.
The Umbrella-School Forecast: An Original Composite Case
Imagine a school science club preparing an outdoor observation session. A fictional forecast board shows the following information for the school’s district:
| Field | Displayed value |
|---|---|
| Product | 6-hour QPF |
| Valid period | 12:00–18:00 |
| Forecast accumulation | 25 mm |
| Probability of precipitation | 70% |
| Forecast issued | 07:00 |
These numbers can coexist without contradiction because they answer different questions. The 25 mm value concerns how much precipitation is forecast to accumulate if described by that QPF field over the valid period. The 70% value concerns a probability statement defined by the forecast provider. Neither value says rain must fall at a constant rate. Neither says the accumulation has already happened. And the two percentages and amounts cannot be swapped merely because both are printed on the same screen.
Observed, claimed and inferred
| Layer | What belongs there? |
|---|---|
| Observed on the graphic | The label QPF, the number 25 mm, the valid period, issue time and map location. |
| Claim made by the product | A forecast precipitation accumulation for that defined period and spatial support. |
| Possible inference | There may be substantial rainfall during the period, subject to forecast uncertainty and local variation. |
| Unsupported leap | “It will rain at 25 mm/h for six hours,” “there is a 25% chance of rain,” or “25 mm has already been measured.” |
The Four Locks: Quantity, Time, Space, Status
1. Quantity: accumulation is not rate
An accumulation is a total over an interval. A rate describes how quickly something is happening at a moment or over a shorter averaging period. If 25 mm accumulates during six hours, it does not follow that rain fell at 25 mm/h. The total might come from one intense burst, several showers, steady rain, or some other time pattern. Without a time-resolved record or forecast, the accumulation alone does not reveal the shape of the rainfall through the interval.
A useful check is to ask: “If I moved the word total in front of the number, would my interpretation change?” If the product means a total accumulation, then a sentence about an instantaneous rate has introduced evidence the product did not provide.
2. Time: every QPF belongs to a valid period
A number without its valid period is incomplete evidence. A 25 mm forecast for six hours and a 25 mm forecast for seven days describe very different rates of accumulation and very different practical situations. Forecast graphics often show start time, end time or phrases such as “Day 1,” “6-hour,” “24-hour” or “Days 1–3.” Read those labels before comparing values.
Also separate issue time from valid time. A forecast issued at 07:00 for 12:00–18:00 does not describe conditions at 07:00. This is the same evidence habit used whenever a scientific object contains more than one clock: ask which time says when the information was produced and which time says when the information applies.
3. Space: a map value may represent a point, cell, basin or area average
Rain can vary sharply over short distances, especially in convective weather. A forecast product therefore has a spatial definition. NOAA’s Weather Prediction Center describes some QPF products in terms of expected areal-average rainfall over a grid. Other forecast services may present point forecasts or basin forecasts. The learner’s job is not to memorize one universal grid size. It is to inspect the product definition and avoid pretending that one displayed value guarantees the same total at every drain, rooftop and rain gauge inside the coloured region.
4. Status: forecast is not observation
QPF is predictive. QPE—Quantitative Precipitation Estimate—is an estimate of precipitation that has already occurred over a stated period. A rain gauge is another observational source. These can later be compared to evaluate forecast performance, but they should not be collapsed into one category. A map labelled “forecast” is not made more observational merely because its colours look precise.
A Number Can Be Precise Without the Future Being Certain
A forecast may display 24.8 mm, 25 mm or a tightly bounded colour range. The number of digits is a property of how the product is represented; it is not a guarantee that nature will follow the forecast to the last digit. Forecast uncertainty comes from uncertainty in the atmospheric starting state, model representation, storm placement and development, local terrain and other factors. The right response is not “forecasts are useless.” The right response is to match the strength of the conclusion to the strength and type of the evidence.
This is healthy scepticism rather than cynicism. Healthy scepticism asks, “What exactly is being claimed, and what would make me update my conclusion?” It does not reject a forecast simply because it is uncertain. Science often provides useful decisions under uncertainty.
Worked Case 1: Same 25 mm, Different Rainfall Histories
Consider three fictional six-hour rainfall histories. Each totals 25 mm:
| Hour | Case A | Case B | Case C |
|---|---|---|---|
| 1 | 4 mm | 0 mm | 20 mm |
| 2 | 4 mm | 0 mm | 5 mm |
| 3 | 4 mm | 0 mm | 0 mm |
| 4 | 4 mm | 10 mm | 0 mm |
| 5 | 4 mm | 15 mm | 0 mm |
| 6 | 5 mm | 0 mm | 0 mm |
| Total | 25 mm | 25 mm | 25 mm |
The same accumulation does not identify the same timing pattern. If the forecast graphic gives only the accumulation, it cannot support the claim that rainfall will be steady, concentrated at the beginning, or concentrated at the end. A separate hourly forecast might provide additional evidence, but that is a different evidence object.
Worked Case 2: 25 mm of Snow?
A winter forecast map says QPF = 25 mm while the weather symbol shows snow. A learner concludes that 25 mm of snow depth is forecast. That conclusion is not secure. QPF commonly expresses liquid-equivalent precipitation. For frozen precipitation, the forecast amount may describe how much liquid water the precipitation would produce when melted, not the physical depth of the snow layer. Snow depth depends on additional properties such as snow-to-liquid ratio and compaction.
The evidence habit is transferable: the unit may be familiar while the physical meaning is not. Read the product definition before translating one quantity into another.
Worked Case 3: The Forecast Updated From 10 mm to 30 mm
A morning forecast gives 10 mm for tomorrow afternoon. An evening forecast gives 30 mm for the same period. A student says, “The first forecast was wrong.” But tomorrow has not happened yet. The evening forecast may incorporate newer observations and model guidance. The change is evidence that the forecast estimate changed; it is not yet evidence about which version will verify better.
To evaluate accuracy, wait until the valid period has passed and compare the relevant forecast with suitable observations or precipitation estimates. Keep revision history separate from outcome verification.
Worked Case 4: One Gauge Reports 42 mm
The regional QPF was 25 mm. A rain gauge later reports 42 mm. Does that prove the forecast was useless? Not automatically. First check whether the QPF value represented an areal average or a specific grid/point forecast, whether the gauge lies inside the same spatial unit, whether the time windows match, and whether the gauge record is complete and quality-controlled. A local thunderstorm can produce a higher point total than an areal-average forecast.
After alignment, the difference may indeed be evidence of forecast error at that location. But the comparison must be like with like before the conclusion is made.
Tempting Reasoning That Fails
- “25 mm means 25%.” A millimetre is an amount unit here, not a percent probability.
- “25 mm means 25 mm/h.” Accumulation needs a valid interval; rate needs time in the denominator.
- “The map is coloured, so the rain already fell.” Forecast maps can look observational. Read the product status.
- “Every place inside the colour receives exactly 25 mm.” Spatial support and local variability matter.
- “The forecast changed, so science failed.” Forecasts are updated as evidence changes. Revision is not the same as failure.
- “The number has one decimal place, so the outcome is known to one decimal place.” Display precision is not forecast certainty.
What Evidence Would Strengthen the Forecast Claim?
Suppose a headline says, “Expect 25 mm of rain in the district tomorrow afternoon.” Evidence becomes stronger when the forecast source clearly defines the valid window and spatial unit; several updated forecasts converge rather than swing widely; independent model guidance and observed atmospheric conditions support the same scenario; and later verification from gauges or quality-controlled precipitation estimates is broadly consistent with the forecast. None of those makes the future certain. They make the evidence chain clearer and the claim better bounded.
What Evidence Would Weaken It?
The claim weakens if the graphic is cropped so that the valid period is missing, if the value belongs to a different location, if a probabilistic exceedance map is mistaken for deterministic QPF, if an old forecast is presented as current, if a snow liquid-equivalent amount is reported as snow depth, or if a point observation is compared with an areal average without acknowledging the difference in spatial scale.
How Far Can the Conclusion Travel?
A QPF can support a statement about forecast precipitation accumulation for the place, period and product definition it actually covers. It cannot by itself tell you the exact minute rain will begin, the exact rate at every moment, the amount at every metre of ground, the probability of exceeding another threshold, or the final observed total. Each extra conclusion needs extra evidence.
That “travel distance” idea matters throughout Science. Strong reasoning does not merely ask whether evidence is relevant. It asks how far the evidence is allowed to carry the conclusion before we have crossed into invention.
PSLE-Style Transfer Case
A fictional weather service publishes the following forecast:
| Valid period | 18:00–00:00 |
| QPF | 18 mm |
| Probability of precipitation | 60% |
A learner writes: “Rain has a 18% chance of falling, and when it falls it will fall at 18 mm each hour.” Explain two errors.
Explained answer: First, the probability is separately stated as 60%; the 18 mm QPF is a forecast precipitation amount, not a percent chance. Second, 18 mm is the accumulation for the stated six-hour valid period, not an hourly rate. The distribution of rainfall within those six hours is not given by the accumulation alone.
Delayed Independent Return: Use Q-T-S-S
Come back later to a weather graphic you have not seen before and ask four questions without looking at this article:
- Quantity: amount, probability, rate, index or something else?
- Time: issue time, valid time and accumulation window?
- Space: point, grid, basin or area average?
- Status: observation, estimate, analysis or forecast?
If those four checks arrive before the conclusion, the learner has gained a transferable evidence habit rather than memorised a weather definition.
Explained Practice
- A map says “24-h QPF = 40 mm.” Has 40 mm already fallen? No. QPF is a forecast. Check the valid 24-hour period.
- Does 40 mm mean a 40% chance of rain? No. Probability and forecast amount are different quantities.
- Can a six-hour 30 mm QPF include one dry hour? Yes. An accumulation does not require a constant rate.
- Can a gauge later record 45 mm while an areal-average QPF was 30 mm? Yes. Point and area values can differ; then evaluate whether the difference is plausible and whether the time windows align.
- Why check issue time? To know which forecast version you are judging and whether it was available before the event.
- Why check liquid equivalent? Frozen precipitation depth is not automatically the same as its melted-water amount.
For Parents and Tutors: Teach the Noun Before the Number
When a learner sees a scientific number, resist asking for the calculation first. Ask, “Twenty-five millimetres of what?” Then ask, “Over what time?” and “For what place?” This forces the noun and evidence object to stay attached to the number. Many weak answers are not arithmetic failures. They are quantity-identification failures.
A useful teaching move is to place four cards on a table: amount, rate, probability, observation. Read a weather statement and ask the learner to point to the correct card before explaining. Then change only one detail—such as “QPF 20 mm” to “70% chance of at least 20 mm”—and ask what changed. The contrast makes the evidence job visible without teaching a rigid answer template.
Authoritative Sources
- NOAA California-Nevada River Forecast Center: Quantitative Precipitation Forecasts — explains QPF as total liquid precipitation expected during specific periods and distinguishes it from QPE.
- NOAA Weather Prediction Center: About Quantitative Precipitation Forecasts — describes 24-hour QPF products and expected areal-average rainfall.
- National Weather Service: Probabilistic QPF Explanation — distinguishes probability-of-precipitation and probabilistic exceedance products from ordinary QPF amounts.
- SEAB: 2026 PSLE Science syllabus — current assessment objectives include interpreting, analysing and evaluating scientific information and communicating reasoning.
The Quiet Return
The strongest habit here is not “remember what QPF stands for.” It is quieter: never let a number travel without its meaning. Keep the quantity, time, space and evidence status attached. Then the weather graphic stops being a coloured answer to memorise and becomes what science communication should be: a claim you can read carefully, test against its definition, and use without pretending it says more than it does.
