Construction · Rain
Rain forecasts for construction and site planning
Rain matters differently for deliveries, earthworks, finishes and pack-up. Put the work window beside the forecast before interpreting the daily total.
An example of a more useful rain check
Suppose the task runs from 9 am to noon, with an hour needed to pack up. Review the hourly forecast across that whole period, including the time after the task. Record the forecast issue time and the place. Then compare recent radar frames, looking at their timestamps rather than assuming the image represents this instant.
This example is a planning workflow, not a prediction that the work can proceed.
Rain chance, amount and radar answer different questions
Chance describes likelihood over a stated period. Millimetres describe an amount over a period. Radar reflectivity describes returned energy, displayed as echoes. These are not interchangeable measurements.
Read the rain-chance guide for how to compare provider labels. Small changes in where showers develop can produce very different rainfall between nearby sites.
What the hourly percentage on Atmosphr means
Atmosphr’s hourly rain chance comes from Open-Meteo. Its documentation defines it as the probability of more than 0.1 mm of precipitation in the preceding hour, based on ensemble weather models at about 27 km resolution. The Bureau’s daily forecast works differently: it gives chances of reaching stated amounts over the day, and treats 0.2 mm as the smallest measurable amount because that is what its rain gauges record.
So an hourly 40% on Atmosphr and a daily 40% from the Bureau are answers to different questions. Compare like with like, and read the period each number covers.
Showers and steady rain behave differently
The Bureau separates two broad types of rain. Showers and thunderstorms (convective rain) fall in isolated, variable bursts: one suburb can get a heavy fall while the next gets nothing, so forecast ranges are wide, for example 5 to 50 mm. This is typical of the warmer months over northern and eastern Australia. Widespread rain from layered cloud (stratiform rain) varies less from place to place, so ranges are narrower, for example 5 to 10 mm. This is typical of cold fronts and cloud bands over southern Australia in the cooler months.
On a showery day, expect the forecast to be least reliable for your exact site. Plan for the range, watch recent radar and check for severe thunderstorm warnings. On a day of steady frontal rain, the timing of the band’s arrival is usually the key planning question.
Wet-weather questions by task
The questions below turn a forecast into a site discussion. They are prompts, not thresholds; the site’s own plan, specifications and contract decide what happens.
| Task | Forecast question | On-site check |
|---|---|---|
| Earthworks and subgrade | How much rain is possible before and after the work, and how long is the dry spell afterwards? | Condition of material, drainage paths and access for plant |
| Deliveries and lifts | Does the rain window coincide with wind, gusts or storms? | Ground conditions for outriggers, visibility and the lift plan |
| Concrete and finishes | Is rain possible during placement or before curing starts? | Covers ready. See the concreting guide for specification limits. |
| Roofing and work at height | Will surfaces be wet during the work? | SafeWork NSW notes slips, trips and falls are more likely when wind combines with wet weather |
| Leaving the site | What is forecast overnight and over the weekend? | Pits and trenches covered, materials secured, runoff controls in place |
For pours, the concreting guide summarises the rain-protection clause in Queensland TMR’s concrete specification.
Recent radar shows history, not a rain forecast
Recent radar helps you see whether a band of showers is approaching. The Bureau lists its limits: drizzle can form below the beam and go undetected, the radar cannot see directly above itself, echoes weaken with distance, and aircraft, flocks of birds or smoke from fires can look like rain. Radar colours often do not match what rain gauges measure. Atmosphr shows RainViewer composite frames from the recent past, so check the timestamp on the latest frame before deciding anything.
For the record, write down what was observed on site and when. Use gauge data or the source named in your contract when evidence is required.
Why two apps can disagree about your site
The Bureau explains that its point forecast for a location comes from the 6 km by 6 km grid square containing it, while the rainfall shown as observed usually comes from a single automatic weather station, which may be many kilometres away. Other providers use different models, grids and update times, so two apps can show different rainfall for the same place and both be working as designed.
For site planning, pick one forecast source, note its issue time, and compare it with what the site actually experiences. A simple diary line might read: “Forecast checked 6:10 am (Open-Meteo via Atmosphr): showers likely from about 2 pm. Rain observed on site 1:40 pm to 3:15 pm; excavation paused at 1:45 pm by the supervisor; resumed 3:30 pm after inspection.” Keep forecasts and observations as separate entries.
Keep forecasts separate from the site record
Write down what was forecast and what was observed as separate entries. Include the work affected, observed times and the person making the entry. A model estimate or screenshot of a radar loop is not a certified historical rainfall record.
For contractual weather evidence, check the contract’s accepted source and recording requirements. The planning checklist helps organise a review; it does not determine entitlement to a delay claim.
Sources and further reading
South East Queensland rain event
Compare the Brisbane rain outlook with observations and reported impacts.