Most of us built for wet.
Dry is a different job.

If you have been running detached residential sites in eastern Australia for the last few years, every habit you have is a wet-weather habit. You programme around ground that stays unusable for days after the rain clears. You keep a pump on site. You know which of your trades will come back after a washout and which will fill the slot elsewhere. You have priced weather into build periods and argued about it in extension of time claims. That is a real skill set, built the expensive way.

It is also a skill set tuned to one set of conditions. When a dry cycle arrives, the constraints that stopped work do not disappear so much as change identity. Saturated ground is replaced by shrinking ground. Rain days are replaced by extreme heat days and total fire ban days. Water you were trying to get off the site becomes water you have to get onto it. The builders who get the benefit of a dry run are the ones who plan for the new constraints early, rather than the ones who assume the old ones simply stopped. This guide is deliberately conservative about what a dry cycle would deliver, because the honest answer is that nobody knows, and specific about what changes on site, because that part is knowable in advance.

Everything in this section is a forecast, attributed and dated, and forecast probabilities move. Read it as the position at the time it was published, not as a description of what happened.

• ENSO returned to neutral following the end of the 2025/26 La Nina.
• As at July 2026, the NOAA Climate Prediction Center held an El Nino Watch, with the probability of El Nino emerging assessed at 82 per cent by May to July 2026 and a 96 per cent probability of persisting through the northern hemisphere winter of 2026/27 (NOAA Climate Prediction Center, ENSO Diagnostic Discussion, 14 May 2026).
• The World Meteorological Organization noted in April 2026 that climate models were strongly aligned, with confidence in onset described as high (WMO, April 2026).
• The Bureau of Meteorology's long-range outlook showed rainfall likely to be below average across most of NSW, Victoria, south-eastern Queensland and parts of South Australia for June to August 2026 (Bureau of Meteorology, outlook issued April/May 2026).
• For contrast, the 2023/24 El Nino was brief and was followed almost immediately by La Nina. Persistence is not guaranteed by onset, and a higher assessed probability of persistence is still a probability.

Those numbers are quoted here so that the reasoning in the rest of the guide is traceable, not so that anyone plans against them. They were current when this page was written and they will have moved since. Before you make any programming decision that touches weather, check the current Bureau of Meteorology long-range outlook for your region rather than relying on figures in this or any other article.

One more limit, and it is the important one. A rainfall outlook is a statement about rainfall. It is not a forecast of build times, completions or housing supply, and it should never be presented as one. The connection between weather and what gets built is real, but it runs through trades, contracts, approvals and finance, and every one of those can move independently of the sky.

The following is inference, conditional on the dry actually arriving and persisting. It is an operator's reasoning about mechanisms, not a forecast of outcomes, and it should be read with the word “if” attached to every sentence.

If drier conditions persist, there are reasonable grounds to expect that build times may begin compressing. The mechanism is not mysterious: a substantial part of the build-time blowout since 2020 came from stages that physically cannot run in the wet, and from the resequencing cascade that follows when they slip. Remove some of the stoppages and the cascade has less to feed on. How much of the difference shows up, and how quickly, is not something a rainfall probability can tell you.

If the dry holds, the trade absorption effect created by flood recovery should also unwind, but progressively. Recovery and remediation work is committed well ahead and runs for months or years after the event that created it, so trade capacity in an affected region returns on its own timetable rather than the weather's. The practical version of this for a builder is covered in flood recovery and trade availability: plan on the capacity you can confirm, and let any easing be something you observe rather than something you assume. Both of those are conditional statements about mechanisms. Neither is a prediction of housing completions, and neither belongs in a programme as a date.

Reactive clay changes volume with its moisture content. It swells when it takes up water and shrinks when it loses it, and a prolonged dry drives that loss deep into the soil profile rather than just at the surface. That much is well understood. What makes it a site problem is that the drying is never uniform. Shaded ground dries slower than exposed ground. Ground under a slab or a driveway holds moisture that the open yard has already lost. Established trees pull water out of one side of a building envelope and not the other. The result is differential movement, which is what actually loads a footing system, rather than movement as such.

The reason this gets underestimated is straightforward: a dry site is a pleasure to work on. Trucks are not bogging, the excavation stands up, the trades turn up. Nothing about the experience of the site says the ground underneath it is on the move, where the wet-weather version of ground risk announces itself.

The practical discipline is to treat prolonged dry as a site condition the engineer should know about, in the same way that a saturated site would be. Site classification, footing design, founding depths, the treatment of nearby trees and the timing of the excavation relative to the pour are engineering questions on specific ground, and nothing here substitutes for the design for your site. What the builder contributes is the observation that conditions have changed, made early enough to matter, and the record of what the ground was like when the footings went in.

Hot-weather concreting is a water balance problem. Air temperature, low humidity, wind and direct sun all increase the rate at which moisture evaporates from the surface of fresh concrete. When that rate runs ahead of the rate at which bleed water reaches the surface, the concrete behaves differently from the concrete you placed in April: less working time, a compressed finishing window, and a curing regime doing real work rather than sitting in the specification. The response is timing and preparation. Pour early, agree a temperature at which the pour does not go ahead, have water and curing organised before the truck arrives, and give someone on site the authority to postpone. The specifics belong to your concreter, your supplier and the relevant standards.

Steel work runs into the same day from a different direction. Sections sitting in the sun become a handling and burn risk, tolerances behave differently, and the hot works themselves become the constraint. Grinding, cutting and welding are precisely the activities a total fire ban is written to restrict, so on a bad day the question is not how hot the steel is but whether the work is permitted at all.

Bushfire conditions operate at a different scale again. A serious fire day can shut a region rather than a site. Roads close, deliveries do not arrive, and trades who live locally are either on a fire ground or at home protecting their own property. Smoke reaches well beyond the fire itself and makes outdoor work untenable. This is not weather that stops a specific task; it removes a day from everyone at once, which is the same shape as a flood day and deserves the same place in a programme. Sites in bushfire-prone areas also carry construction requirements of their own, which are a design question separate from the operational risk described here.

This is the part that costs money, so it is worth stating plainly. A builder who prices a fixed-price contract on a dry assumption and then gets a wet quarter has no recourse. The weather did not agree to anything. The client did not agree to anything about the weather either, beyond whatever allowance the contract states. The optimism was a unilateral decision by the builder, and a unilateral decision is exactly the kind that comes back as an absorbed cost rather than a claim.

The machinery that deals with weather on a residential job is unchanged by a favourable outlook. The build period still carries an allowance for anticipated weather. The extension of time clause still defines which causes qualify and what notice is required. Weather inside the allowance is still the builder's to absorb. A dry outlook does not change any of those positions; all it can do is make it less likely that the allowance is consumed. The correct way to enjoy a good run is to finish inside your dates with the allowance unused, not to have sold the allowance away in advance.

The same discipline applies to the programme. Bringing weather-exposed stages forward because the odds look better is sound sequencing, and it costs nothing if the odds turn out to be wrong. Shortening the programme because the odds look better is a bet, and it is a bet with an asymmetric payoff, because the upside is a slightly earlier handover and the downside is a late job under a contract that has a rate attached to lateness.

The most reliable mistake available in a turning season is to assume the trade market turns with it. It does not, for a mechanical reason. When a region floods, trades move onto repair and remediation work because it is urgent, funded and immediately available, and that work is contracted months ahead and runs long after the water has gone. A builder whose own site was never flooded still cannot sequence trades who are committed somewhere else across the same region. When the weather improves, those commitments do not evaporate; they are worked through.

The same lag applies to the ordinary backlog. Every job that slipped through a wet period is still in the system and still consuming a slot, so the first effect of better weather is often that everybody catches up at once rather than that anybody has capacity. It returns progressively, and unevenly by trade and by region.

What this means for a programme is undramatic. Book on confirmed availability. Keep confirming it, because a confirmation given three months out is a statement of intent rather than a booking. Hold the same lead-time discipline you built during the wet, and treat any genuine easing as something you notice on your own jobs, in your own call logs, before you plan the next one around it. The mechanics of the cascade that follows a missed booking are the same in any weather, and they are covered in lead times and sequencing.

The site record is the one habit from the wet years that should not be relaxed. A daily site record written through a good season establishes what normal looked like on your jobs, which is the only baseline against which an abnormal period can later be described. It also captures the dry-cycle events that are easy to leave out because they are not rain: the total fire ban day, the pour postponed for heat, the smoke day, the week the site was on carted water.

The practical test of a record system is whether it gets written on the days when nothing is at stake. If keeping it depends on someone deciding it is worth the effort, it will be complete for the wet fortnight and empty either side of it, which is the pattern that makes a claim look assembled rather than observed. In VIABUILD the ViaSite daily diary is filled in on site as part of the supervisor's day, with conditions and photos filed against the job, and the schedule holds the dates those conditions moved. What any of it means stays the builder's judgement; the record simply exists for the period before anyone knew it would matter.

• Bureau of Meteorology long-range forecasts, the current rainfall and temperature outlooks by region. This is the source to check before any weather-sensitive programming decision, in preference to any figure quoted on this page.
• Bureau of Meteorology ENSO history, the record of past El Nino and La Nina events, useful mainly for how variable their duration and effects have been.
• NOAA Climate Prediction Center ENSO Diagnostic Discussion, the monthly update that carries the watch and advisory status and the current probabilities. Issued monthly, and the figures change with each issue.
• Your site's geotechnical report and footing design, for what a change in soil moisture conditions means on your particular ground. Site-specific engineering advice is the only answer to a site-specific question.