What a NatHERS or HERS rating actually assumes about your climate file

Last month I sat with a couple in Newcomb, near Geelong, going through their NatHERS certificate line by line. They'd had a rough winter, condensation on the windows, a heating bill that made them wince, and a certificate on file claiming 6.8 stars. "How can it be nearly seven stars and feel like this," the husband asked. Fair question. The answer sits in a part of the rating almost nobody reads: the climate file.

Every NatHERS or HERS assessment runs a thermal simulation against a specific weather dataset tied to a location, not the actual weather the house experiences in any given year. Understanding what that climate file is, and isn't, changes how you read a star rating for good.

The climate file is an average year, not a real one

NatHERS assigns properties to one of 69 climate zones across Australia, each built from a reference weather year, a Typical Meteorological Year (TMY) style dataset assembled from decades of Bureau of Meteorology records for a representative station near that zone. The US HERS system, administered under RESNET's protocols, does something structurally similar with TMY3 data tied to weather stations and county-level assignments.

Neither system pulls last July's actual overnight lows or this January's heatwave. They pull a statistically constructed "typical" year, smoothed from historical patterns. That's a deliberate design choice, not an oversight, and it exists so a rating issued in a mild decade and one issued after three brutal summers stay comparable. But it also means the certificate is answering a different question than the one most owners think it's answering. It's not "how will this house perform this year." It's "how would this house perform against this location's long-run averaged conditions, assuming the rest of the model's occupancy defaults hold." We've covered those occupancy assumptions in more depth in what your NatHERS or HERS rating assumes about how you'll live in the house, and it's worth reading alongside this one because the two blind spots compound.

Why the nearest weather station matters more than people think

The climate zone your assessor selects is based on postcode and, in some cases, a nominated weather station rather than your exact street. A property on Bellerine Street in Geelong and one twenty minutes inland toward Bannockburn can fall in the same nominal zone, despite meaningfully different frost frequency and diurnal temperature swings. Coastal moderation, elevation, and local microclimate get flattened into a zone-level average.

This isn't a flaw unique to Australia. RESNET's methodology has the same tension in the US, particularly in states with large elevation changes, think a HERS rating in Denver versus one in a mountain town forty minutes away in a different microclimate entirely. The station is representative, not personal.

I'd argue this is one of the more defensible simplifications in the whole system, honestly, because building a bespoke climate file for every address would be computationally absurd and wouldn't meaningfully change design decisions for most homes. Where it does bite is in genuinely marginal locations, valley floors with unusual cold pooling, or ridge lines with wind exposure well above the zone norm.

What this means for comfort in an actual bad year

A star rating tells you how a design performs against averaged conditions. It says very little about how it copes with a genuine outlier, a week of unusually still, cold nights, or the kind of heatwave that's becoming less of an outlier each decade. We wrote about this gap directly in can you get a 7-star NatHERS rating and still freeze in July, and the climate file is a big part of why that happens. A high star rating doesn't guarantee comfort in extremes; it guarantees reasonable modelled performance across a typical spread of conditions.

For the Newcomb couple, the issue wasn't a bad rating exactly. It was a run of unusually cold, still nights that sat well outside the averaged reference year, combined with a heating system undersized for those outliers and gaps in the door seals that no simulation would have caught without a blower door test. The certificate wasn't wrong. It just wasn't answering the question they needed answered.

Where climate zones get updated, and why that changes existing ratings

Climate zone boundaries and reference years aren't permanent. NatHERS climate files get periodically reviewed as the Bureau of Meteorology's long-run records extend and as the national methodology committee revisits zone boundaries. When a zone gets reassigned or a reference file gets refreshed, houses rated under an older file can, in principle, return a different result if re-rated today, even with zero physical changes. That's a related but distinct issue from the assessor-variability question we detailed in why the same house can get two different NatHERS star ratings, and from the drift-over-time issue covered in why your NatHERS star rating can change without touching the house. Climate file revisions are one of the mechanisms behind that drift, alongside software version updates and default assumption changes.

In the US, RESNET periodically updates its accredited software providers' data feeds too, and a HERS Index Score calculated under an older data vintage isn't always perfectly comparable to one run today. If you're comparing two certificates from different years, check which software version and climate data vintage each one used before assuming they're apples to apples.

Orientation, shading and why the climate file compounds other blind spots

The climate file interacts badly with a couple of other known simplifications in these ratings. We've written before about how NatHERS and HERS ratings ignore your actual orientation in certain respects, defaulting to modelled solar exposure rather than confirmed site conditions. Stack an averaged climate file on top of an approximated orientation and approximated occupancy, and you get three layers of abstraction between the certificate and the lived experience of the house.

None of those abstractions are wrong individually. They exist because a national or state-wide rating scheme has to be standardised to mean anything comparable across postcodes. But a homeowner reading "6.8 stars" as a guarantee of comfort is asking the number to do a job it was never built for.

What to actually ask your assessor about the climate data

Most assessors won't volunteer this unless you ask, mostly because most clients don't ask. A few questions worth putting to whoever does your assessment: which climate zone and reference weather station has this been modelled against, and how far is that station from the actual site? Which software version and climate data vintage was used? And, more practically, given the local microclimate, are there known cold pockets, frost patterns, or exposure issues on this specific block that the zone-level file wouldn't capture?

None of that changes the certificate. But it tells you where the gap between the model and your actual winter is likely to be widest, which is far more useful than the star number on its own. For a broader look at what an assessor is actually doing on-site before any of this modelling happens, our piece on what a NatHERS or HERS assessor actually checks on-site is a good companion read.

If you're building new and choosing glazing or insulation levels based partly on your rating target, it's also worth reading double vs triple glazing: when the extra pane is actually worth it and air sealing before insulation, because both of those decisions matter more in the outlier weeks the climate file smooths over than in the average ones it models well.

The Bureau of Meteorology publishes the underlying climate data NatHERS zones draw from, and it's publicly available if you want to see how your zone's reference year compares to recent actual conditions at your nearest station. The US Department of Energy's Building America program has similarly published documentation on how TMY3 datasets get constructed, for anyone who wants to go deeper than the certificate itself ever will.

Reading the number for what it is

A star rating or HERS Index Score is a comparison tool, built on averaged, standardised inputs so that one house can be meaningfully weighed against another. It's genuinely useful for that. It's a poor tool for predicting how a specific week of genuinely severe weather will feel in a specific room. Both things are true at once, and conflating them is where most of the disappointment I hear from homeowners actually comes from.

Back in Newcomb, we ended up talking less about the certificate and more about door seals, a proper load calculation for their heating system, and what their specific block does in a southerly. That conversation should probably happen before the certificate gets filed away, not four winters after settlement.