Can quantum computers predict the weather?
No. As of August 2026 no quantum computer has ever run a weather forecast, and there is no evidence that current or near-term quantum hardware would help. The relevant algorithms exist mostly on paper with severe input/output caveats, and the atmosphere's chaotic dynamics limit every computer — quantum included.
Paper only — with severe caveats Quantum linear-system solvers (HHL and successors) promise exponential speedups, but only if loading the input state is cheap and you never need the full solution back. A forecast needs the full field back. Reading out a solution stored in amplitudes takes roughly as many measurements as there are numbers in it — the readout problem — which erases the exponential win. And the atmosphere's governing equations (Navier-Stokes) are nonlinear, so they must first be forced into linear form, inflating resource counts further. No end-to-end advantage has been shown even on paper for a realistic forecast workload.
Toy fluid solves exist — no speedup Researchers have solved miniature incompressible Navier-Stokes benchmarks (lid-driven cavity flows) on real noisy hardware using hybrid schemes. This is proof-of-principle plumbing, not capability: a laptop solves the same problems instantly, the grids are microscopic next to the ~10^9 grid points of a real forecast model, and no speedup exists at any size.
Quantum ML for weather explored — no advantage shown Met Office-affiliated researchers have studied quantum scientific machine learning for weather modelling: small parameterised quantum circuits fitted to simple atmospheric data. The work is careful and explicitly exploratory — it demonstrates feasibility of the approach at toy scale and makes no claim of advantage over the classical ML models that already run operationally.
No operational use — exploratory research only Forecasting centres such as ECMWF are watching the field: ECMWF's centre of excellence with Atos covers HPC, AI and quantum computing, and its researchers are candid that whether quantum computing can help at all is not yet clear. No credible organisation has put a date on operational quantum forecasting. Meanwhile classical machine-learning models have delivered the decade's real forecasting gains — without a single qubit.
Chaos limits any computer — quantum included Tiny errors in the measured initial state of the atmosphere grow exponentially (the Lorenz butterfly effect), which is why deterministic forecasts degrade after about two weeks regardless of compute power. A quantum computer ingests the same imperfect satellite and station observations as a classical one; it cannot out-compute sensitivity to initial conditions. More computing of any kind buys sharper probability estimates and modest horizon extensions, not certainty about week-three rain.
Why is weather such a bad fit for quantum computers?
Three separate walls, each fatal on its own today.
The physics mismatch. A quantum computer natively evolves linear quantum dynamics — amplitudes rotating and interfering. The atmosphere is a classical, turbulent, strongly nonlinear fluid. Forcing nonlinear equations into a linear quantum framework (Carleman linearisation, variational tricks) is possible on paper but multiplies the resource requirements, often past the point where the theoretical speedup survives.
Getting data in. A forecast model starts from roughly a billion numbers describing the current atmosphere. Loading classical data into quantum amplitudes is expensive, and the elegant speedup proofs usually assume this step is free. It is not.
Getting data out. A quantum solver does not hand you the answer; it hands you a quantum state whose amplitudes encode the answer. Measurement gives you one collapsed sample at a time. Extracting a full weather field requires on the order of as many shots as there are numbers in the field — which quietly erases the exponential speedup you came for. This is honest amplitude arithmetic, not pessimism; the measurement lesson covers why.
Wouldn't a quantum computer at least beat chaos?
No, because chaos is not a compute problem. The two-week forecast horizon comes from the atmosphere amplifying tiny errors in the starting state exponentially fast. Those errors come from imperfect observations — sparse stations, noisy satellites — and no processor, quantum or classical, can compute its way past not knowing today's atmosphere exactly.
What more compute genuinely buys is better probabilities: larger ensembles of perturbed forecasts, finer grids, better-resolved storms. That is exactly where the last decade's real revolution happened — classical machine-learning models now rival traditional simulation at a fraction of the cost. The honest comparison for quantum computing is not 'quantum vs today's forecasts' but 'quantum vs rapidly improving classical ML', and quantum is not on the board yet.
What would have to change for the answer to change?
First, fault-tolerant machines with enough logical qubits to run deep linear-algebra circuits — vendor roadmaps put the first such systems around 2029-2030, at scales still far below what PDE solving demands. Second, end-to-end algorithms whose advantage survives after counting data loading and readout, not just the middle of the computation. Neither exists today, even on paper, for a realistic forecast workload.
If those arrive, the plausible entry points are narrow: linear-algebra subroutines inside data assimilation (the step that blends observations into the model), or sampling for stochastic sub-grid physics. In every credible scenario quantum hardware accelerates pieces of the pipeline; 'a quantum computer predicting the weather' is not how any serious proposal reads. Track how actual devices are progressing on the QPU index, and see the quantum advantage assessment for how far demonstrated capability really extends.
Classifications follow the QPU137 editorial policy: every applied label carries a date, source, scale, and hardware, or it does not render. Found an error? Report it.