Statistical downscaling is a crucial component of the weather modeling field, where high-resolution outputs must be reconstructed from coarse-resolution inputs with the full cost of dynamical refinement. In this work, we investigate a hybrid quantum-classical corrective diffusion model for probabilistic statistical downscaling of weather fields. The proposed model inserts variational quantum circuit layers into the most compressed bottleneck of the diffusion UNet while leaving the regression branch fully classical. This placement tests whether quantum circuits can act as compact nonlinear feature maps for latent-channel mixing. We evaluate intra-channel and cross-channel ansätze on 10m wind components. On the 2020 validation set, the hybrid models remain stable, preserve the large-scale spatial organization of the generated wind fields, and improve both MAE and CRPS relative to a classical corrective diffusion model in several configurations. Structural diagnostics further show that the hybrid variants preserve kinetic-energy spectra and windspeed distributions similar to its classical counterpart while producing controlled changes in tail behavior, extreme-windspeed localization, and joint wind field components structure. Backend studies on the 2020 validation set show negligible impact from simulated device noise at the tested circuit scale, whereas real-hardware deployment remains limited by qubit availability and execution fidelity. The 2021 out-of-distribution test shows that these in-distribution gains do not transfer uniformly under temporal shift, revealing a generalization gap that motivates future mitigation through stabilization and regularization. These results show that bottleneck-level quantum hybridization can make a nontrivial contribution to weather statistical downscaling, while also highlighting that circuit scale and hardware deployment remain key limiting factors.
Statistical downscaling is a crucial component of the weather modeling field, where high-resolution outputs must be reconstructed from coarse-resolution inputs with the full cost of dynamical refinement. In this work, we investigate a hybrid quantum-classical corrective diffusion model for probabilistic statistical downscaling of weather fields. The proposed model inserts variational quantum circuit layers into the most compressed bottleneck of the diffusion UNet while leaving the regression branch fully classical. This placement tests whether quantum circuits can act as compact nonlinear feature maps for latent-channel mixing. We evaluate intra-channel and cross-channel ansätze on 10m wind components. On the 2020 validation set, the hybrid models remain stable, preserve the large-scale spatial organization of the generated wind fields, and improve both MAE and CRPS relative to a classical corrective diffusion model in several configurations. Structural diagnostics further show that the hybrid variants preserve kinetic-energy spectra and windspeed distributions similar to its classical counterpart while producing controlled changes in tail behavior, extreme-windspeed localization, and joint wind field components structure. Backend studies on the 2020 validation set show negligible impact from simulated device noise at the tested circuit scale, whereas real-hardware deployment remains limited by qubit availability and execution fidelity. The 2021 out-of-distribution test shows that these in-distribution gains do not transfer uniformly under temporal shift, revealing a generalization gap that motivates future mitigation through stabilization and regularization. These results show that bottleneck-level quantum hybridization can make a nontrivial contribution to weather statistical downscaling, while also highlighting that circuit scale and hardware deployment remain key limiting factors.
Weather forecasting plays a critical role across various sectors, supporting strategic planning and reducing the impact of hazardous climatic conditions. However, the inherently chaotic and nonlinear nature of atmospheric systems limits the performance of conventional forecasting approaches, often resulting in prediction inaccuracies and heightened risk. This study proposes a quantum-driven predictive framework that exploits the computational strengths of Quantum Machine Learning (QML) and hybrid quantum–classical optimization to enhance the accuracy and efficiency of modern forecasting models. The framework employs Quantum Neural Networks (QNNs) combined with Variational Quantum Circuits (VQCs) to learn complex spatial–temporal dynamics embedded within large meteorological datasets. Through quantum feature encoding, key atmospheric indicators—including temperature, humidity, wind velocity, and barometric pressure—are projected into high-dimensional Hilbert spaces, enabling more expressive pattern extraction and robust predictive behavior. Simulation results reveal that the quantum-augmented approach surpasses traditional deep learning architectures in both training convergence and forecast precision when applied to extensive weather records. The hybrid design further supports scalability by intelligently distributing computational workloads between classical processors and quantum hardware. Overall, this research demonstrates the transformative potential of quantum computing in atmospheric modeling, offering a foundation for future real-time forecasting systems capable of managing the increasing complexity of global climate patterns.
Abstract The major computational bottleneck of operational Numerical Weather Prediction (NWP) is data assimilation, in which a high-dimensional, prohibitively expensive variational cost function must be minimised repeatedly, with computational costs scaling alongside model resolution. This work presents a systematic empirical evaluation of an Adaptive Hybrid Variational Quantum-Classical (VQE)-based framework for Four-Dimensional Variational Data Assimilation (4DVAR), benchmarked against Classical BFGS optimisation and three deep learning baselines: a standard Long Short-Term Memory network (LSTM), a task-aligned Data Assimilation LSTM (DA-LSTM), and a Neural Ordinary Differential Equation (Neural ODE) across two canonical chaotic atmospheric models of increasing complexity: the three-variable Lorenz-1963 (L63) and the forty-variable Lorenz-1996 (L96) systems. All experiments were conducted using PennyLane quantum simulation across 50 independent runs per method, with rigorous statistical validation including paired t-tests, Wilcoxon signed-rank tests, bootstrap 95% confidence intervals, and formal effect size estimation. Results are reported with full timing decomposition separating quantum circuit execution (T 1 ), classical optimiser time (T 2 ), and infrastructure overhead (T 3 ). The central finding is that the Adaptive Hybrid Quantum-Classical model with a Hardware-Efficient Ansatz (Hybrid QC-HEA) achieves a mean 4DVAR cost of 2.10 on L63 and 21.0 on L96 representing only 11.7% and 14.8% above the Classical 4DVAR baseline respectively—differences that are not statistically significant at p < 0.05. Crucially, Hybrid QC-HEA outperforms all three deep learning baselines by substantial margins on L96, achieving 5.5x lower cost than DA-LSTM, 5.2x lower than LSTM, and 8.3x lower than Neural ODE. The Hybrid QC-HEA also exhibits the tightest cost distribution across 50 runs on L96, demonstrating superior operational reliability compared to all competing methods. Conversely, all-quantum VQE-based methods with no refinement by classical methods scale to cost 32 and 75x more than classical on L96, and thus gradient-free quantum optimisation without the classical BFGS refinement step is ultimately impractical to high-precision data assimilation and the classical BFGS refinement step is in any case a fundamental element of the architecture. Computational scaling analysis reveals that Classical 4DVAR wall-clock time grows 126-fold from L63 to L96, consistent with theoretical predictions of exponentially increasing complexity in chaotic optimisation landscapes, while the hybrid framework maintains near-classical accuracy at both scales. These results constitute an empirical proof-of-concept for hybrid quantum-classical data assimilation on realistic atmospheric models, establishing a reproducible benchmarking framework for future investigation on near-term quantum processors based NISQ devices.
Quantum diffusion models are rapidly advancing generative remote-sensing analytics by harnessing quantum computation to deliver higher-quality, more realistic satellite images, while converging faster and with substantially fewer trainable parameters than comparable classical diffusers. In this context, we first realise a fully-quantum latent diffusion model (QLDM) whose denoiser is a variational-quantum circuit acting on a 10-dimensionaI latent code. On EuroSAT, focusing on three land-cover classes Forest, Herbaceous Vegetation and SeaLake, QLDM lowers the Fréchet Inception Distance (FID) and Kernel Inception Distance (KID) by 21.5 % and 29.9 %, respectively, relative to a parameter-matched classical latent diffuser. Yet the severe spatial compression intrinsic to latent processing limits fine-grained detail. To restore image fidelity, we develop a hybrid quantum—classical architecture—the Quanvolutional Conditioned U-Net (QCU-Net)—which inserts entangling quantum layers both at the U-Net bottleneck and as a quanvolutional filter early in the encoder. Trained on the full ten-class EuroSAT RGB set, QCU-Net reaches an FID of 2.57 and a KID of 0.0008, representing 64 % and 76 % improvements over the best classical diffusion baseline (FID = 7.22; KID = 0.0034), and boosts class-conditioning accuracy to 81.7 % versus 62.2 % for the classical model. Thèse gains confirm that embedding quantum circuits within the feature-extraction pipeline yields richer spatial— spectral representations than latent-space quantum processing alone. Our results outline a clear progression in quantum generative modelling for Earth Observation: by first demonstrating with the QLDM that fully quantum denoising is not only feasible but also highly parameter- efficient, then showing how the hybrid QCU-Net architecture restores perceptual fidelity by embedding quantum feature extractors within a classical U-Net backbone, and finally pointing toward a fully pixel-space quantum diffusion model as future work—one that would leverage deep quantum circuits to perform both the forward and reverse diffusion directly in Hilbert space, remove classical auto-encoding steps, and promise even greater image quality and efficiency on fault-tolerant hardware.
Rare events such as financial crashes, climate extremes, and biological anomalies are notoriously difficult to model due to their scarcity and heavy-tailed distributions. Classical deep generative models often struggle to capture these rare occurrences, either collapsing low-probability modes or producing poorly calibrated uncertainty estimates. In this work, we propose the Quantum-Enhanced Generative Model (QEGM), a hybrid classical quantum framework that integrates deep latent-variable models with variational quantum circuits. The framework introduces two key innovations: (i) a hybrid loss function that jointly optimizes reconstruction fidelity and tail-aware likelihood, and (ii) quantum randomness–driven noise injection to enhance sample diversity and mitigate mode collapse. Training proceeds via a hybrid loop where classical parameters are updated through backpropagation while quantum parameters are optimized using parameter-shift gradients. We evaluate QEGM on synthetic Gaussian mixtures and real-world datasets spanning finance, climate, and protein structure. Results demonstrate that QEGM consistently reduces tail KL-divergence by up to 50% compared to state-of-the-art baselines (GAN, VAE, Diffusion), while improving rare-event recall and coverage calibration. These findings highlight the potential as a principled approach for rare-event prediction, offering robustness beyond what is achievable with purely classical methods.
We present a quantum-enhanced deep learning framework for short-term forecasting of extreme convective weather systems, specifically targeting the formation of bow echo structures through prediction clouds. Our approach integrates a quantum convolutional neural network (QCNN) into a classical convolutional neural network-long short-term memory (CNN-LSTM) pipeline, incorporating a 10-qubit variational quantum circuit executed on the lightning.qubit simulator. The model is trained using heterogeneous meteorological data (satellite images, convective available potential energy (CAPE) fields, lightning activity) from the extreme storm that hit the Corsican coast on August 18, 2022, preceding the convective event and evaluated during the critical intensification window (0700–0800 h). We compare this hybrid quantum model with classical CNN-LSTM and CNN-LSTM-Transformer architectures using three complementary metrics: Root mean squared error (RMSE), structural similarity index (SSIM), and Wasserstein distance. The quantum model demonstrates superior robustness to input noise and adversarial perturbations fast gradient sign method (FGSM), maintaining spatial coherence and stable prediction under degradation. These findings highlight the capacity of quantum circuits to encode more resilient representations for meteorological inference. This study provides a concrete application of hybrid QCNN architectures in the domain of extreme weather forecasting and lays the groundwork for future integration of topological data analysis to assess the preservation of topological features in quantum predictions.
Reliable estimation of tail risk and detection of low-prevalence anomalies remain challenging under distribution shift and limited simulation budgets. This paper introduces a statistical mechanics-inspired hybrid quantum generative model that learns an energy-shaped proposal for rare-event simulation while producing a calibrated anomaly score for deployment. Across simulation budgets from 200 to 6400 samples, relative error decreased from 0.31 to 0.08, outperforming normalizing flows (0.44 to 0.12), diffusion sampling (0.48 to 0.14), and a classical energy model (0.46 to 0.13). At a fixed evaluation budget, the proposed estimator achieved relative error 0.10 with effective sample size 1120 and tail hit rate 6.8%, compared with 0.16, 760, and 4.1% for flows. For anomaly detection, AUROC reached 0.962 and AUPRC reached 0.714, with precision 0.61 at 80% recall, exceeding flow scoring (AUROC 0.935, AUPRC 0.624, precision 0.49) and VAE reconstruction (AUROC 0.928, AUPRC 0.598, precision 0.46). Under temporal drift after a single calibration, realized false positive rate stayed within 0.0004 of a 0.5% target, whereas baselines drifted by up to 0.0084. Ablations showed peak AUPRC near moderate circuit depth and mid-scale qubits, while monotone annealing reduced rare-event error from 0.12 to 0.10 and improved stability. Under increasing noise, mitigation preserved AUPRC from 0.74 at low noise to 0.69 at moderate noise, supporting feasibility for NISQ-constrained monitoring and tail estimation.