★ Dynamical-generative downscaling of climate model ensembles
Combining RCM dynamical downscaling with diffusion refinement lets large CMIP6 ensembles be downscaled cheaply, giving better uncertainty bounds than BCSD and smaller dynamical ensembles.
Downscaling & super-resolution
Statistical and generative downscaling of weather or climate fields to higher resolution, including bias-aware downscaling of climate projections.
Search and filter this topic Plain list on GitHub
★ Highlights
★ Residual Corrective Diffusion Modeling for Km-scale Atmospheric Downscaling
Two-step UNet plus diffusion residual model downscales 25 km reanalysis to 2 km over Taiwan, recovering spectra and sharpening fronts and typhoon structures.
All papers (282)
★ Dynamical-generative downscaling of climate model ensembles
Combining RCM dynamical downscaling with diffusion refinement lets large CMIP6 ensembles be downscaled cheaply, giving better uncertainty bounds than BCSD and smaller dynamical ensembles.
Machine learning emulation of precipitation from km-scale UK regional climate simulations using a diffusion model
A diffusion model (CPMGEM) emulates 2.2 km UK convection-permitting precipitation from 60 km GCM inputs, reproducing extremes up to about 100-year return times at much lower cost.
Enhancing Regional Climate Downscaling through Advances in Machine Learning
Review of machine learning for empirical climate downscaling, covering out-of-distribution generalization, extremes, and a proposed evaluation framework for downscaling algorithms.
Spatial Downscaling of Near-Surface Air Temperature Based on Deep Learning Cross-Attention Mechanism
SNCA-CLDASSD, a Swin cross-attention network with DEM input, downscales CLDAS air temperature from 0.05° to 0.01° and beats interpolation and super-resolution baselines in mountains.
Nanjing University of Information Science & Technology (NUIST)
Investigating transformer‐based models for spatial downscaling and correcting biases of near‐surface temperature and wind‐speed forecasts
SwinIR and Uformer transformers downscale and bias-correct ECMWF HRES 2 m temperature and 10 m wind to 0.01°, outperforming UNet and RCAN baselines.
spateGAN: Spatio‐Temporal Downscaling of Rainfall Fields Using a cGAN Approach
spateGAN, a conditional GAN, downscales coarse German radar rainfall from 32 km/1 h to 2 km/10 min, with FSS of 0.6 for intensities over 15 mm/h.
Leveraging Deep Learning for Climate Change Prediction Models : A Dive into Cutting-Edge Techniques
Proposes deep ensembles to improve uncertainty in deep-learning statistical downscaling of climate projections, tested with pseudo-reality experiments for non-stationary future climates.
Interpretable AI-Driven Discovery of Terrain-Precipitation Relationships for Enhanced Climate Insights
A genetic-algorithm plus geographically weighted regression framework finds explicit terrain-precipitation equations that beat empirical models and support fine-scale precipitation downscaling under future climate.
★ Residual Corrective Diffusion Modeling for Km-scale Atmospheric Downscaling
Two-step UNet plus diffusion residual model downscales 25 km reanalysis to 2 km over Taiwan, recovering spectra and sharpening fronts and typhoon structures.
Enhancing wind field resolution in complex terrain through a knowledge-driven machine learning approach
ESRGAN-inspired network upscales coarse wind fields to high-resolution 3D flow over complex terrain at a Norwegian wind farm, outperforming trilinear interpolation.
Predicting rainfall response to climate change and uncertainty analysis: Introducing a novel downscaling CMIP6 models technique based on the stacking ensemble machine learning
Proposes a stacking ensemble of machine learning algorithms to statistically downscale three CMIP6 GCMs for rainfall at two Iranian basins under SSP scenarios.
Statistical Downscaling of SEVIRI Land Surface Temperature to WRF Near-Surface Air Temperature Using a Deep Learning Model
A 3D-convolution U-Net downscales SEVIRI land surface temperature to hourly 1 km WRF air temperature, with MAE 1.36 °C versus 1.49 °C for ridge regression.
Downscaling Daily Satellite-Based Precipitation Estimates Using MODIS Cloud Optical and Microphysical Properties in Machine-Learning Models
Random forest downscales daily IMERG precipitation from 10 km to 1 km using MODIS cloud properties, cutting RMSE by 29% against gauges.
A high-precision prediction method for coarse grids based on deep learning and the Weather Research and Forecasting model
Deep learning method improves WRF predictions on coarse grids; the abstract is unavailable, so the approach is inferred from the title only.
Probabilistic Rainfall Downscaling: Joint Generalized Neural Models with Censored Spatial Gaussian Copula
Joint generalised neural networks combined with a censored latent Gaussian copula produce spatially coherent probabilistic rainfall downscaling from coarse weather variables over the UK.
IMPROVING THE ACCURACY OF SATELLITE-BASED NEAR SURFACE AIR TEMPERATURE AND PRECIPITATION PRODUCTS
Random Forest spatial interpolation with satellite covariates downscales reanalysis and satellite air temperature and precipitation products over Turkey, with AgERA5 and MSWEP the best products.
Deep-learning statistical downscaling of precipitation in the middle reaches of the Yellow River: A Residual in Residual Dense Block based network
RRDBNet residual-in-residual dense network statistically downscales Yellow River precipitation and improves extreme precipitation over regression and deep baselines.
On the use of convolutional neural networks for downscaling daily temperatures over southern South America in a climate change scenario
Applies CNN statistical downscaling to daily temperatures over southern South America under climate change scenarios.
Multi-variable Hard Physical Constraints for Climate Model Downscaling
Introduces multi-variable hard constraints so deep-learning downscaled climate variables keep physical relationships, demonstrated on temperature downscaling from GCM output.
Downscaling CORDEX Through Deep Learning to Daily 1 km Multivariate Ensemble in Complex Terrain
Deep-learning perfect-prognosis downscaling of CORDEX-EUR11 yields a daily 1 km multivariate ensemble for Saxony through 2100 with validated transfer functions.
Downscaling ERA5 wind speed data: a machine learning approach considering topographic influences
Machine learning downscales ERA5 wind speed from about 31 km to 1 km in complex terrain using station data and topographic metrics, aimed at energy applications.
Comparison of machine learning statistical downscaling and regional climate models for temperature, precipitation, wind speed, humidity and radiation over Europe under present conditions
Compares a machine-learning statistical downscaling model with six EURO-CORDEX regional climate models over Europe; the ML model scores better on most metrics before bias correction.
The skill of statistical downscaling in future climate with high-resolution climate models as pseudo-reality
Evaluates 21 statistical downscaling methods, including ML, against a 4 km pseudo-reality for Belgian extreme rainfall, recommending an ensemble of downscaling approaches.
Uncertainty Quantification of Deep Learning–Based Statistical Downscaling of Climatic Parameters
Downscales CanESM5 temperature and precipitation at two Iranian stations with feedforward and LSTM networks, adding LUBE prediction intervals; LSTM gains about 55% higher NSE.
Bivariate DeepKriging for Large-scale Spatial Interpolation of Wind Fields
Bivariate DeepKriging uses a spatially embedded DNN to interpolate wind fields at about 500,000 locations, beating cokriging and running roughly 20 times faster.
Comparison of a novel machine learning approach with dynamical downscaling for Australian precipitation
An MLP with a novel loss downscales Australian precipitation from 12–48 km to 5 km, outperforming dynamical downscaling on extremes with 15–85% lower bias at 12 km.
Efficient Super‐Resolution of Near‐Surface Climate Modeling Using the Fourier Neural Operator
Applies a Fourier neural operator with a Clausius–Clapeyron constraint for zero-shot super-resolution of WRF near-surface heat fields over the Great Lakes, reaching 1-km resolution.
An improved deep learning procedure for statistical downscaling of climate data
Applies a CNN to downscale CMIP6 CanESM5 temperatures over Egypt to 0.1°, projecting warming of 4.8 to 6.3 °C by 2100 under SSP scenarios.
Downscaling Seasonal Weather Forecasting with Generative Adversarial Networks
A GAN combined with a digital elevation model downscales seasonal temperature forecasts from 1° to 0.1° over Greece, with some over- and underestimation.
Statistical downscaling and projection of climatic extremes using machine learning algorithms
Statistically downscales and projects climatic extremes using machine learning algorithms; abstract unavailable, so scope is inferred from the title.
Evaluating Loss Functions and Learning Data Pre-Processing for Climate Downscaling Deep Learning Models
L2 loss outperforms L1 for imbalanced precipitation downscaling, and a learned nonlinear pre-processing function further improves accuracy in climate downscaling networks.
General Circulation Model Downscaling Using Interpolation—Machine Learning Model Combination—Case Study: Thailand
IDW-ANN combines interpolation with a neural network to downscale IPSL GCM temperature and precipitation to 1 km over Thailand under SSP scenarios.
Downscaling of GPM satellite precipitation products based on machine learning method in complex terrain and limited observation area
Downscales GPM satellite precipitation from 0.1° to 1 km using a BP neural network with terrain and vegetation predictors, improving accuracy in complex terrain with sparse gauges.
Surrogate Downscaling of Mesoscale Wind Fields Using Ensemble Superresolution Convolutional Neural Networks
A 20-member SRCNN ensemble downscales 5 km surface winds to 1 km around Tokyo, with the ensemble mean beating the input winds all year.
Generating high-resolution climate maps from sparse and irregular observations using a novel hybrid RBF network
A hybrid RBF network downscales sparse station data into high-resolution precipitation, temperature and humidity maps for Bangladesh, outperforming MLR and kriging.
Downscaling daily wind speed with Bayesian deep learning for climate monitoring
Bayesian deep learning downscales daily wind speed for climate monitoring; the abstract is missing, so the scope is inferred from the title.
Assessment of Drought Disasters (EDI) Based on ENSO and NOAA Climate Data Using ANN in Bondowoso Regency
Backpropagation ANN statistically downscales ENSO and NOAA indices to monthly rainfall in Bondowoso, Indonesia, to drive EDI drought assessment with R2 of 0.97.
DownScaleBench for developing and applying a deep learning based urban climate downscaling- first results for high-resolution urban precipitation climatology over Austin, Texas
DownScaleBench uses an iterative super-resolution CNN to produce 300 m urban precipitation from 10 km GsMAP data over Austin, improving on cubic interpolation.
An Intercomparison of Deep-Learning Methods for Super-Resolution Bias-Correction (SRBC) of Indian Summer Monsoon Rainfall (ISMR) Using CORDEX-SA Simulations
Intercomparison of deep-learning super-resolution bias-correction methods for Indian summer monsoon rainfall in CORDEX South Asia simulations; the abstract is unavailable, so the classification relies on the title.
Fourier Neural Operators for Arbitrary Resolution Climate Data Downscaling
A Fourier neural operator downscales ERA5 climate fields zero-shot to arbitrary unseen resolutions, outperforming CNN and GAN downscaling baselines.
Climate change effects on sub-daily precipitation in Spain
Machine-learning downscaling of CORDEX projections to hourly rainfall over Spain projects intensified sub-daily extremes and 100-year events, increasing flash flood risk.
Downscaling wind speed based on coupled environmental factors and machine learning
Neural-network downscaling of ERA5 daily wind speed to a 1 km China dataset for 2015, with R2 of 0.95 against station observations.
Towards an Accurate and Reliable Downscaling Scheme for High-Spatial-Resolution Precipitation Data
Compares ML downscaling of satellite precipitation, finding SRF and XGBoost best and GDA calibration beneficial, while residual Kriging degrades performance.
Climate downscaling for regional models with a neural network: A Hawaiian example
Applies a neural network to downscale climate for a regional Hawaiian model; the abstract is missing, so the target variables and performance cannot be verified.
Deep Learning for Downscaling Tropical Cyclone Rainfall to Hazard‐Relevant Spatial Scales
A Wasserstein GAN downscales tropical cyclone rainfall from about 100 km to 10 km, reproducing spatial structure and spectra with mean bias within 5% and extrapolating to extreme storms.
Assessing Three Perfect Prognosis Methods for Statistical Downscaling of Climate Change Precipitation Scenarios
Compares GLMs, CNNs and a posteriori random forests for perfect-prognosis precipitation downscaling of EC-Earth; APRFs best reproduce the GCM's future climate change signal.
Unpaired Downscaling of Fluid Flows with Diffusion Bridges
Unpaired diffusion bridges downscale geophysical fluid simulations without paired data, correcting biases including in extremes and promising use for climate precipitation and temperature.
Deep Ensembles to Improve Uncertainty Quantification of Statistical Downscaling Models under Climate Change Conditions
Deep ensembles improve uncertainty quantification of statistical downscaling under climate change, tested in a pseudo-reality experiment.
On the use of Deep Generative Models for Perfect Prognosis Climate Downscaling
Deep generative models for perfect-prognosis precipitation downscaling improve spatial consistency of high-resolution fields compared with independent per-location estimates.
Improving trajectory calculations by FLEXPART 10.4+ using single-image super-resolution
EDSR super-resolution of ERA5 winds halves interpolation RMSE and cuts FLEXPART trajectory errors by at least 49.5% after 48 hours.
Physics-informed deep learning framework to model intense precipitation events at super resolution
A physics-informed recurrent deep learning model super-resolves regional climate precipitation from 2.5 km to 250 m, producing realistic intense short-duration events over Montreal.
Downscaling long lead time daily rainfall ensemble forecasts through deep learning
Deep-learning downscaling of ACCESS-S1 seasonal rainfall ensembles to 12 km, trained on CRPS, beats quantile mapping for lead times up to 217 days.
Spatiotemporal Super-Resolution with Generative Machine Learning for Creating Renewable Energy Resource Data Under Climate Change Scenarios
Generative ML downscales daily CMIP GCM output to hourly 4 km weather for wind, solar, temperature and humidity, enabling grid studies under climate change.
A Comparison of the Statistical Downscaling and Long-Short-Term-Memory Artificial Neural Network Models for Long-Term Temperature and Precipitations Forecasting
LSTM outperforms statistical downscaling (SDSM) for monthly temperature and precipitation at five Lake Chad Basin stations using coarse GCM output.
Improving risk reduction potential of weather index insurance by spatially downscaling gridded climate data - a machine learning approach
ML downscaling of gridded precipitation, temperature and soil moisture improves weather index insurance hedging in 70% of Kazakh and Mongolian county cases.
Statistical Downscaling in the Tropics and Midlatitudes: A Comparative Assessment over Two Representative Regions
Compares statistical downscaling methods, including XGBoost, for temperature and precipitation in Spain and Central America, finding model output statistics best for temperature.
Super-Resolution of Three-Dimensional Temperature and Velocity for Building-Resolving Urban Micrometeorology Using Physics-Guided Convolutional Neural Networks with Image Inpainting Techniques
Physics-guided gated-convolution CNN super-resolves 3D urban air temperature and wind from coarse building-resolving simulations, inferring near-surface flow from flow above buildings.
Projection of climate variables by general circulation and deep learning model for Lahore, Pakistan
Projects future local climate variables for Lahore, Pakistan by combining GCM output with a deep learning model. No abstract available; scope inferred from title.
Machine learning-assisted mapping of city-scale air temperature: Using sparse meteorological data for urban climate modeling and adaptation
Uses machine learning to map city-scale air temperature from sparse meteorological station data for urban climate adaptation. No abstract available; scope inferred from title.
Deep learning models for generation of precipitation maps based on numerical weather prediction
Conditional GAN and U-Net generate high-resolution short-range precipitation maps from full NWP variables, reducing over-forecast bias; CGAN gives the most realistic output.
Deutscher Wetterdienst (DWD)Forschungszentrum Jülich / Jülich Supercomputing Centre
Using Machine Learning to Cut the Cost of Dynamical Downscaling
Machine learning emulators (MLP and error-corrected regression with random forest) downscale GCM evapotranspiration from 12.5 km to 1.5 km, reaching KGE of 0.86 and beating statistical downscaling.
Predictor Selection for CNN-based Statistical Downscaling of Monthly Precipitation
Studies predictor selection for CNN-based statistical downscaling of monthly precipitation, aiming to improve regional precipitation fields from large-scale predictors.
Institute of Atmospheric Physics, Chinese Academy of SciencesChina Meteorological Administration (CMA)
Algorithmic Hallucinations of Near-Surface Winds: Statistical Downscaling with Generative Adversarial Networks to Convection-Permitting Scales
GANs statistically downscale near-surface winds to convection-permitting WRF scales, with spectral analysis exposing spatial-structure biases and a partial frequency-separation loss.
Evaluation of near-surface air temperature reanalysis datasets and downscaling with machine learning based Random Forest method for complex terrain of Turkey
Evaluates near-surface air temperature reanalyses over Turkey and downscales them with a random forest for complex terrain; no abstract available.
SPATIAL DOWNSCALING OF GPM IMERG V06 GRIDDED PRECIPITATION USING MACHINE LEARNING ALGORITHMS
Downscales GPM IMERG precipitation from 10 km to 1 km over Taiwan using decision trees, SVR and random forests with NDVI and terrain predictors, validated against gauges.
Using Explainability to Inform Statistical Downscaling Based on Deep Learning Beyond Standard Validation Approaches
Intercompares deep learning perfect-prognosis temperature downscaling models for CORDEX North America and uses XAI to expose extrapolation behaviour beyond standard validation.
Customized deep learning for precipitation bias correction and downscaling
Customized deep learning with weighted losses, multitask learning and covariates improves hourly precipitation bias correction and downscaling, better capturing extremes than quantile mapping.
Prediction of rainfall response to the 21st-century climate change in Ghana using machine learning empirical statistical downscaling
Machine learning empirical statistical downscaling projects 21st-century rainfall change in Ghana under emission scenarios; the abstract is minimal.
Material for "A Comparison between Perfect-Prognosis and Super-Resolution Deep Learning for Statistical Downscaling of Climate Change Simulations"
Repository with code and material comparing perfect-prognosis and super-resolution deep learning for statistical downscaling of climate change simulations.
Estimating Site-Specific Wind Speeds Using Gridded Data: A Comparison of Multiple Machine Learning Models
A deep CNN estimates site-specific surface wind speed from gridded ECMWF ERA5 across eastern China, reducing RMSE 23–26% versus linear interpolation.
An evaluation of satellite precipitation downscaling models using machine learning algorithms in Hashtgerd Plain, Iran
Evaluates machine learning models for downscaling satellite precipitation over Hashtgerd Plain, Iran, checked against local observations.
Combining Long-Short Term Memory and Genetic Programming for Monthly Rainfall Downscaling in Southern Thailand's Thale Sap Songkhla River Basin
Combines LSTM and genetic programming to downscale CMIP5 and CMIP6 monthly rainfall at three Thai stations, projecting 2030s to 2080s trends.
Super-Resolution for Renewable Energy Resource Data with Climate Change Impacts (Sup3rCC)
Releases 4 km hourly wind, solar and meteorological fields for the contiguous US under climate scenarios, generatively downscaled from GCMs for energy studies.
Machine Learning-Based Downscaling of Urban Air Temperature Using Lidar Data
Machine learning downscaling of urban air temperature using lidar data; the abstract is missing, so the method and results are limited.
Offshore Wind Field Super-Resolution Using Deep Learning Methods
Applies deep learning to super-resolve offshore wind fields; the abstract is unavailable, so the method and gains cannot be verified.
Strictly Enforcing Invertibility and Conservation in CNN-Based Super Resolution for Scientific Datasets
Differentiable operator enforces exact conservation of low-resolution values in CNN super-resolution, tested on radar, satellite and climate model data.
A XGBoost-Based Downscaling-Calibration Scheme for Extreme Precipitation Events
XGBoost used for both downscaling and calibration of satellite precipitation during typhoons, better reproducing extreme rainfall spatial variability than single-step alternatives.
Downscaling the Midsummer Temperature-Humidity Index Based on Multiple Machine Learning Methods
Random Forest downscales the summer temperature-humidity index from 6 km to 1 km over Anhui, achieving correlations above 0.97 with site observations.
Statistical Downscaling of Temperature Distributions in Southwest China by Using Terrain-Guided Attention Network
Terrain-guided attention CNN downscales 2 m air temperature from 0.0625° to 0.01° over Southwest China, reaching 1.12 °C RMSE using DEM attention blocks.
Nanjing University of Information Science & Technology (NUIST)China Meteorological Administration (CMA)
A precipitation downscaling method using a super-resolution deconvolution neural network with step orography
Super-resolution deconvolution network with orography downscales hourly precipitation from 100 to 12.5 km over Australia, improving fine-scale extremes over DeepSD.
Debias Coarsely, Sample Conditionally: Statistical Downscaling through Optimal Transport and Probabilistic Diffusion Models
A two-stage statistical downscaling method combines optimal-transport debiasing with diffusion upsampling, enabling unpaired 8x and 16x super-resolution on fluid flows relevant to weather and climate.
Deep Learning Regional Climate Model Emulators: a comparison of two downscaling training frameworks
Deep learning emulator downscales a global climate model to regional Antarctic surface mass balance; the imperfect-model training framework outperforms the perfect-model one.
Emulating the Adaptation of Wind Fields to Complex Terrain with Deep Learning
DEVINE, a CNN trained on ARPS simulations, downscales AROME winds to 30 m over the Alps and cuts the mean wind-speed bias by 27% at 61 stations.
Machine Learning Downscaling of Regional Precipitation Patterns in Mountainous Terrain
CNN downscales 25 km regional climate precipitation to 1 km in mountain terrain, cutting RMSE 34% versus raw output and 12% versus quantile mapping.
Downscaling Extreme Rainfall Using Physical-Statistical Generative Adversarial Learning
A physics-statistical GAN downscales 0.25° climate model rainfall to 0.01° fields, quantifying uncertainty and matching observed extreme rainfall distributions.
Statistical treatment of convolutional neural network super-resolution of inland surface wind for subgrid-scale variability quantification
CNNs super-resolve surface wind from 25-100 km to 3 km, with coarse wind plus fine topography giving the best out-of-sample extreme-wind skill.
Machine learning emulation of a local-scale UK climate model
Diffusion models generate realistic high-resolution UK rainfall samples conditioned on low-resolution simulations, matching the high-resolution model's quantiles and time-mean.
Development of the Indian Future Weather File Generator Based on Representative Concentration Pathways
Generates future weather files for India from CanESM2 using spatial and temporal downscaling plus ML bias correction across three RCP scenarios and five future years.
Generating high-resolution climatological precipitation data using SinGAN
A single-image GAN with a weighted patchGAN discriminator generates high-resolution precipitation from one low-resolution sample, improving detail in extreme-value regions.
Machine learning-based downscaling: application of multi-gene genetic programming for downscaling daily temperature at Dogonbadan, Iran, under CMIP6 scenarios
Multi-gene genetic programming downscales daily temperature at Dogonbadan, Iran, under CMIP6 scenarios, producing local projections from global climate models.
Contrastive Learning for Climate Model Bias Correction and Super-Resolution
Contrastive-learning GAN performs super-resolution and bias correction of CMIP6 daily precipitation and temperature, reaching double NEX-GDDP resolution with comparable or better bias reduction.
High-resolution downscaling with interpretable deep learning: Rainfall extremes over New Zealand
A simple CNN with tailored loss functions outperforms regression for downscaling New Zealand rainfall, raising wet-day explained variance from 0.35 to 0.52.
A Multi-Scale Deep Learning Framework for Projecting Weather Extremes
Multi-resolution deep learning debiases GCM tail statistics against reanalysis and reconstructs finer scales, producing realistic Western European weather-extreme realizations for risk assessment.
Repeatable high-resolution statistical downscaling through deep learning
U-Net and U-Net++ deep learning perfect-prognosis downscaling produces daily 1 km precipitation for Saxony with improved skill and full reproducibility.
Downscaling of Historical Wind Fields over Switzerland Using Generative Adversarial Networks
GAN with gradient-penalized Wasserstein loss downscales ERA5 hourly winds over Switzerland from 25 km to 1.1 km, capturing orographic effects missed by ERA5.
Deep Learning-Based Downscaling of Temperatures for Monitoring Local Climate Change Using Global Climate Simulation Data
An overview of ML downscaling of GCM output, with a deep-learning case study producing local weekly temperature projections for a New Jersey watershed.
Deep generative model super-resolves spatially correlated multiregional climate data
Physics-informed SRGAN downscales climate fields up to 50x while preserving inter-regional spatial correlations of temperature and precipitation.
Downscaling of Precipitation for Climate Change Projections Using Multiple Machine Learning Techniques: Case Study of Shenzhen City, China
SVM and MLP downscale GCM output to daily precipitation at 10 Shenzhen stations, outperforming SDSM and projecting ensemble precipitation trends to 2100.
Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44
DeepESD, a CNN multi-model ensemble, downscales eight CMIP5 GCMs to CORDEX EUR-44, matching RCM climate change signals with smaller precipitation spread.
Physically Constrained Generative Adversarial Networks for Improving Precipitation Fields from Earth System Models
A physically constrained GAN corrects Earth system model precipitation, fixing local distributions and the double-peaked ITCZ while preserving global precipitation sums under future scenarios.
Potsdam Institute for Climate Impact Research (PIK)Technical University of Munich (TUM)
Urban precipitation downscaling using deep learning: a smart city application over Austin, Texas, USA
Iterative SRCNN downscales 10 km JAXA GsMAP satellite precipitation to 300 m over Austin, Texas, beating cubic interpolation on PSNR and mutual information.
Hard-Constrained Deep Learning for Climate Downscaling
Hard constraints enforce conservation laws in deep learning climate downscaling models, improving accuracy and physical consistency across architectures and datasets.
Regional climate model emulator based on deep learning: concept and first evaluation of a novel hybrid downscaling approach
A neural-network RCM emulator hybrid-downscales daily 12 km near-surface temperature from large-scale predictors, reproducing RCM variability with training in under 2 hours and inference in under a minute.
A Methodology for Bridging the Gap between Regional- and City-Scale Climate Simulations for the Urban Thermal Environment
Combines artificial neural networks and regression to downscale EURO-CORDEX temperatures from 12 km to 1 km over Athens, projecting Tmax and Tmin rises under RCP scenarios.
Temporal downscaling of precipitation from climate model projections using machine learning
Applies machine learning to temporally downscale precipitation from climate model projections to finer time steps; no abstract was available, so details are limited.
Intercomparison of deep learning architectures for the prediction of precipitation fields
Compares deep learning architectures for European precipitation fields; U-Net performs best, and a single-level U-Net matches deeper versions on extremes.
On the modern deep learning approaches for precipitation downscaling
Compares DeepSD, ConvLSTM, U-Net and SR-GAN for precipitation downscaling against IMD station data; SR-GAN performs best.
Is Climate Change Affecting the Monsoon Months in Pune, India?
ANN statistical downscaling of five GCMs projects rising non-monsoon and falling monsoon rainfall for Pune, India.
Application of copula‐based approach as a new data‐driven model for downscaling the mean daily temperature
Vine copula downscaling of daily mean temperature outperforms ANN, SVR, GP and ANFIS, reducing RMSE by 20–82% at one Iranian station.
Spatio-temporal Downscaling Emulator for Regional Climate Models: a Comparative Study
Compares INLA and varycoef for a spatio-temporal varying-coefficient statistical emulator that downscales NARCCAP regional climate model output.
Improving trajectory calculations using deep learning inspired single image superresolution
Applies EDSR-style super-resolution to ERA5 winds, halving RMSE versus linear interpolation and reducing 48-hour FLEXPART trajectory errors by at least 49.5%.
Using GAN Neural Networks for Super-Resolution Reconstruction of Temperature Fields
GAN with sub-pixel convolution performs 8x super-resolution of ERA5 temperature fields, improving RMSE 37% over bicubic interpolation.
IBM ResearchNanjing University of Information Science & Technology (NUIST)
FC-ZSM: Spatiotemporal Downscaling of Rain Radar Data Using a Feature Constrained Zooming Slow-Mo Network
Feature-constrained Zooming Slow-Mo network spatiotemporally downscales rain radar fields, outperforming U-Net variants and interpolation on a public radar dataset.
Nanjing University of Information Science & Technology (NUIST)
Function-Based Troposphere Tomography Technique for Optimal Downscaling of Precipitation
GNSS tomography humidity fields improve statistical and ANN precipitation downscaling, cutting monthly precipitation RMSE by about 38% on average.
A Hybrid Statistical Downscaling Framework Based on Nonstationary Time Series Decomposition and Machine Learning
Bivariate EMD decomposition combined with machine learning downscales CanESM monthly precipitation and temperature at California stations, beating wavelet-based baselines.
Assessing statistical downscaling in Argentina: Daily maximum and minimum temperatures
Compares analog, GLM and neural network statistical downscaling of daily Tx and Tn at 101 Argentine stations across three reanalyses, including a warm-period extrapolation test.
Downscaling model for rainfall based on the influence of typhoon under climate change
GA-RBF neural network downscales CMIP5 GCM output to monthly rainfall at Taichung and Hualien, Taiwan, with separate typhoon and non-typhoon rainfall modelling.
Predictive Modelling of Statistical Downscaling Based on Hybrid Machine Learning Model for Daily Rainfall in East-Coast Peninsular Malaysia
Two-phase SVC-RVM hybrid statistically downscales daily rainfall on Malaysia's east coast, outperforming SVR and ANN variants in wet-day classification and extremes.
Merging and Downscaling Soil Moisture Data From CMIP6 Projections Using Deep Learning Method
A deep learning model fuses 25 CMIP6 ESMs and downscales soil moisture to 0.25° over China, outperforming individual models and the multi-model mean.
Nanjing University of Information Science & Technology (NUIST)
High-resolution wind speed forecast system coupling numerical weather prediction and machine learning for agricultural studies — a case study from South Korea
Random forests best downscale NWP wind speed to 100 m over South Korea for agriculture, with 48-hour RMSE near 0.8 m/s.
On deep learning-based bias correction and downscaling of multiple climate models simulations
Deep learning bias correction and downscaling of multiple CMIP climate model simulations; abstract not available, so scope is inferred from the title.
Case Study: Development of the CNN Model Considering Teleconnection for Spatial Downscaling of Precipitation in a Climate Change Scenario
CNN with teleconnection predictors downscales GCM precipitation at three Korean stations, outperforming quantile mapping with mean correlation of 69%.
A critical view on the suitability of machine learning techniques to downscale climate change projections: Illustration for temperature with a toy experiment
Toy experiments show neural networks and SVR downscaling temperature extrapolate poorly outside calibration range, unlike multiple linear regression.
A Generative Deep Learning Approach to Stochastic Downscaling of Precipitation Forecasts
GAN and VAE-GAN downscale low-resolution forecasts to radar-scale precipitation maps with CRPS matching state-of-the-art pointwise post-processing, including heavy rain.
A 148-year precipitation oxygen isoscape for China generated based on data fusion and bias correction of iGCMs simulations
Builds a 148-year, 50-60 km precipitation oxygen isoscape for China by CNN fusion and bias correction of eight isotope-enabled GCMs, correlation above 0.8.
Statistical downscaling of numerical weather prediction based on convolutional neural networks
CNN-based statistical downscaling of NWP fields to turbine hub height matches physical-model accuracy for wind power forecasting at one farm.
An attention mechanism based convolutional network for satellite precipitation downscaling over China
Attention-based CNN downscales GPM IMERG monthly precipitation over China, using a physics-based degradation loss and in-situ measurements for correction.
A Novel Reference-Based and Gradient-Guided Deep Learning Model for Daily Precipitation Downscaling
Proposes a reference-based, gradient-guided deep network for daily precipitation downscaling from GPM and ERA5 with topography, outperforming baseline super-resolution methods.
Increasing the accuracy and resolution of precipitation forecasts using deep generative models
CorrectorGAN jointly bias-corrects and super-resolves global ensemble precipitation forecasts over the US against radar, approaching an operational regional model in seconds.
Evaluation of statistical downscaling methods for climate change projections over Spain: Present conditions with imperfect predictors (global climate model experiment)
Compares five statistical downscaling methods, including ANN, SVM and kernel ridge, over 18 CMIP5 GCMs for Spain, quantifying errors from imperfect GCM predictors.
Wind‐Topo: Downscaling near‐surface wind fields to high‐resolution topography in highly complex terrain with deep learning
Wind-Topo uses deep learning to downscale COSMO-1 near-surface winds to 50 m over Swiss Alpine topography, reproducing orographic effects at 60 validation stations.
Combining computational fluid dynamics and neural networks to characterize microclimate extremes: Learning the complex interactions between meso-climate and urban morphology
Trains feedforward and deep neural networks on CFD simulations to downscale mesoscale wind and temperature to urban microscale, reaching R2 of 0.91 and 0.94.
A Comparison between Analog Ensemble and Convolutional Neural Network Empirical-Statistical Downscaling Techniques for Reconstructing High-Resolution Near-Surface Wind
Compares analog ensemble and CNN empirical-statistical downscaling for 10 m winds over complex terrain, finding the analog ensemble has smaller errors than the CNN.
Machine learning‐based statistical downscaling of wind resource maps using multi‐resolution topographical data
Machine learning with new topography-derived variables downscales Weibull wind resource maps over South Korea, cutting RMSE by about 15% versus baseline inputs.
Wavelet-based predictor screening for statistical downscaling of precipitation and temperature using the artificial neural network method
Wavelet coherence predictor screening improves ANN statistical downscaling of CMIP5/CMIP6 precipitation and temperature, projecting precipitation declines of 7–29% for the study region.
Multi-model Ensemble Analysis with Neural Network Gaussian Processes
NN-GPR combines multiple climate models into a 50 km projection without regional models, rivaling RCMs for temperature and precipitation tails.
Reconstructing High Resolution ESM Data Through a Novel Fast Super Resolution Convolutional Neural Network (FSRCNN)
FSRCNN-ESM super-resolves coarse Earth system model output from 1° to 0.25° over North America, outperforming other super-resolution methods on temperature, fluxes and precipitation.
AtmoDist: Self-supervised Representation Learning for Atmospheric Dynamics
Self-supervised network learns atmospheric dynamics by predicting temporal distance between ERA5 fields; the resulting distance loss improves downscaling and missing-data interpolation.
A station-data-based model residual machine learning method for fine-grained meteorological grid prediction
Model residual machine learning uses station forecasts to fit ECMWF residuals, producing high-resolution grids of temperature, humidity and wind that beat interpolation.
Institute of Atmospheric Physics, Chinese Academy of Sciences
To Accurately and Lightly Downscale the Temperature Field by Deep Learning
Light-CLDASSD, a lightweight deep model with station data, downscales temperature from 1/16° to 0.01° and beats HRCLDAS by 0.08 °C RMSE.
Super Resolution Reconstruction of E3SM Data Using a FSRCNN
FSRCNN-ESM, a modified fast super-resolution CNN, downscales E3SM climate output from 1° to 0.25° over North America with better accuracy than baselines.
Convolutional conditional neural processes for local climate downscaling
Convolutional conditional neural processes provide off-grid multisite temperature and precipitation downscaling over Europe, beating existing methods and improving extreme precipitation representation.
Machine Learning as a Downscaling Approach for Prediction of Wind Characteristics under Future Climate Change Scenarios
An M5' decision tree with model output statistics downscales BNU-ESM CMIP5 output, projecting North Atlantic wind speed declines and a 13–19% drop in wind energy by 2030–2040.
ResDeepD: A residual super-resolution network for deep downscaling of daily precipitation over India
ResDeepD, a residual super-resolution CNN, downscales 1° Indian monsoon rainfall to 0.25° and outperforms SRCNN, DeepSD and Nest-UNet, including during cyclonic storms.
DL4DS - Deep Learning for empirical DownScaling
DL4DS is a Python library for training configurable CNN downscaling models on gridded Earth data, demonstrated on CAMS air-quality fields over the Mediterranean.
Spatial downscaling method and application of CMIP5 global climate models based on improved U-Net in Lake Poyang Basin
An improved U-Net downscales CMIP5 projections over Lake Poyang Basin, outperforming linear regression and projecting warming and changing precipitation under RCP2.6 and RCP8.5.
Regression Analysis on Global Wind Speed Using Reanalysis Data
Random forest on topographic features reproduces global ERA5 wind speed with R2 of 1.00, suggesting reanalysis wind can be downscaled for wind resource assessment.
DeepUrbanDownscale: A physics informed deep learning framework for high-resolution urban surface temperature estimation via 3D point clouds
Physics-informed network downscales urban land surface temperature to 30 m using 3D point clouds, with errors under 0.2 K against satellite measurements.
Machine learning for Earth System Science (ESS): A survey, status and future directions for South Asia
Survey of machine learning across Earth system science in South Asia, covering atmospheric, ocean and other domains, with emphasis on downscaling and forecasting applications.
Simulation of Daily Snow Depth Data in China Based on the NEX-GDDP
Backpropagation neural network simulates daily snow depth from NEX-GDDP downscaled projections across China to 2065, validated with r = 0.94 for monthly snow cover duration.
Extension of Convolutional Neural Network along Temporal and Vertical Directions for Precipitation Downscaling
3D CNNs extended in time and vertical directions improve watershed-scale daily precipitation downscaling over a 2D CNN, with the vertical variant best on RMSE and NSE.
Deep-Learning-Based Harmonization and Super-Resolution of Near-Surface Air Temperature from CMIP6 Models (1850–2100)
Super-resolution deep learning merges 31 CMIP6 models into a 0.5° monthly near-surface temperature dataset (1850–2100) with bias correction and downscaling.
Nanjing University of Information Science & Technology (NUIST)
Evaluation of statistical downscaling methods for climate change projections over Spain: Future conditions with pseudo reality (transferability experiment)
Intercomparison of five statistical downscaling methods, including ANN, SVM and kernel ridge regression, reveals transferability problems for future climate over Spain.
A Machine Learning Based Downscaling Approach to Produce High Spatio-Temporal Resolution Land Surface Temperature of the Antarctic Dry Valleys from MODIS Data
Machine learning downscales MODIS land surface temperature from 1 km to 30 m over Antarctic Dry Valleys using Landsat 8 as reference, achieving R² of 0.78.
MSG-GAN-SD: A Multi-Scale Gradients GAN for Statistical Downscaling of 2-Meter Temperature over the EURO-CORDEX Domain
MSG-GAN-SD, a multi-scale gradient GAN, downscales ERA-Interim 2-m temperature from 83 km to 14 km over the EURO-CORDEX domain.
Effect of Climate Change on the Rainfall Trends: A case study of Pune, India
Artificial neural network downscales five GCMs' rainfall over Pune, India, giving a higher correlation than distribution-based scaling for the best models.
Improving the accuracy and spatial resolution of precipitable water vapor dataset using a neural network-based downscaling method
Neural network downscaling raises the spatial resolution of a precipitable water vapor dataset; the abstract is missing, so accuracy gains are unreported.
Regional analysis of climate projections using Bias Corrected spatial disaggregated super-resolution convolutional neural networks
Bias-corrected super-resolution CNN downscales six GCMs' daily precipitation and temperature from 1.25° to 0.25° over India, targeting extremes better than SRCNN and LSTM.
Prediction of Cloud Fractional Cover Using Machine Learning
Compares ConvLSTM and multiple regression for statistical downscaling of future cloud fractional cover; regression performs best overall, while ConvLSTM wins in coastal and mountain areas.
A Rainfall Disaggregation Scheme for Generating Fine Time-scale Extreme Rainfall under Climate Change
An artificial neural network disaggregates 3-hourly rainfall to hourly extreme rainfall for Cambodia using ERA5 covariates, reaching R2 of 0.93.
Application of ensemble machine learning model in downscaling and projecting climate variables over different climate regions in Iran
Ensemble machine learning downscales and projects climate variables across Iran's climate regions, using GCM output for regional climate projections.
Impact of climate change on hydro-climatological parameters in North Cyprus: application of artificial intelligence-based statistical downscaling models
ANN, ANFIS and MLR statistical downscaling of BNU-ESM CMIP5 predictors projects up to 22% less rainfall in Famagusta and Nicosia by 2040 under RCP4.5.
Learning to Correct Climate Projection Biases
Regularized adversarial domain adaptation corrects climate model daily precipitation biases over the contiguous US across moments while keeping spatiotemporal consistency.
Generating 1 km Spatially Seamless and Temporally Continuous Air Temperature Based on Deep Learning over Yangtze River Basin, China
Deep belief network fuses satellite, model and station data to produce 1 km daily air temperature over the Yangtze basin, with RMSE 1.96 °C.
Long-Term Maximum and Minimum Temperature Projections Over Metro Vancouver, Canada
GRU recurrent network statistically downscales WRF/MPI-ESM-LR RCP8.5 output to project Metro Vancouver daily temperature extremes to 2100, with warming near 3°C.
Deep Learning‐Based Super‐Resolution Climate Simulator‐Emulator Framework for Urban Heat Studies
Adversarial sub-pixel CNN super-resolves 2.5 km regional climate simulations of Montreal to 250 m, improving urban temperature errors and heat-stress indicators.
WiSoSuper: Benchmarking Super-Resolution Methods on Wind and Solar Data
Benchmarks ESRGAN, EDSR and PhIREGAN for super-resolving NWP wind and solar fields, and releases an ML-ready dataset for renewable forecasting.
Massachusetts Institute of Technology (MIT)California Institute of Technology (Caltech)University of California, Berkeley
Hybrid precipitation downscaling over coastal watersheds in Japan using WRF and CNN
Hybrid WRF and CNN downscaling produces 6-km daily precipitation from ERA-Interim over a Japanese watershed, cutting cost versus pure dynamical downscaling.
Spatio-temporal variations of land surface temperature and precipitation due to climate change in the Jhelum river basin, India
Wavelet-ANN statistical downscaling of CanESM2 under RCP8.5 projects 17-25% wetter and 2-3 °C warmer conditions in the Jhelum basin by 2100.
Computationally-Efficient Climate Predictions using Multi-Fidelity Surrogate Modelling
Multi-fidelity Gaussian processes fuse a coarse GCM and a regional climate model to predict Peruvian temperatures, evaluating only 6% of the high-fidelity region.
Super-Resolution of Near-Surface Temperature Utilizing Physical Quantities for Real-Time Prediction of Urban Micrometeorology
A CNN super-resolution model with attention upscales urban near-surface temperature from 20 m to 5 m, beating bicubic interpolation and enabling fast real-time inference.
Comparisons of Machine Learning Methods of Statistical Downscaling Method: Case Studies of Daily Climate Anomalies in Thailand
Compares linear regression, Gaussian process, SVM and deep learning for statistical downscaling of Thai temperature and precipitation, with deep learning performing best.
Statistical downscaling of daily precipitation over southeastern South America: Assessing the performance in extreme events
Compares analogue, neural network and GLM statistical downscaling of daily precipitation extremes over southeastern South America; no single method performs best across all aspects.
Evaluation of statistical downscaling methods for climate change projections over Spain: Present conditions with perfect predictors
Compares analog, regression, ANN, SVM and kernel ridge downscaling over Spain with perfect predictors, finding nonlinear methods add value for temperature.
On the suitability of deep convolutional neural networks for continental-wide downscaling of climate change projections
Assesses whether deep CNN downscaling of GCM projections over Europe yields plausible end-of-century changes, finding smaller departures from raw GCM signals than GLMs.
Modeling high-resolution precipitation by coupling a regional climate model with a machine learning model: an application to Sai Gon–Dong Nai Rivers Basin in Vietnam
Couples a regional climate model with a machine learning model to produce high-resolution precipitation for the Sai Gon–Dong Nai basin in Vietnam; abstract unavailable, classified from title.
A data-driven framework for the stochastic reconstruction of small-scale features with application to climate data sets
A data-driven stochastic framework reconstructs small-scale features in climate data sets; no abstract was provided, so impact is hard to assess.
Statistical Downscaling of Rainfall Under Climate Change in Krishna River Sub-basin of Andhra Pradesh, India Using Artificial Neural Network (ANN)
Applies ANN with PCA to downscale CanESM2 rainfall projections for three Krishna sub-basin districts under three RCP scenarios, validated by R² and RMSE.
Deconditional Downscaling with Gaussian Processes
Bayesian Gaussian-process deconditioning method for statistical downscaling of atmospheric fields, with a proven minimax-optimal rate and gains over existing methods.
A Downscaling–Merging Scheme for Improving Daily Spatial Precipitation Estimates Based on Random Forest and Cokriging
Random forest downscales GPM satellite daily precipitation to 0.01° and cokriging merges it with gauges, retaining accuracy while improving spatial detail.
Spatio-Temporal Downscaling of Climate Data Using Convolutional and Error-Predicting Neural Networks
Deep convolutional networks reconstruct high-resolution weather and climate fields from low-resolution simulations, and generalize to observed data for precipitation and other variables.
Deep Learning for Daily Precipitation and Temperature Downscaling
SRDRN residual CNN downscales daily precipitation and temperature from 100 km to 4 km, reproduces extremes, and transfers across regions.
DeepDT: Generative Adversarial Network for High-Resolution Climate Prediction
DeepDT trains a GAN with separate generator and discriminator steps to reduce artifacts in CNN-based climate downscaling, outperforming most CNN baselines.
Novel statistical downscaling emulator for precipitation projections using deep Convolutional Autoencoder over Northern Africa
Deep convolutional autoencoder emulates statistical downscaling of precipitation projections over Northern Africa; abstract unavailable.
Nanjing University of Information Science & Technology (NUIST)
Statistical downscaling using principal component regression for climate change impact assessment at the Cauvery river basin
Applies principal component regression to downscale 26 CMIP5 models over the Cauvery basin and ranks them by precipitation skill; modest machine learning novelty.
Intercomparison of statistical downscaling models: a case study of a large-scale river basin
Proposes XGBoost-based statistical downscaling of ACCESS1.0 GCM output for Indian river basin, with bias near zero and NSE 0.75-0.85 against 35 stations.
Creating 1-km long-term (1980–2014) daily average air temperatures over the Tibetan Plateau by integrating eight types of reanalysis and land data assimilation products downscaled with MODIS-estimated temperature lapse rates based on machine learning
Gradient boosting fuses eight reanalyses with MODIS-based lapse rates to produce a 1-km daily air temperature dataset over the Tibetan Plateau for 1980-2014.
Fast and accurate learned multiresolution dynamical downscaling for precipitation
CNNs trained with MSE and conditional GANs emulate 12 km precipitation from 50 km simulations, running 30 years in 14 minutes versus one month.
Downscaling diurnal temperature over west and southwest Iran : A comparison of statistical downscaling approaches
Compares artificial neural networks with SDSM for downscaling diurnal temperature in western Iran; ANN gives lower errors at mountain stations.
An Application of ANN Ensemble for Estimating of Precipitation Using Regional Climate Models
ANN ensemble combines five regional climate models to correct Korean precipitation against observations, improving on individual RCMs, most strongly inland.
Research on downscaling of TRMM precipitation products based on deep learning: Exemplified by northeast China
Deep learning is applied to downscale TRMM satellite precipitation products over northeast China; the abstract is missing, so methods and results are unverified.
CliGAN: A Structurally Sensitive Convolutional Neural Network Model for Statistical Downscaling of Precipitation from Multi-Model Ensembles
CliGAN, a Wasserstein GAN with structural similarity loss, downscales multi-model CMIP6 annual maximum precipitation to regional gridded fields.
Downscaling Satellite and Reanalysis Precipitation Products Using Attention-Based Deep Convolutional Neural Nets
Attention-based U-Net downscales ERA5-Land, TRMM and IMERG monthly precipitation to 1 km over Central Texas, using PRISM as target over 2001–2018.
Downscaling TRMM Monthly Precipitation Using Google Earth Engine and Google Cloud Computing
Gradient boosting, SVR and ANN downscale TRMM precipitation from 25 km to 1 km using vegetation indices, with the ANN performing best in the Mekong region.
FireSRnet: Geoscience-Driven Super-Resolution of Future Fire Risk from Climate Change
A super-resolution CNN enhances fire risk maps from climate models, outperforming interpolation at 4x and 8x and projecting risk for 2040 and 2100.
Assessment of predictability in Downscaling GEFS Precipitation Forecasts
Feedforward neural networks statistically downscale GEFS precipitation forecasts, assessing which predictors matter at each lead time up to 16 days.
Deep-learning based down-scaling of summer monsoon rainfall data over Indian region
Compares SRCNN variants and DeepSD for 4x downscaling of Indian summer monsoon rainfall, finding DeepSD gives the best spatial distribution and lowest RMSE.
Change in Rainfall Patterns in the Hilly Region of Uttarakhand due to the Impact of Climate Change
An ANN with PCA statistically downscales CanESM2 projections to Uttarakhand districts, projecting monsoon rainfall changes of up to 200% under RCP8.5.
Deep-Learning-Based Gridded Downscaling of Surface Meteorological Variables in Complex Terrain. Part II: Daily Precipitation
Nest-UNet, a UNet variant with nested skip connections, downscales daily precipitation from 0.25° to 4 km over western North America, outperforming bias-correction baselines.
Deep-Learning-Based Gridded Downscaling of Surface Meteorological Variables in Complex Terrain. Part I: Daily Maximum and Minimum 2-m Temperature
UNet-AE adds an elevation autoencoder branch to downscale daily TMAX/TMIN from 0.25° to 4 km, cutting MAE by over 10% in unseen regions.
A Climate Downscaling Deep Learning Model considering the Multiscale Spatial Correlations and Chaos of Meteorological Events
A two-stage deep learning climate downscaling network with stepwise reconstruction and ensemble inference captures multiscale spatial correlations and extremes better than simple SISR analogies.
Comparison of Different Methods for Spatial Downscaling of GPM IMERG V06B Satellite Precipitation Product Over a Typical Arid to Semi-Arid Area
Compares two regression and three ML methods (ANN, SVM, RF) for downscaling IMERG to 1 km over Gansu; ML beats regression against 80 rain gauges.
Strict Enforcement of Conservation Laws and Invertibility in CNN-Based Super Resolution for Scientific Datasets
Adds a differentiable operator to CNN super-resolution so outputs exactly conserve low-resolution averages, demonstrated on radar, satellite and climate model data.
Radar Super Resolution Using a Deep Convolutional Neural Network
A deep CNN super-resolves NEXRAD radar reflectivity, outperforming interpolation at 4x and 8x resolution increases on objective and perceptual metrics after six months of training data.
Projection of Temperature and Precipitation using Multiple Linear Regression and Artificial Neural Network as a Downscaling Methodology for Upper Bhima Basin
An ANN outperforms multiple linear regression for downscaling GCM temperature and precipitation projections over the Upper Bhima basin in India.
Augmented Convolutional LSTMs for Generation of High-Resolution Climate Change Projections
Auxiliary-variable-informed convolutional LSTM statistically downscales CMIP-style precipitation from 115 km to 25 km over India, with better extreme-event prediction than baselines.
Understanding climate change with statistical downscaling and machine learning
Philosophical analysis of how statistical downscaling and machine learning contribute to understanding climate change; a perspective piece rather than a new method.
Deep-learning based climate downscaling using the super-resolution method: a case study over the western US
Super-resolution CNN downscales daily temperature and precipitation to 4 km over the western US in about 23 ms per day, competitive with dynamical downscaling.
TRU-NET: A Deep Learning Approach to High Resolution Prediction of Rainfall
TRU-NET, a convolutional-recurrent U-Net with cross attention, predicts high-resolution rainfall from climate model fields with lower RMSE than a dynamical model's rainfall output.
A comparative study of convolutional neural network models for wind field downscaling
Compares CNN architectures and proposes DeepRU, a U-Net variant, to downscale ERA5-driven 100 m wind forecasts from 31 km to 9 km over the Alps.
ClimAlign: Unsupervised statistical downscaling of climate variables via normalizing flows
ClimAlign uses normalizing flows for unsupervised statistical downscaling of daily temperature and precipitation, matching supervised baselines while enabling conditional and unconditional sampling.
Statistical downscaling of daily temperature and precipitation over China using deep learning neural models: Localization and comparison with other methods
CNN statistical downscaling of daily temperature and precipitation over China outperforms GLM and BCSD/BCCI for precipitation, though ERA-Interim is better for temporal correlation.
Nanjing University of Information Science & Technology (NUIST)
Statistical Downscaling of Temperature Distributions from the Synoptic Scale to the Mesoscale Using Deep Convolutional Neural Networks
A deep CNN downscales 22-km synoptic temperature fields to 5-km mesoscale, cutting RMSE over high terrain from 2.7 K to 1.0 K.
Adversarial super-resolution of climatological wind and solar data
Adversarial deep learning super-resolves wind and solar outputs from global climate models by up to 50x, producing physically consistent fields for renewable resource assessment.
Spatially Downscaling IMERG at Daily Scale Using Machine Learning Approaches Over Zhejiang, Southeastern China
SVM and BPNN downscale IMERG daily precipitation over Zhejiang; SVM raises R² from 0.344 to 0.615 against gauges.
Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields with a Generative Adversarial Network
A recurrent stochastic GAN generates ensembles of temporally consistent high-resolution precipitation radar and cloud optical thickness fields from low-resolution sequences.
Fine scale surface climate in complex terrain using machine learning
Machine learning with PCA, ANN and quantile mapping downscales GCM output to 30 m sub-daily temperature, humidity and wind in complex terrain, improving on GCM accuracy.
Spatial and Temporal Downscaling of TRMM Precipitation with Novel Algorithms
Two downscaling methods (GWR and BP-ANN, plus bias correction and PCA) refine TRMM rainfall from 25 km to 1 km over Hubei, validated against 27 gauges.
Optimization and Variants of Quantile-Based Methods for Bias Corrections of Statistically Downscaled Precipitation Data
Neural-network-augmented quantile mapping bias corrects GCM-downscaled precipitation at 18 Florida gauges, reproducing extreme precipitation indices better than existing quantile methods.
Configuration and intercomparison of deep learning neural models for statistical downscaling
Systematic VALUE-based comparison of CNN downscaling against linear benchmarks over Europe finds CNNs help mainly for precipitation and temperature extremes, including under warm-period extrapolation.
Climate change projections and extremes for Costa Rica using tailored predictors from CORDEX model output through statistical downscaling with artificial neural networks
Perfect-prognosis artificial neural networks statistically downscale precipitation and temperature for Costa Rica using tailored CORDEX predictors and station data, with projected extremes to 2100.
PreciPatch: A Dictionary-based Precipitation Downscaling Method
PreciPatch uses dictionary learning to downscale hourly GCM precipitation, outperforming bicubic, RainFARM and DeepSD on IMERG data.
Sequence-based statistical downscaling and its application to hydrologic simulations based on machine learning and big data
Sequence-based machine-learning statistical downscaling of climate data for hydrologic simulation; abstract not available, so classification is based on the title only.
Downscaling and Projection of Multi-CMIP5 Precipitation Using Machine Learning Methods in the Upper Han River Basin
Bayesian model averaging with SVR, MLP and random forest downscales eight CMIP5 models' precipitation in the Upper Han basin, with best PCC 0.82.
New Approach to Multisite Downscaling of Precipitation by Identifying Different Set of Atmospheric Predictor Variables
Support vector regression multisite precipitation downscaling with climatological sub-regions, producing 2012–2050 rainfall projections for Ireland under four RCP scenarios.
Hyper-local, efficient extreme heat projection and analysis using machine learning to augment a hybrid dynamical-statistical downscaling technique
Uses machine learning to augment a hybrid dynamical-statistical downscaling method for hyper-local extreme heat projections; based on title only, abstract missing.
Performance of statistical and machine learning ensembles for daily temperature downscaling
Compares statistical and machine learning ensembles for daily temperature downscaling; abstract unavailable, so judgment rests on the title.
CHARACTERISTICS OF PRECIPITATION DOWNSCALING BY MEANS OF DEEP LEARNING METHOD
Examines how CNN precipitation downscaling from atmospheric fields depends on input domain, pressure levels and variables. Finds more inputs do not always improve basin rainfall accuracy.
ResLap: Generating High-Resolution Climate Prediction Through Image Super-Resolution
ResLap applies a residual-dense Laplacian pyramid super-resolution network to downscale China Meteorological Administration climate data and releases datasets.
Statistical downscaling of monthly precipitation over India using long short term memory neural network (LSTM)
LSTM network statistically downscales monthly precipitation over India, aiming to produce finer regional rainfall estimates from large-scale predictors.
Using Artificial Neural Networks to predict monthly precipitation for the Cali river basin, Colombia
Artificial neural networks statistically downscale CMIP5 models to monthly rainfall for the Cali river basin, projecting a shift of rainfall toward the middle and lower basin by 2100.
Long‐Term Precipitation Estimates Generated by a Downscaling‐Calibration Procedure Over the Tibetan Plateau From 1983 to 2015
Cubist regression downscales PERSIANN-CDR to 1 km and calibrates it, yielding 1983–2015 Tibetan Plateau precipitation with bias reduced from 41% to about 5%.
Downscaling Precipitation in the Data-Scarce Inland River Basin of Northwest China Based on Earth System Data Products
Statistical downscaling of monthly precipitation to 1 km for the Tarim River Basin, combining polynomial regression with batch gradient descent; it outperforms ordinary polynomial regression.
Machine learning for downscaling: the use of parallel multiple populations in genetic programming
Parallel multi-population genetic programming downscales large-scale predictors to daily station Tmin and Tmax, generalising better than single-deme GP with fewer outliers.
Optimized Artificial Neural Networks-Based Methods for Statistical Downscaling of Gridded Precipitation Data
Neural networks optimized by PSO, ICA and GA downscale satellite precipitation over Austria, with ICA best and RMSE of 5.26 mm against 54 rain gauges.
Efficacy of hybrid neural networks in statistical downscaling of precipitation of the Bagmati River basin
Feed-forward and wavelet neural networks statistically downscale GCM predictors to Bagmati basin precipitation, projecting 2014–2036 decreases at the study sites.
Geographically Weighted Machine Learning and Downscaling for High-Resolution Spatiotemporal Estimations of Wind Speed
Geographically weighted ensemble of ML learners plus residual-network downscaling of reanalysis yields high-resolution wind speed over China, improving R2 by 12–16%.
Event selection for dynamical downscaling: a neural network approach for physically-constrained precipitation events
Neural network selects physically constrained precipitation events to make dynamical downscaling of climate simulations more efficient; abstract unavailable.
Stepwise extreme learning machine for statistical downscaling of daily maximum and minimum temperature
Stepwise extreme learning machine is applied to statistically downscale daily maximum and minimum temperature; abstract unavailable.
Improving Monsoon Precipitation Prediction Using Combined Convolutional and Long Short Term Memory Neural Network
CNN-LSTM statistical downscaling of ECMWF precipitation over a monsoon basin beats quantile mapping and SVM, with gains up to two-week lead times.
Machine Learning for Projecting Extreme Precipitation Intensity for Short Durations in a Changing Climate
Machine learning temporally downscales climate projections to estimate short-duration extreme precipitation IDF curves, trained on about 2,000 US stations.
Downscaling and Disaggregation of Wind Speed to River Basin in India for IPCC SRES Scenarios
Support vector machine downscales CGCM3 wind speed projections to an Indian river basin under SRES scenarios, then disaggregates to daily values with a k-nearest neighbour method.
Potential Predictability of Regional Precipitation and Discharge Extremes Using Synoptic-Scale Climate Information via Machine Learning: An Evaluation for the Eastern Continental United States
Archetypal analysis plus a 2D CNN predicts eastern US regional precipitation and discharge extremes from synoptic-scale fields, with the most reliable skill from 300 hPa geopotential height.
Downscaling rainfall using deep learning long short‐term memory and feedforward neural network
LSTM and feedforward networks downscale CMIP5 precipitation for the Mekong Delta, projecting 20-25% wet-season increases and dry-season decreases for 2036-2065.
Observed and model simulated twenty-first century hydro-climatic change of Northern Ethiopia
ANNs downscale 30 CMIP5 GCMs for the Tekeze basin, projecting falling precipitation and streamflow to 2099 under RCP4.5 and RCP8.5, with multi-model ensembles more robust.
Integrating precipitation zoning with random forest regression for the spatial downscaling of satellite‐based precipitation: A case study of the Lancang–Mekong River basin
Random forest with precipitation zoning downscales satellite rainfall to 1 km over the Lancang–Mekong basin, cutting RMSE by 1–17%.
Artificial Neural Network Technique for Statistical Downscaling of Global Climate Model
Applies artificial neural networks for statistical downscaling of global climate model output to regional scales.
Improving Precipitation Estimation Using Convolutional Neural Network
CNN predicts daily precipitation at 14 US grid points from resolved dynamics, outperforming reanalysis and linear, forest and dense-network baselines.
Subdaily Rainfall Estimation through Daily Rainfall Downscaling Using Random Forests in Spain
Random forests downscale daily rainfall to subdaily statistics at over 700 Spanish gauges, improving skill and the Poisson cluster rainfall simulations built on it.
Downscaling Satellite Precipitation Estimates With Multiple Linear Regression, Artificial Neural Networks, and Spline Interpolation Techniques
Compares regression, ANN and spline downscaling of IMERG satellite rainfall in Austria; all improve accuracy, with spline best at correlation 0.56.
A hierarchical analysis of the impact of methodological decisions on statistical downscaling of daily precipitation and air temperatures
Systematic comparison of GLM and ANN statistical downscaling choices at 10 US sites; ANNs beat GLMs for temperature, while PCA predictors help only extremes.
Spatial and Temporal Analysis of Climate Change in the Future and Comparison of SDSM, LARS-WG and Artificial Neural Network Downscaling Methods (Case Study: Khuzestan Province)
Compares SDSM, LARS-WG and ANN downscaling for future climate change in Khuzestan Province; the abstract is missing, so findings are unclear.
Performance Evaluation of Artificial Neural Network Models for Downscaling and Predicting of Climate Variables
Evaluates ANN models for downscaling and predicting climate variables; the abstract is only a title, so results are unclear.
Climate modelling using ANN
ANN downscales GCM predictors to Jhelum basin stations, projecting 2100 warming of 1.4 to 1.6 C and a 31 to 35% precipitation drop.
Super-Resolution Simulation for Real-Time Prediction of Urban Micrometeorology
CNN super-resolution applied to low-resolution urban micrometeorology simulations gives 4x finer predictions faster than high-resolution runs, beating interpolation.
Assessing future rainfall uncertainties of climate change in Taiwan with a bootstrapped neural network‐based downscaling model
A bootstrapped RBF neural network downscales three GCMs to project Taiwan station rainfall for 2050–2090 with quantified uncertainty.
Uncertainty in Rainfall Intensity Duration Frequency Curves of Peninsular Malaysia under Changing Climate Scenarios
Combines MOS downscaling of GCM rainfall with ANN temporal disaggregation to project IDF curves for Peninsular Malaysia under RCP4.5 and RCP8.5 with uncertainty bounds.
ANN-based statistical downscaling of climatic parameters using decision tree predictor screening method
ANN-based statistical downscaling of climate parameters with decision-tree predictor screening; abstract unavailable, classification based on title only.
Climate modeling using ANN
An ANN downscales GCM predictors to four Jhelum basin stations, projecting 1.4–1.6°C warming and 31–35% precipitation decline by 2100.
Analysis of Precipitation Variability using Memory Based Artificial Neural Networks
Gamma memory neural networks, with and without GA optimization, downscale daily precipitation for the Barak basin, with GMNN-GA matching the standard SDSM model.
Future Predictions of Rainfall and Temperature Using GCM and ANN for Arid Regions: A Case Study for the Qassim Region, Saudi Arabia
Combines GCMs, a weather generator and ANN/ANFIS models to project Qassim rainfall and temperature; ANNs outperformed ANFIS for long-term predictions.
Intercomparison of machine learning methods for statistical downscaling: the case of daily and extreme precipitation
Intercomparison of machine learning methods for statistical downscaling of daily and extreme precipitation; abstract unavailable, classification based on title.
Evaluation of data-driven models to downscale rainfall parameters from global climate models outputs: the case study of Latyan watershed
Compares ANN, K-nearest-neighbour and ANFIS for downscaling GCM precipitation at a Latyan basin station; KNN performed best monthly.
Data-driven ensemble model to statistically downscale rainfall using nonlinear predictor screening approach
Ensemble statistical downscaling of rainfall using nonlinear predictor screening to select inputs; the abstract is missing, so details are inferred from the title.
Application of multivariate recursive nesting bias correction, multiscale wavelet entropy and AI-based models to improve future precipitation projection in upstream of the Heihe River, Northwest China
Combines multivariate recursive nesting bias correction with AI-based models to improve future precipitation projections over the Heihe River upstream; abstract missing.
DeepDownscale: a Deep Learning Strategy for High-Resolution Weather Forecast
DeepDownscale learns high-resolution weather forecast fields from low-resolution model output with a lightweight deep network, improving on interpolation and ensembles.
PERFORMANCE ASSESSMENT OF DIFFERENT BIAS CORRECTION METHODS IN STATISTICAL DOWNSCALING OF PRECIPITATION
An MLP downscales arid-region monthly precipitation, and four bias correction methods are compared, with power transformation found most reliable.
Downscaling of daily extreme temperatures in the Yarlung Zangbo River Basin using machine learning techniques
Applies machine learning to downscale daily extreme temperatures in the Yarlung Zangbo River Basin; abstract not available, so details are limited.
Statistical downscaling of precipitation using machine learning techniques
Applies machine learning techniques to statistically downscale precipitation; no abstract was available, so skill and data specifics are unconfirmed.
Multivariate Interpolation of Wind Field Based on Gaussian Process Regression
Gaussian process regression with a multi-scale anisotropic kernel interpolates NWP 10 m wind fields, cutting RMSE by up to 95% versus neural networks in cyclone cases.
Coupling dynamical and statistical downscaling for high-resolution rainfall forecasting: case study of the Red River Delta, Vietnam
Hybrid RCM and neural-network downscaling reproduces 6-km Red River Delta rainfall (r>0.9) at an 89% reduction in computing power.
Development of an Urban High-Resolution Air Temperature Forecast System for Local Weather Information Services Based on Statistical Downscaling
Statistical and machine-learning downscaling produces 25 m urban air temperature maps for Seoul, improving on the 1.5 km UM-LDAPS model.
Bat optimisation neural networks for rainfall forecasting: case study for Kuching city
A metaheuristic bat-optimised neural network downscales climate model predictors to Kuching station precipitation, reporting average r of 0.96 across three GCMs.
Conjunction of wavelet-entropy and SOM clustering for multi-GCM statistical downscaling
Wavelet-entropy and SOM predictor selection with ANN downscales multi-GCM station rainfall, projecting a 21–29% rainfall decrease for 2017–2050.
Quantifying Uncertainty in Discrete-Continuous and Skewed Data with Bayesian Deep Learning
Discrete-continuous Bayesian deep learning with lognormal likelihood gives calibrated aleatoric and epistemic uncertainty for DeepSD precipitation downscaling, especially at extremes.
Statistical downscaling of global climate model outputs to monthly precipitation via extreme learning machine: A case study
Extreme learning machine downscales reanalysis predictors to station monthly precipitation in Iran, outperforming ANN, genetic programming and quantile mapping.
Super-Resolution and Deep Learning for Climate Downscaling
Conference abstract on applying super-resolution deep learning to climate downscaling; no quantitative results are available from the abstract.
Deep neural networks for statistical downscaling of climate change projections
Deep neural networks are applied to statistically downscale climate change projections; no abstract or quantitative results are available.
Comparison of MLP-ANN Scheme and SDSM as Tools for Providing Downscaled Precipitation for Impact Studies at Daily Time Scale
An MLP-ANN downscaling scheme for daily precipitation in the Clutha watershed, New Zealand, is shown to outperform the standard SDSM regression model.
FORECASTING LONG TERM PRECIPITATION USING CUCKOO SEARCH OPTIMIZATION NEURAL NETWORK MODELS
Cuckoo search-trained feedforward and recurrent networks downscale GCM output to project long-term precipitation, forecasting a 6–7% rise by 2071–2100.
DeepSD: Generating High Resolution Climate Change Projections through Single Image Super-Resolution
DeepSD applies a multi-scale stacked super-resolution CNN to downscale daily precipitation from 1 degree to 1/8 degree over the continental US for climate projections.
Probabilistic Multisite Precipitation Downscaling by an Expanded Bernoulli–Gamma Density Network
An expanded Bernoulli-gamma neural density network downscales multisite daily precipitation from synoptic predictors, reproducing spatial covariances and generating realistic synthetic series in coastal British Columbia.
Also relevant (45)
Prithvi WxC: Foundation Model for Weather and Climate
Prithvi WxC is a 2.3-billion-parameter MERRA-2 foundation model that transfers to forecasting, downscaling, gravity-wave parameterization and extreme-event tasks.
Application of Wavelet Transform for Bias Correction and Predictor Screening of Climate Data
Wavelet-based bias correction outperforms quantile mapping for climate model precipitation in Iran, with wavelet-ANN predictor screening for downscaling.
All-weather precipitable water vapor map reconstruction using data fusion and machine learning-based spatial downscaling
Fuses multiple data sources with machine-learning spatial downscaling to reconstruct all-weather precipitable water vapour maps; no abstract available, so results are unverified.
Further analysis of cGAN: A system for Generative Deep Learning Post-processing of Precipitation
Trains a conditional GAN on ERA5 to generate downscaled 1 km precipitation ensembles over UK and US regions, competitive with IFS ensembles at 8–16 h lead.
AI Foundation Models for Weather and Climate: Applications, Design, and Implementation
Review and perspective arguing that transformer-based weather and climate foundation models are mature enough to build, outlining criteria for downscaling, wildfire and hurricane tasks.
Foundation modelsIBM ResearchKarlsruhe Institute of Technology (KIT)NASA
SuperBench: A Super-Resolution Benchmark Dataset for Scientific Machine Learning
SuperBench provides a super-resolution benchmark across fluids, cosmology and weather data, finding that deep SR methods miss fine-scale physics; weather is one of three domains.
A Review of Recent and Emerging Machine Learning Applications for Climate Variability and Weather Phenomena
Review of ML for climate variability modes, extreme weather precursors, observation-model integration, downscaling and bias correction, aimed at domain scientists.
Climate analysis & attributionNSF NCAR / UCARUniversity of California, BerkeleyLawrence Berkeley National Laboratory / NERSCOak Ridge National Laboratory
Spiking Neural Network based Water management in Drought Area of Maharashtra India: A Case Study of Osmanabad District
A spiking neural network downscales coarse precipitation to map meteorological drought (SPI) in Osmanabad, India; reported efficiency is claimed without rigorous benchmarking.
Diffusion Models for High-Resolution Solar Forecasts
Score-based diffusion models super-resolve coarse NWP to satellite-resolution day-ahead solar irradiance, giving probabilistic forecasts with many samples.
SwinRDM: Integrate SwinRNN with Diffusion Model Towards High-Resolution and High-Quality Weather Forecasting
SwinRDM couples a SwinRNN forecaster with a diffusion super-resolution step to reach 0.25° forecasts that beat IFS up to 5 days on Z500, T850, T2M and precipitation.
Detail Enhancement of AIRS/AMSU Temperature and Moisture Profiles Using a 3D Deep Neural Network
A 3D deep neural network sharpens AIRS/AMSU temperature and humidity retrievals, better capturing planetary boundary layer top inversions and improving PBL height estimates.
Observations & remote sensingNASAMassachusetts Institute of Technology (MIT)
Artificial Neural Network-Empowered Projected Future Rainfall Intensity-Duration-Frequency Curves Under Changing Climate
Artificial neural networks project future rainfall intensity-duration-frequency curves under climate change; the abstract is missing, so results are limited.
Spatiotemporal model based on transformer for bias correction and temporal downscaling of forecasts
Swin-transformer U-Net corrects NWP 2 m temperature and 10 m wind biases and produces hourly temporal downscaling, cutting RMSE by 0.78 versus the original forecasts.
Projection of Flight Cancellations and Economic Losses Caused by Future Climate Change
Applies machine learning and nearest-neighbour with genetic algorithms to downscale SSP projections, estimating flight cancellations and losses rising up to 8.29 times by 2100.
Deep Learning Provides Substantial Improvements to County‐Level Fire Weather Forecasting Over the Western United States
CFS-SR applies image super-resolution to CFSv2 to give 4 km county-level fire weather forecasts to 7 days, improving accuracy over the western US.
Extremes & hazardsUniversity of WashingtonMax Planck Institute for Biogeochemistry
AI-Based Climate Change Forecasting Models
Compares six AI approaches across four climate tasks, claiming hybrid physics-ML best at 6-month temperature RMSE; claims are unverified and methodology thin.
Boosting solar radiation predictions with global climate models, observational predictors and hybrid deep-machine learning algorithms
Hybrid deep and machine learning models use global climate model and observational predictors to boost solar radiation predictions; abstract not available, so scope is inferred from the title.
Convolutional autoencoders for spatially-informed ensemble post-processing
Convolutional autoencoders compress spatial NWP fields into features that augment station-based neural post-processing of 2 m temperature ensemble forecasts over Germany.
Post-processing & calibrationKarlsruhe Institute of Technology (KIT)
Simulation-based high-resolution fire danger mapping using deep learning
DeepFire neural emulator of fire spread turns high-resolution weather forecasts into next-day fire danger maps and ensembles for 13 Corsican fires.
From emission scenarios to spatially resolved projections with a chain of computationally efficient emulators: coupling of MAGICC (v7.5.1) and MESMER (v0.8.3)
Chains the MAGICC global emulator with the MESMER spatial emulator to produce millions of local temperature projections per emission pathway, validated against CMIP6 ensembles.
Importance of Weather Conditions in a Flight Corridor
Compares kriging, RBF, neural network and tree interpolation of GFS weather to high resolution for flight trajectories, with the choice shaping trajectory optimization.
A statistical bias correction technique for global climate model predicted near-surface temperature in India using the generalized regression neural network
A generalized regression neural network bias-corrects CNRM-CM5 near-surface temperature over India, and the trained model is applied to RCP4.5 and RCP8.5 projections to 2100.
Toward an Improved Global Longwave Downward Radiation Product by Fusing Satellite and Reanalysis Data
A machine-learning fusion of CERES, ERA5 and GLDAS yields a global hourly 0.25° longwave downward radiation product with R=0.97 against ground measurements over land.
Weather Radar Image Superresolution Using a Nonlocal Residual Network
Nonlocal residual CNN super-resolves weather radar echo images at x2 and x4, outperforming interpolation in quantitative metrics and edge detail.
Observations & remote sensingChina Meteorological Administration (CMA)
Easy-to-use spatial random-forest-based downscaling-calibration method for producing precipitation data with high resolution and high accuracy
Spatial random forest downscales and calibrates IMERG satellite precipitation in Sichuan using gauges and covariates, outperforming GWR, ANN and kriging.
Towards Representation Learning for Atmospheric Dynamics
AtmoDist is a self-supervised learned metric for atmospheric dynamics, trained to predict temporal distance between shuffled wind fields, and it improves GAN super-resolution of vorticity over MSE-based baselines.
Latent Linear Adjustment Autoencoder v1.0: a novel method for estimating and emulating dynamic precipitation at high resolution
A variational autoencoder with a linear latent penalty separates circulation-driven daily precipitation at 12 km over Europe, enabling dynamical adjustment of high-resolution fields.
Bias adjustment of Cordex database based on a nonlinear artificial neural network method to simulate summer precipitation in the southeast Iran
Nonlinear neural network bias adjustment of CORDEX regional projections for summer precipitation in southeast Iran. No abstract was provided.
RainNet: A Large-Scale Imagery Dataset and Benchmark for Spatial Precipitation Downscaling
Releases RainNet, a 62,400-pair precipitation downscaling dataset with 17 years of hourly maps and benchmarks 14 models, including deep learning baselines.
Generating future weather files under climate change scenarios to support building energy simulation – A machine learning approach
Uses a machine learning approach to generate future weather files under climate change scenarios for building energy simulation; no abstract available.
Enabling Smart Dynamical Downscaling of Extreme Precipitation Events With Machine Learning
CNNs pre-screen GCM circulation data so only a small subset needs kilometre-scale dynamical downscaling, retaining 92–98% of extreme precipitation events.
A Model Output Deep Learning Method for Grid Temperature Forecasts in Tianjin Area
End-to-end CNN post-processing (MODL) of TIGGE ECMWF forecasts improves Tianjin 2 m temperature RMSE over ECMWF, MOS and other ML baselines, especially in the first three days.
Post-processing & calibrationChina Meteorological Administration (CMA)
Estimation of All-Weather 1 km MODIS Land Surface Temperature for Humid Summer Days
Random forest fuses AMSR2 microwave data with MODIS annual-cycle terms to produce all-weather 1 km land surface temperature over South Korea, including cloudy days, with daytime RMSE of 2.5 °C.
Estimating near‐surface air temperature across Israel using a machine learning based hybrid approach
Two-stage random forest hybrid estimates gridded 1 km intra-daily near-surface air temperature over Israel from satellite skin temperature, with out-of-sample R2 of 0.96 for air temperature.
Observations & remote sensingPotsdam Institute for Climate Impact Research (PIK)
Gridded Extreme Precipitation Intensity–Duration–Frequency Estimates for the Canadian Landmass
Produces a gridded Canadian extreme precipitation IDF dataset by linking reanalysis predictors to extremes with a support vector machine, with 5–25% RMSE against observations.
Emulating Earth system model temperatures with MESMER: from global mean temperature trajectories to grid-point-level realizations on land
MESMER statistically emulates grid-point land temperature from CMIP5 models, producing a 40-model superensemble at negligible computational cost.
Deep learning-based air temperature mapping by fusing remote sensing, station, simulation and socioeconomic data
A deep belief network fuses satellite, station, simulation and socioeconomic data to map daily maximum air temperature across China at 0.01°, achieving RMSE 1.996°C.
Evaluating Multivariable Statistical Methods for Downscaling Nighttime Land Surface Temperature in Urban Areas
Compares multivariable ML methods for downscaling coarse nighttime urban land surface temperature; daytime visible-NIR predictors perform as well as physical variables, MAE within 2 K.
Study on Wind Simulations Using Deep Learning Techniques during Typhoons: A Case Study of Northern Taiwan
Deep neural network emulates WRF-simulated typhoon wind speeds at arbitrary locations in Northern Taiwan, matching WRF accuracy across 47 typhoon events.
Generative Adversarial Networks Capabilities for Super-Resolution Reconstruction of Weather Radar Echo Images
GAN-based super-resolution of degraded weather radar echo images produces sharper reconstructions with higher PSNR and SSIM than classical methods.
Observations & remote sensingChina Meteorological Administration (CMA)
Regression‐based regionalization for bias correction of temperature and precipitation
Regression-based regionalization trains an ANN on a few pixels per region to bias-correct CCSM temperature and precipitation over northern South America.
Prediction of Long-Term Near-Surface Temperature Based on NA-CORDEX Output
Neural networks (MLP, TLFN, NARX) correct daily near-surface temperature from NA-CORDEX output at Big Trout Lake, reaching R2 above 0.84.
Seasonal prediction of high‐resolution temperature at 2‐m height over Mongolia during boreal winter using both coupled general circulation model and artificial neural network
Neural networks post-process PNU CGCM seasonal hindcasts with sea-ice and SST indices to downscale winter 2-m temperature over Mongolia to 20 km, with improved skill.
Artificial neural network model for downscaling of temperature forecast over Western Himalaya
Artificial neural network statistically downscales MM5 temperature forecasts at three Western Himalaya stations, reducing RMSE to about 2 °C for day-2 forecasts.
The Climate Generator: Stochastic climate representation for glacial cycle integration
Bayesian neural networks emulate GCM monthly temperature and precipitation for glacial cycles, adding calibrated noise to mimic internal variability from 22 to 120 ka.