Adversarial super-resolution of climatological wind and solar data

Stengel, Karen; Glaws, Andrew; Hettinger, Dylan; King, Ruth

doi:10.1073/pnas.1918964117

Cited by 135 publications

(130 citation statements)

References 56 publications

(51 reference statements)

Supporting

Mentioning

129

Contrasting

Order By: Relevance

“…This general approach has achieved promising results but also has problems. The trained SR model often performs well when applied to a low-resolution data (upscaled from high-resolution data) and compared with the same high-resolution data, capturing both spatial patterns and sharp gradients (Geiss and Hardin, 2019), especially when using a GAN (Stengel et al, 2020). This result is not surprising,…”

mentioning

confidence: 88%

“…CC BY 4.0 License. SR methods have recently been applied to the challenging problems of downscaling precipitation (Vandal et al, 2017;Geiss and Hardin, 2019) and wind and solar radiation (Stengel et al, 2020), quantities that can vary sharply over spatial scales of 10 km or less depending on location. Downscaling with an SR model proceeds as follows (Vandal et al, 2017;Stengel et al, 2020): (1) take high-resolution data (either climate model output or gridded observations), upscale the data to a low resolution;…”

mentioning

confidence: 99%

See 1 more Smart Citation

Fast and accurate learned multiresolution dynamical downscaling for precipitation

Wang¹,

Liu²,

Foster³

et al. 2021

Preprint

View full text Add to dashboard Cite

Abstract. This study develops a neural network-based approach for emulating high-resolution modeled precipitation data with comparable statistical properties but at greatly reduced computational cost. The key idea is to use combination of low- and high- resolution simulations to train a neural network to map from the former to the latter. Specifically, we define two types of CNNs, one that stacks variables directly and one that encodes each variable before stacking, and we train each CNN type both with a conventional loss function, such as mean square error (MSE), and with a conditional generative adversarial network (CGAN), for a total of four CNN variants.We compare the four new CNN-derived high-resolution precipitation results with precipitation generated from original high resolution simulations, a bilinear interpolater and the state-of-the-art CNN-based super-resolution (SR) technique. Results show that the SR technique produces results similar to those of the bilinear interpolator with smoother spatial and temporal distributions and smaller data variabilities and extremes than the high resolution simulations. While the new CNNs trained by MSE generate better results over some regions than the interpolator and SR technique do, their predictions are still not as close as ground truth. The CNNs trained by CGAN generate more realistic and physically reasonable results, better capturing not only data variability in time and space but also extremes such as intense and long-lasting storms. The new proposed CNN-based downscaling approach can downscale precipitation from 50 km to 12 km in 14 min for 30 years once the network is trained (training takes 4 hours using 1 GPU), while the conventional dynamical downscaling would take 1 months using 600 CPU cores to generate simulations at the resolution of 12 km over contiguous United States.

show abstract

mentioning

confidence: 88%

mentioning

confidence: 99%

Fast and accurate learned multiresolution dynamical downscaling for precipitation

Wang¹,

Liu²,

Foster³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, while Percolator has demonstrated impressive performance for general feature setsespecially compared to other popular post-processors 45 -the use of shallow machine learning models (such as SVMs and gradient boosted decision trees) potentially leaves identifiable peptides on the table. In particular, Deep Learning 3 , has lead to many recent groundbreaking advances in other fields, such as computer vision, 29,33 speech recognition, 17,18 genomics, 2, 48 particle physics, 4 climate analysis, 43 and medical diagnosis. 32,35,41 We show that deep neural networks (DNNs) improve MS/MS universal post-processing accuracy across a large number of diverse datasets, identifying more PSMs than Percolator for both Prosit analysis and the post-processing of a recently developed scoring algorithm designed for new MS/MS machines and datasets.…”

Section: Introductionmentioning

confidence: 99%

“…However, while Percolator has demonstrated impressive performance for general feature sets– especially compared to other popular post-processors 45 –the use of shallow machine learning models (such as SVMs and gradient boosted decision trees) potentially leaves identifiable peptides on the table. In particular, Deep Learning 3 , has lead to many recent groundbreaking advances in other fields, such as computer vision, 29,33 speech recognition, 17,18 genomics, 2,48 particle physics, 4 climate analysis, 43 and medical diagnosis. 32,35,41…”

Section: Introductionmentioning

confidence: 99%

Deep Semi-Supervised Learning Improves Universal Peptide Identification of Shotgun Proteomics Data

Urban

Rocke

et al. 2020

Preprint

View full text Add to dashboard Cite

In proteomic analysis pipelines, machine learning post-processors play a critical role in improving the accuracy of shotgun proteomics analysis. Most often performed in a semi-supervised manner, such post-processors accept the peptide-spectrum matches (PSMs) and corresponding feature vectors resulting from a database search, train a machine learning classifier, and recalibrate PSM scores based on the resulting trained parameters, often leading to significantly more identified peptides across q-value thresholds. However, current state-of-the-art post-processors rely on shallow machine learning methods, such as SVMs, gradient boosted decision trees, and linear discriminant analysis. In contrast, the powerful learning capabilities of deep models have displayed superior performance to shallow models in an ever-growing number of other fields. In this work, we show that deep neural networks (DNNs) significantly improve the recalibration of shotgun proteomics data compared to the most accurate and widely used post-processors, such as Percolator and PeptideProphet. Furthermore, we show that DNNs are able to adaptively analyze complex datasets and features for more accurate universal post-processing, leading to both improved Prosit analysis and markedly better recalibration of recently developed p-value scoring functions.

show abstract

“…Ref. [59] used adversarial learning to downscale wind and solar output from several AOGCM climate scenarios to regional level high-resolution. Cheng et al, 2020 also used adversarial learning to downscale precipitation.…”

Section: Introductionmentioning

confidence: 99%

CliGAN: A Structurally Sensitive Convolutional Neural Network Model for Statistical Downscaling of Precipitation from Multi-Model Ensembles

Chaudhuri

Robertson

2020

Water

View full text Add to dashboard Cite

Despite numerous studies in statistical downscaling methodologies, there remains a lack of methods that can downscale from precipitation modeled in global climate models to regional level high resolution gridded precipitation. This paper reports a novel downscaling method using a Generative Adversarial Network (GAN), CliGAN, which can downscale large-scale annual maximum precipitation given by simulation of multiple atmosphere-ocean global climate models (AOGCM) from Coupled Model Inter-comparison Project 6 (CMIP6) to regional-level gridded annual maximum precipitation data. This framework utilizes a convolution encoder-dense decoder network to create a generative network and a similar network to create a critic network. The model is trained using an adversarial training approach. The critic uses the Wasserstein distance loss function and the generator is trained using a combination of adversarial loss Wasserstein distance, structural loss with the multi-scale structural similarity index (MSSIM), and content loss with the Nash-Sutcliff Model Efficiency (NS). The MSSIM index allowed us to gain insight into the model’s regional characteristics and shows that relying exclusively on point-based error functions, widely used in statistical downscaling, may not be enough to reliably simulate regional precipitation characteristics. Further use of structural loss functions within CNN-based downscaling methods may lead to higher quality downscaled climate model products.

show abstract

Adversarial super-resolution of climatological wind and solar data

Cited by 135 publications

References 56 publications

Fast and accurate learned multiresolution dynamical downscaling for precipitation

Fast and accurate learned multiresolution dynamical downscaling for precipitation

Deep Semi-Supervised Learning Improves Universal Peptide Identification of Shotgun Proteomics Data

CliGAN: A Structurally Sensitive Convolutional Neural Network Model for Statistical Downscaling of Precipitation from Multi-Model Ensembles

Contact Info

Product

Resources

About