Showing 1–14 of 14 results for author: Harder, P

Climate Variable Downscaling with Conditional Normalizing Flows

Authors: Christina Winkler, Paula Harder, David Rolnick

Abstract: Predictions of global climate models typically operate on coarse spatial scales due to the large computational costs of climate simulations. This has led to a considerable interest in methods for statistical downscaling, a similar process to super-resolution in the computer vision context, to provide more local and regional climate information. In this work, we apply conditional normalizing flows… ▽ More Predictions of global climate models typically operate on coarse spatial scales due to the large computational costs of climate simulations. This has led to a considerable interest in methods for statistical downscaling, a similar process to super-resolution in the computer vision context, to provide more local and regional climate information. In this work, we apply conditional normalizing flows to the task of climate variable downscaling. We showcase its successful performance on an ERA5 water content dataset for different upsampling factors. Additionally, we show that the method allows us to assess the predictive uncertainty in terms of standard deviation from the fitted conditional distribution mean. △ Less

Submitted 31 May, 2024; originally announced May 2024.

Multi-variable Hard Physical Constraints for Climate Model Downscaling

Authors: Jose González-Abad, Álex Hernández-García, Paula Harder, David Rolnick, José Manuel Gutiérrez

Abstract: Global Climate Models (GCMs) are the primary tool to simulate climate evolution and assess the impacts of climate change. However, they often operate at a coarse spatial resolution that limits their accuracy in reproducing local-scale phenomena. Statistical downscaling methods leveraging deep learning offer a solution to this problem by approximating local-scale climate fields from coarse variable… ▽ More Global Climate Models (GCMs) are the primary tool to simulate climate evolution and assess the impacts of climate change. However, they often operate at a coarse spatial resolution that limits their accuracy in reproducing local-scale phenomena. Statistical downscaling methods leveraging deep learning offer a solution to this problem by approximating local-scale climate fields from coarse variables, thus enabling regional GCM projections. Typically, climate fields of different variables of interest are downscaled independently, resulting in violations of fundamental physical properties across interconnected variables. This study investigates the scope of this problem and, through an application on temperature, lays the foundation for a framework introducing multi-variable hard constraints that guarantees physical relationships between groups of downscaled climate variables. △ Less

Submitted 2 August, 2023; originally announced August 2023.

Fourier Neural Operators for Arbitrary Resolution Climate Data Downscaling

Authors: Qidong Yang, Alex Hernandez-Garcia, Paula Harder, Venkatesh Ramesh, Prasanna Sattegeri, Daniela Szwarcman, Campbell D. Watson, David Rolnick

Abstract: Climate simulations are essential in guiding our understanding of climate change and responding to its effects. However, it is computationally expensive to resolve complex climate processes at high spatial resolution. As one way to speed up climate simulations, neural networks have been used to downscale climate variables from fast-running low-resolution simulations, but high-resolution training d… ▽ More Climate simulations are essential in guiding our understanding of climate change and responding to its effects. However, it is computationally expensive to resolve complex climate processes at high spatial resolution. As one way to speed up climate simulations, neural networks have been used to downscale climate variables from fast-running low-resolution simulations, but high-resolution training data are often unobtainable or scarce, greatly limiting accuracy. In this work, we propose a downscaling method based on the Fourier neural operator. It trains with data of a small upsampling factor and then can zero-shot downscale its input to arbitrary unseen high resolution. Evaluated both on ERA5 climate model data and on the Navier-Stokes equation solution data, our downscaling model significantly outperforms state-of-the-art convolutional and generative adversarial downscaling models, both in standard single-resolution downscaling and in zero-shot generalization to higher upsampling factors. Furthermore, we show that our method also outperforms state-of-the-art data-driven partial differential equation solvers on Navier-Stokes equations. Overall, our work bridges the gap between simulation of a physical process and interpolation of low-resolution output, showing that it is possible to combine both approaches and significantly improve upon each other. △ Less

Submitted 30 May, 2023; v1 submitted 23 May, 2023; originally announced May 2023.

Comments: Presented at the ICLR 2023 workshop on "Tackling Climate Change with Machine Learning"

Pyrocast: a Machine Learning Pipeline to Forecast Pyrocumulonimbus (PyroCb) Clouds

Authors: Kenza Tazi, Emiliano Díaz Salas-Porras, Ashwin Braude, Daniel Okoh, Kara D. Lamb, Duncan Watson-Parris, Paula Harder, Nis Meinert

Abstract: Pyrocumulonimbus (pyroCb) clouds are storm clouds generated by extreme wildfires. PyroCbs are associated with unpredictable, and therefore dangerous, wildfire spread. They can also inject smoke particles and trace gases into the upper troposphere and lower stratosphere, affecting the Earth's climate. As global temperatures increase, these previously rare events are becoming more common. Being able… ▽ More Pyrocumulonimbus (pyroCb) clouds are storm clouds generated by extreme wildfires. PyroCbs are associated with unpredictable, and therefore dangerous, wildfire spread. They can also inject smoke particles and trace gases into the upper troposphere and lower stratosphere, affecting the Earth's climate. As global temperatures increase, these previously rare events are becoming more common. Being able to predict which fires are likely to generate pyroCb is therefore key to climate adaptation in wildfire-prone areas. This paper introduces Pyrocast, a pipeline for pyroCb analysis and forecasting. The pipeline's first two components, a pyroCb database and a pyroCb forecast model, are presented. The database brings together geostationary imagery and environmental data for over 148 pyroCb events across North America, Australia, and Russia between 2018 and 2022. Random Forests, Convolutional Neural Networks (CNNs), and CNNs pretrained with Auto-Encoders were tested to predict the generation of pyroCb for a given fire six hours in advance. The best model predicted pyroCb with an AUC of $0.90 \pm 0.04$. △ Less

Submitted 22 November, 2022; originally announced November 2022.

Comments: 5 pages, 2 figures, Tackling Climate Change with Machine Learning: workshop at NeurIPS 2022

Identifying the Causes of Pyrocumulonimbus (PyroCb)

Authors: Emiliano Díaz Salas-Porras, Kenza Tazi, Ashwin Braude, Daniel Okoh, Kara D. Lamb, Duncan Watson-Parris, Paula Harder, Nis Meinert

Abstract: A first causal discovery analysis from observational data of pyroCb (storm clouds generated from extreme wildfires) is presented. Invariant Causal Prediction was used to develop tools to understand the causal drivers of pyroCb formation. This includes a conditional independence test for testing $Y$ conditionally independent of $E$ given $X$ for binary variable $Y$ and multivariate, continuous vari… ▽ More A first causal discovery analysis from observational data of pyroCb (storm clouds generated from extreme wildfires) is presented. Invariant Causal Prediction was used to develop tools to understand the causal drivers of pyroCb formation. This includes a conditional independence test for testing $Y$ conditionally independent of $E$ given $X$ for binary variable $Y$ and multivariate, continuous variables $X$ and $E$, and a greedy-ICP search algorithm that relies on fewer conditional independence tests to obtain a smaller more manageable set of causal predictors. With these tools, we identified a subset of seven causal predictors which are plausible when contrasted with domain knowledge: surface sensible heat flux, relative humidity at $850$ hPa, a component of wind at $250$ hPa, $13.3$ micro-meters, thermal emissions, convective available potential energy, and altitude. △ Less

Submitted 18 November, 2022; v1 submitted 16 November, 2022; originally announced November 2022.

Comments: 14 pages 9 figures. To be published in the 2022 NeurIPS Workshop on Causal Machine Learning for Real-World Impact

Hard-Constrained Deep Learning for Climate Downscaling

Authors: Paula Harder, Alex Hernandez-Garcia, Venkatesh Ramesh, Qidong Yang, Prasanna Sattigeri, Daniela Szwarcman, Campbell Watson, David Rolnick

Abstract: The availability of reliable, high-resolution climate and weather data is important to inform long-term decisions on climate adaptation and mitigation and to guide rapid responses to extreme events. Forecasting models are limited by computational costs and, therefore, often generate coarse-resolution predictions. Statistical downscaling, including super-resolution methods from deep learning, can p… ▽ More The availability of reliable, high-resolution climate and weather data is important to inform long-term decisions on climate adaptation and mitigation and to guide rapid responses to extreme events. Forecasting models are limited by computational costs and, therefore, often generate coarse-resolution predictions. Statistical downscaling, including super-resolution methods from deep learning, can provide an efficient method of upsampling low-resolution data. However, despite achieving visually compelling results in some cases, such models frequently violate conservation laws when predicting physical variables. In order to conserve physical quantities, here we introduce methods that guarantee statistical constraints are satisfied by a deep learning downscaling model, while also improving their performance according to traditional metrics. We compare different constraining approaches and demonstrate their applicability across different neural architectures as well as a variety of climate and weather data sets. Besides enabling faster and more accurate climate predictions through downscaling, we also show that our novel methodologies can improve super-resolution for satellite data and natural images data sets. △ Less

Submitted 29 February, 2024; v1 submitted 8 August, 2022; originally announced August 2022.

Abstract error analysis for Cahn--Hilliard type equations with dynamic boundary conditions

Authors: Paula Harder

Abstract: This work addresses the problem of solving the Cahn-Hilliard equation numerically. For that we introduce an abstract formulation for Cahn-Hilliard type equations with dynamic boundary conditions, we conduct the spatial semidiscretization via finite elements and prove error bounds based on the technique of energy estimates. The variational formulation for Cahn-Hilliard/Cahn-Hilliard coupling, will… ▽ More This work addresses the problem of solving the Cahn-Hilliard equation numerically. For that we introduce an abstract formulation for Cahn-Hilliard type equations with dynamic boundary conditions, we conduct the spatial semidiscretization via finite elements and prove error bounds based on the technique of energy estimates. The variational formulation for Cahn-Hilliard/Cahn-Hilliard coupling, will apply to a larger abstract class of problems and is similar to the usual weak formulation of parabolic problems. In contrast to problems with non dynamic boundary conditions, the Hilbert spaces $L^2(Ω)$ and $H^1(Ω)$ are exchanged with the spaces $L^2(Ω)\times L^2(Γ)$ and $\lbrace v\in H^1(Ω): γv \in H^1(Γ)\rbrace$, respectively. Because we are considering a fourth-order differential equation, which will be described by a system of two second-order differential equations, the variational formulation also consists of a system of two equations. △ Less

Submitted 8 August, 2022; originally announced August 2022.

Comments: Masterthesis

Physics-Informed Learning of Aerosol Microphysics

Authors: Paula Harder, Duncan Watson-Parris, Philip Stier, Dominik Strassel, Nicolas R. Gauger, Janis Keuper

Abstract: Aerosol particles play an important role in the climate system by absorbing and scattering radiation and influencing cloud properties. They are also one of the biggest sources of uncertainty for climate modeling. Many climate models do not include aerosols in sufficient detail due to computational constraints. In order to represent key processes, aerosol microphysical properties and processes have… ▽ More Aerosol particles play an important role in the climate system by absorbing and scattering radiation and influencing cloud properties. They are also one of the biggest sources of uncertainty for climate modeling. Many climate models do not include aerosols in sufficient detail due to computational constraints. In order to represent key processes, aerosol microphysical properties and processes have to be accounted for. This is done in the ECHAM-HAM global climate aerosol model using the M7 microphysics, but high computational costs make it very expensive to run with finer resolution or for a longer time. We aim to use machine learning to emulate the microphysics model at sufficient accuracy and reduce the computational cost by being fast at inference time. The original M7 model is used to generate data of input-output pairs to train a neural network on it. We are able to learn the variables' tendencies achieving an average $R^2$ score of $77.1\% $. We further explore methods to inform and constrain the neural network with physical knowledge to reduce mass violation and enforce mass positivity. On a GPU we achieve a speed-up of up to over 64x compared to the original model. △ Less

Submitted 24 July, 2022; originally announced July 2022.

Comments: arXiv admin note: text overlap with arXiv:2109.10593

Detecting AutoAttack Perturbations in the Frequency Domain

Authors: Peter Lorenz, Paula Harder, Dominik Strassel, Margret Keuper, Janis Keuper

Abstract: Recently, adversarial attacks on image classification networks by the AutoAttack (Croce and Hein, 2020b) framework have drawn a lot of attention. While AutoAttack has shown a very high attack success rate, most defense approaches are focusing on network hardening and robustness enhancements, like adversarial training. This way, the currently best-reported method can withstand about 66% of adversar… ▽ More Recently, adversarial attacks on image classification networks by the AutoAttack (Croce and Hein, 2020b) framework have drawn a lot of attention. While AutoAttack has shown a very high attack success rate, most defense approaches are focusing on network hardening and robustness enhancements, like adversarial training. This way, the currently best-reported method can withstand about 66% of adversarial examples on CIFAR10. In this paper, we investigate the spatial and frequency domain properties of AutoAttack and propose an alternative defense. Instead of hardening a network, we detect adversarial attacks during inference, rejecting manipulated inputs. Based on a rather simple and fast analysis in the frequency domain, we introduce two different detection algorithms. First, a black box detector that only operates on the input images and achieves a detection accuracy of 100% on the AutoAttack CIFAR10 benchmark and 99.3% on ImageNet, for epsilon = 8/255 in both cases. Second, a whitebox detector using an analysis of CNN feature maps, leading to a detection rate of also 100% and 98.7% on the same benchmarks. △ Less

Submitted 20 February, 2024; v1 submitted 16 November, 2021; originally announced November 2021.

Comments: accepted at ICML 2021 workshop for robustness

Emulating Aerosol Microphysics with Machine Learning

Authors: Paula Harder, Duncan Watson-Parris, Dominik Strassel, Nicolas Gauger, Philip Stier, Janis Keuper

Abstract: Aerosol particles play an important role in the climate system by absorbing and scattering radiation and influencing cloud properties. They are also one of the biggest sources of uncertainty for climate modeling. Many climate models do not include aerosols in sufficient detail. In order to achieve higher accuracy, aerosol microphysical properties and processes have to be accounted for. This is don… ▽ More Aerosol particles play an important role in the climate system by absorbing and scattering radiation and influencing cloud properties. They are also one of the biggest sources of uncertainty for climate modeling. Many climate models do not include aerosols in sufficient detail. In order to achieve higher accuracy, aerosol microphysical properties and processes have to be accounted for. This is done in the ECHAM-HAM global climate aerosol model using the M7 microphysics model, but increased computational costs make it very expensive to run at higher resolutions or for a longer time. We aim to use machine learning to approximate the microphysics model at sufficient accuracy and reduce the computational cost by being fast at inference time. The original M7 model is used to generate data of input-output pairs to train a neural network on it. By using a special logarithmic transform we are able to learn the variables tendencies achieving an average $R^2$ score of $89\%$. On a GPU we achieve a speed-up of 120 compared to the original model. △ Less

Submitted 29 September, 2021; v1 submitted 22 September, 2021; originally announced September 2021.

SpectralDefense: Detecting Adversarial Attacks on CNNs in the Fourier Domain

Authors: Paula Harder, Franz-Josef Pfreundt, Margret Keuper, Janis Keuper

Abstract: Despite the success of convolutional neural networks (CNNs) in many computer vision and image analysis tasks, they remain vulnerable against so-called adversarial attacks: Small, crafted perturbations in the input images can lead to false predictions. A possible defense is to detect adversarial examples. In this work, we show how analysis in the Fourier domain of input images and feature maps can… ▽ More Despite the success of convolutional neural networks (CNNs) in many computer vision and image analysis tasks, they remain vulnerable against so-called adversarial attacks: Small, crafted perturbations in the input images can lead to false predictions. A possible defense is to detect adversarial examples. In this work, we show how analysis in the Fourier domain of input images and feature maps can be used to distinguish benign test samples from adversarial images. We propose two novel detection methods: Our first method employs the magnitude spectrum of the input images to detect an adversarial attack. This simple and robust classifier can successfully detect adversarial perturbations of three commonly used attack methods. The second method builds upon the first and additionally extracts the phase of Fourier coefficients of feature-maps at different layers of the network. With this extension, we are able to improve adversarial detection rates compared to state-of-the-art detectors on five different attack methods. △ Less

Submitted 2 June, 2021; v1 submitted 4 March, 2021; originally announced March 2021.

NightVision: Generating Nighttime Satellite Imagery from Infra-Red Observations

Authors: Paula Harder, William Jones, Redouane Lguensat, Shahine Bouabid, James Fulton, Dánell Quesada-Chacón, Aris Marcolongo, Sofija Stefanović, Yuhan Rao, Peter Manshausen, Duncan Watson-Parris

Abstract: The recent explosion in applications of machine learning to satellite imagery often rely on visible images and therefore suffer from a lack of data during the night. The gap can be filled by employing available infra-red observations to generate visible images. This work presents how deep learning can be applied successfully to create those images by using U-Net based architectures. The proposed m… ▽ More The recent explosion in applications of machine learning to satellite imagery often rely on visible images and therefore suffer from a lack of data during the night. The gap can be filled by employing available infra-red observations to generate visible images. This work presents how deep learning can be applied successfully to create those images by using U-Net based architectures. The proposed methods show promising results, achieving a structural similarity index (SSIM) up to 86\% on an independent test set and providing visually convincing output images, generated from infra-red observations. △ Less

Submitted 8 December, 2020; v1 submitted 13 November, 2020; originally announced November 2020.

Error estimates for the Cahn--Hilliard equation with dynamic boundary conditions

Authors: Paula Harder, Balázs Kovács

Abstract: A proof of convergence is given for bulk--surface finite element semi-discretisation of the Cahn--Hilliard equation with Cahn--Hilliard-type dynamic boundary conditions in a smooth domain. The semi-discretisation is studied in the weak formulation as a second order system. Optimal-order uniform-in-time error estimates are shown in the $L^2$ and $H^1$ norms. The error estimates are based on a consi… ▽ More A proof of convergence is given for bulk--surface finite element semi-discretisation of the Cahn--Hilliard equation with Cahn--Hilliard-type dynamic boundary conditions in a smooth domain. The semi-discretisation is studied in the weak formulation as a second order system. Optimal-order uniform-in-time error estimates are shown in the $L^2$ and $H^1$ norms. The error estimates are based on a consistency and stability analysis. The proof of stability is performed in an abstract framework, based on energy estimates exploiting the anti-symmetric structure of the second order system. Numerical experiments illustrate the theoretical results. △ Less

Submitted 30 November, 2020; v1 submitted 7 May, 2020; originally announced May 2020.

Influence of interface-driven strain on the spectral diffusion properties of core/shell CdSe/CdS dot/rod nanoparticles

Authors: Sven-Hendrik Lohmann, Philip Harder, Felix Bourier, Christian Strelow, Alf Mews, Tobias Kipp

Abstract: By combining an atomistic valence-force field approach and calculations based on the effective-mass approximation we investigate the influence of strain effects on the band alignment and general excitonic properties of core/shell CdSe/CdS dot/rod nanoparticles. We find that the inclusion of strain effects leads to an increase in exciton energy as well as to a decrease in electron and hole wave fun… ▽ More By combining an atomistic valence-force field approach and calculations based on the effective-mass approximation we investigate the influence of strain effects on the band alignment and general excitonic properties of core/shell CdSe/CdS dot/rod nanoparticles. We find that the inclusion of strain effects leads to an increase in exciton energy as well as to a decrease in electron and hole wave function overlap. Importantly, the native type-I band alignment of the CdSe/CdS material system is preserved and does not change into an quasi-type-II or even type-II band offset for the nanoparticles. Furthermore, we analyze the impact of strain on a model in which the spectral diffusion of the fluorescence emission of these nanoparticles is explained by migrating surface charges. Our calculations show that the addition of strain effects leads to increased energy shifts as well as larger changes in the squared electron and hole wave function overlap, while the correlation of both also exhibits a steeper slope than for the unstrained system. For a given CdSe core size, an increase in CdS-shell thickness decreases the possible ranges of energy shift and squared wave function overlap without changing the slope of their correlation. On the other hand, for a given nanoparticle overall thickness, dot/rod systems with a small CdSe core exhibit the strongest influenceability by surface charges. △ Less

Submitted 22 February, 2019; v1 submitted 25 July, 2018; originally announced July 2018.

Comments: This document is the unedited Author's version of a Submitted Work that was subsequently accepted for publication in The Journal of Physical Chemistry C, copyright ©American Chemical Society after peer review. To access the final edited and published work see https://pubs.acs.org/doi/full/10.1021/acs.jpcc.8b12253