Search Results
You are looking at 1 - 10 of 15 items for
- Author or Editor: Imme Ebert-Uphoff x
- Refine by Access: All Content x
Abstract
The method of neural networks (aka deep learning) has opened up many new opportunities to utilize remotely sensed images in meteorology. Common applications include image classification, e.g., to determine whether an image contains a tropical cyclone, and image-to-image translation, e.g., to emulate radar imagery for satellites that only have passive channels. However, there are yet many open questions regarding the use of neural networks for working with meteorological images, such as best practices for evaluation, tuning, and interpretation. This article highlights several strategies and practical considerations for neural network development that have not yet received much attention in the meteorological community, such as the concept of receptive fields, underutilized meteorological performance measures, and methods for neural network interpretation, such as synthetic experiments and layer-wise relevance propagation. We also consider the process of neural network interpretation as a whole, recognizing it as an iterative meteorologist-driven discovery process that builds on experimental design and hypothesis generation and testing. Finally, while most work on neural network interpretation in meteorology has so far focused on networks for image classification tasks, we expand the focus to also include networks for image-to-image translation.
Abstract
The method of neural networks (aka deep learning) has opened up many new opportunities to utilize remotely sensed images in meteorology. Common applications include image classification, e.g., to determine whether an image contains a tropical cyclone, and image-to-image translation, e.g., to emulate radar imagery for satellites that only have passive channels. However, there are yet many open questions regarding the use of neural networks for working with meteorological images, such as best practices for evaluation, tuning, and interpretation. This article highlights several strategies and practical considerations for neural network development that have not yet received much attention in the meteorological community, such as the concept of receptive fields, underutilized meteorological performance measures, and methods for neural network interpretation, such as synthetic experiments and layer-wise relevance propagation. We also consider the process of neural network interpretation as a whole, recognizing it as an iterative meteorologist-driven discovery process that builds on experimental design and hypothesis generation and testing. Finally, while most work on neural network interpretation in meteorology has so far focused on networks for image classification tasks, we expand the focus to also include networks for image-to-image translation.
Abstract
In the last decade, much work in atmospheric science has focused on spatial verification (SV) methods for gridded prediction, which overcome serious disadvantages of pixelwise verification. However, neural networks (NN) in atmospheric science are almost always trained to optimize pixelwise loss functions, even when ultimately assessed with SV methods. This establishes a disconnect between model verification during versus after training. To address this issue, we develop spatially enhanced loss functions (SELF) and demonstrate their use for a real-world problem: predicting the occurrence of thunderstorms (henceforth, “convection”) with NNs. In each SELF we use either a neighborhood filter, which highlights convection at scales larger than a threshold, or a spectral filter (employing Fourier or wavelet decomposition), which is more flexible and highlights convection at scales between two thresholds. We use these filters to spatially enhance common verification scores, such as the Brier score. We train each NN with a different SELF and compare their performance at many scales of convection, from discrete storm cells to tropical cyclones. Among our many findings are that (i) for a low or high risk threshold, the ideal SELF focuses on small or large scales, respectively; (ii) models trained with a pixelwise loss function perform surprisingly well; and (iii) nevertheless, models trained with a spectral filter produce much better-calibrated probabilities than a pixelwise model. We provide a general guide to using SELFs, including technical challenges and the final Python code, as well as demonstrating their use for the convection problem. To our knowledge this is the most in-depth guide to SELFs in the geosciences.
Significance Statement
Gridded predictions, in which a quantity is predicted at every pixel in space, should be verified with spatially aware methods rather than pixel by pixel. Neural networks (NN), which are often used for gridded prediction, are trained to minimize an error value called the loss function. NN loss functions in atmospheric science are almost always pixelwise, which causes the predictions to miss rare events and contain unrealistic spatial patterns. We use spatial filters to enhance NN loss functions, and we test our novel spatially enhanced loss functions (SELF) on thunderstorm prediction. We find that different SELFs work better for different scales (i.e., different-sized thunderstorm complexes) and that spectral filters, one of the two filter types, produce unexpectedly well calibrated thunderstorm probabilities.
Abstract
In the last decade, much work in atmospheric science has focused on spatial verification (SV) methods for gridded prediction, which overcome serious disadvantages of pixelwise verification. However, neural networks (NN) in atmospheric science are almost always trained to optimize pixelwise loss functions, even when ultimately assessed with SV methods. This establishes a disconnect between model verification during versus after training. To address this issue, we develop spatially enhanced loss functions (SELF) and demonstrate their use for a real-world problem: predicting the occurrence of thunderstorms (henceforth, “convection”) with NNs. In each SELF we use either a neighborhood filter, which highlights convection at scales larger than a threshold, or a spectral filter (employing Fourier or wavelet decomposition), which is more flexible and highlights convection at scales between two thresholds. We use these filters to spatially enhance common verification scores, such as the Brier score. We train each NN with a different SELF and compare their performance at many scales of convection, from discrete storm cells to tropical cyclones. Among our many findings are that (i) for a low or high risk threshold, the ideal SELF focuses on small or large scales, respectively; (ii) models trained with a pixelwise loss function perform surprisingly well; and (iii) nevertheless, models trained with a spectral filter produce much better-calibrated probabilities than a pixelwise model. We provide a general guide to using SELFs, including technical challenges and the final Python code, as well as demonstrating their use for the convection problem. To our knowledge this is the most in-depth guide to SELFs in the geosciences.
Significance Statement
Gridded predictions, in which a quantity is predicted at every pixel in space, should be verified with spatially aware methods rather than pixel by pixel. Neural networks (NN), which are often used for gridded prediction, are trained to minimize an error value called the loss function. NN loss functions in atmospheric science are almost always pixelwise, which causes the predictions to miss rare events and contain unrealistic spatial patterns. We use spatial filters to enhance NN loss functions, and we test our novel spatially enhanced loss functions (SELF) on thunderstorm prediction. We find that different SELFs work better for different scales (i.e., different-sized thunderstorm complexes) and that spectral filters, one of the two filter types, produce unexpectedly well calibrated thunderstorm probabilities.
Abstract
Causal discovery seeks to recover cause–effect relationships from statistical data using graphical models. One goal of this paper is to provide an accessible introduction to causal discovery methods for climate scientists, with a focus on constraint-based structure learning. Second, in a detailed case study constraint-based structure learning is applied to derive hypotheses of causal relationships between four prominent modes of atmospheric low-frequency variability in boreal winter including the Western Pacific Oscillation (WPO), Eastern Pacific Oscillation (EPO), Pacific–North America (PNA) pattern, and North Atlantic Oscillation (NAO). The results are shown in the form of static and temporal independence graphs also known as Bayesian Networks. It is found that WPO and EPO are nearly indistinguishable from the cause–effect perspective as strong simultaneous coupling is identified between the two. In addition, changes in the state of EPO (NAO) may cause changes in the state of NAO (PNA) approximately 18 (3–6) days later. These results are not only consistent with previous findings on dynamical processes connecting different low-frequency modes (e.g., interaction between synoptic and low-frequency eddies) but also provide the basis for formulating new hypotheses regarding the time scale and temporal sequencing of dynamical processes responsible for these connections. Last, the authors propose to use structure learning for climate networks, which are currently based primarily on correlation analysis. While correlation-based climate networks focus on similarity between nodes, independence graphs would provide an alternative viewpoint by focusing on information flow in the network.
Abstract
Causal discovery seeks to recover cause–effect relationships from statistical data using graphical models. One goal of this paper is to provide an accessible introduction to causal discovery methods for climate scientists, with a focus on constraint-based structure learning. Second, in a detailed case study constraint-based structure learning is applied to derive hypotheses of causal relationships between four prominent modes of atmospheric low-frequency variability in boreal winter including the Western Pacific Oscillation (WPO), Eastern Pacific Oscillation (EPO), Pacific–North America (PNA) pattern, and North Atlantic Oscillation (NAO). The results are shown in the form of static and temporal independence graphs also known as Bayesian Networks. It is found that WPO and EPO are nearly indistinguishable from the cause–effect perspective as strong simultaneous coupling is identified between the two. In addition, changes in the state of EPO (NAO) may cause changes in the state of NAO (PNA) approximately 18 (3–6) days later. These results are not only consistent with previous findings on dynamical processes connecting different low-frequency modes (e.g., interaction between synoptic and low-frequency eddies) but also provide the basis for formulating new hypotheses regarding the time scale and temporal sequencing of dynamical processes responsible for these connections. Last, the authors propose to use structure learning for climate networks, which are currently based primarily on correlation analysis. While correlation-based climate networks focus on similarity between nodes, independence graphs would provide an alternative viewpoint by focusing on information flow in the network.
Abstract
Convolutional neural networks (CNNs) have recently attracted great attention in geoscience because of their ability to capture nonlinear system behavior and extract predictive spatiotemporal patterns. Given their black-box nature, however, and the importance of prediction explainability, methods of explainable artificial intelligence (XAI) are gaining popularity as a means to explain the CNN decision-making strategy. Here, we establish an intercomparison of some of the most popular XAI methods and investigate their fidelity in explaining CNN decisions for geoscientific applications. Our goal is to raise awareness of the theoretical limitations of these methods and to gain insight into the relative strengths and weaknesses to help guide best practices. The considered XAI methods are first applied to an idealized attribution benchmark, in which the ground truth of explanation of the network is known a priori, to help objectively assess their performance. Second, we apply XAI to a climate-related prediction setting, namely, to explain a CNN that is trained to predict the number of atmospheric rivers in daily snapshots of climate simulations. Our results highlight several important issues of XAI methods (e.g., gradient shattering, inability to distinguish the sign of attribution, and ignorance to zero input) that have previously been overlooked in our field and, if not considered cautiously, may lead to a distorted picture of the CNN decision-making strategy. We envision that our analysis will motivate further investigation into XAI fidelity and will help toward a cautious implementation of XAI in geoscience, which can lead to further exploitation of CNNs and deep learning for prediction problems.
Abstract
Convolutional neural networks (CNNs) have recently attracted great attention in geoscience because of their ability to capture nonlinear system behavior and extract predictive spatiotemporal patterns. Given their black-box nature, however, and the importance of prediction explainability, methods of explainable artificial intelligence (XAI) are gaining popularity as a means to explain the CNN decision-making strategy. Here, we establish an intercomparison of some of the most popular XAI methods and investigate their fidelity in explaining CNN decisions for geoscientific applications. Our goal is to raise awareness of the theoretical limitations of these methods and to gain insight into the relative strengths and weaknesses to help guide best practices. The considered XAI methods are first applied to an idealized attribution benchmark, in which the ground truth of explanation of the network is known a priori, to help objectively assess their performance. Second, we apply XAI to a climate-related prediction setting, namely, to explain a CNN that is trained to predict the number of atmospheric rivers in daily snapshots of climate simulations. Our results highlight several important issues of XAI methods (e.g., gradient shattering, inability to distinguish the sign of attribution, and ignorance to zero input) that have previously been overlooked in our field and, if not considered cautiously, may lead to a distorted picture of the CNN decision-making strategy. We envision that our analysis will motivate further investigation into XAI fidelity and will help toward a cautious implementation of XAI in geoscience, which can lead to further exploitation of CNNs and deep learning for prediction problems.
Abstract
Methods of explainable artificial intelligence (XAI) are used in geoscientific applications to gain insights into the decision-making strategy of neural networks (NNs), highlighting which features in the input contribute the most to a NN prediction. Here, we discuss our “lesson learned” that the task of attributing a prediction to the input does not have a single solution. Instead, the attribution results depend greatly on the considered baseline that the XAI method utilizes—a fact that has been overlooked in the geoscientific literature. The baseline is a reference point to which the prediction is compared so that the prediction can be understood. This baseline can be chosen by the user or is set by construction in the method’s algorithm—often without the user being aware of that choice. We highlight that different baselines can lead to different insights for different science questions and, thus, should be chosen accordingly. To illustrate the impact of the baseline, we use a large ensemble of historical and future climate simulations forced with the shared socioeconomic pathway 3-7.0 (SSP3-7.0) scenario and train a fully connected NN to predict the ensemble- and global-mean temperature (i.e., the forced global warming signal) given an annual temperature map from an individual ensemble member. We then use various XAI methods and different baselines to attribute the network predictions to the input. We show that attributions differ substantially when considering different baselines, because they correspond to answering different science questions. We conclude by discussing important implications and considerations about the use of baselines in XAI research.
Significance Statement
In recent years, methods of explainable artificial intelligence (XAI) have found great application in geoscientific applications, because they can be used to attribute the predictions of neural networks (NNs) to the input and interpret them physically. Here, we highlight that the attributions—and the physical interpretation—depend greatly on the choice of the baseline—a fact that has been overlooked in the geoscientific literature. We illustrate this dependence for a specific climate task, in which a NN is trained to predict the ensemble- and global-mean temperature (i.e., the forced global warming signal) given an annual temperature map from an individual ensemble member. We show that attributions differ substantially when considering different baselines, because they correspond to answering different science questions.
Abstract
Methods of explainable artificial intelligence (XAI) are used in geoscientific applications to gain insights into the decision-making strategy of neural networks (NNs), highlighting which features in the input contribute the most to a NN prediction. Here, we discuss our “lesson learned” that the task of attributing a prediction to the input does not have a single solution. Instead, the attribution results depend greatly on the considered baseline that the XAI method utilizes—a fact that has been overlooked in the geoscientific literature. The baseline is a reference point to which the prediction is compared so that the prediction can be understood. This baseline can be chosen by the user or is set by construction in the method’s algorithm—often without the user being aware of that choice. We highlight that different baselines can lead to different insights for different science questions and, thus, should be chosen accordingly. To illustrate the impact of the baseline, we use a large ensemble of historical and future climate simulations forced with the shared socioeconomic pathway 3-7.0 (SSP3-7.0) scenario and train a fully connected NN to predict the ensemble- and global-mean temperature (i.e., the forced global warming signal) given an annual temperature map from an individual ensemble member. We then use various XAI methods and different baselines to attribute the network predictions to the input. We show that attributions differ substantially when considering different baselines, because they correspond to answering different science questions. We conclude by discussing important implications and considerations about the use of baselines in XAI research.
Significance Statement
In recent years, methods of explainable artificial intelligence (XAI) have found great application in geoscientific applications, because they can be used to attribute the predictions of neural networks (NNs) to the input and interpret them physically. Here, we highlight that the attributions—and the physical interpretation—depend greatly on the choice of the baseline—a fact that has been overlooked in the geoscientific literature. We illustrate this dependence for a specific climate task, in which a NN is trained to predict the ensemble- and global-mean temperature (i.e., the forced global warming signal) given an annual temperature map from an individual ensemble member. We show that attributions differ substantially when considering different baselines, because they correspond to answering different science questions.
Abstract
This paper reports preliminary yet encouraging findings on the use of causal discovery methods to understand the interaction between atmospheric planetary- and synoptic-scale disturbances in the Northern Hemisphere. Specifically, constraint-based structure learning of probabilistic graphical models is applied to the spherical harmonics decomposition of the daily 500-hPa geopotential height field in boreal winter for the period 1948–2015. Active causal pathways among different spherical harmonics components are identified and documented in the form of a temporal probabilistic graphical model. Since, by definition, the structure learning algorithm used here only robustly identifies linear causal effects, we report only causal pathways between two groups of disturbances with sufficiently large differences in temporal and/or spatial scales, that is, planetary-scale (mainly zonal wavenumbers 1–3) and synoptic-scale disturbances (mainly zonal wavenumbers 6–8). Daily reconstruction of geopotential heights using only interacting scales suggest that the modulation of synoptic-scale disturbances by planetary-scale disturbances is best characterized by the flow of information from a zonal wavenumber-1 disturbance to a synoptic-scale circumglobal wave train whose amplitude peaks at the North Pacific and North Atlantic storm-track region. The feedback of synoptic-scale to planetary-scale disturbances manifests itself as a zonal wavenumber-2 structure driven by synoptic-eddy momentum fluxes. This wavenumber-2 structure locally enhances the East Asian trough and western Europe ridge of the wavenumber-1 planetary-scale disturbance that actively modulates the activity of synoptic-scale disturbances. The winter-mean amplitude of the actively interacting disturbances are characterized by pronounced fluctuations across interannual to decadal time scales.
Abstract
This paper reports preliminary yet encouraging findings on the use of causal discovery methods to understand the interaction between atmospheric planetary- and synoptic-scale disturbances in the Northern Hemisphere. Specifically, constraint-based structure learning of probabilistic graphical models is applied to the spherical harmonics decomposition of the daily 500-hPa geopotential height field in boreal winter for the period 1948–2015. Active causal pathways among different spherical harmonics components are identified and documented in the form of a temporal probabilistic graphical model. Since, by definition, the structure learning algorithm used here only robustly identifies linear causal effects, we report only causal pathways between two groups of disturbances with sufficiently large differences in temporal and/or spatial scales, that is, planetary-scale (mainly zonal wavenumbers 1–3) and synoptic-scale disturbances (mainly zonal wavenumbers 6–8). Daily reconstruction of geopotential heights using only interacting scales suggest that the modulation of synoptic-scale disturbances by planetary-scale disturbances is best characterized by the flow of information from a zonal wavenumber-1 disturbance to a synoptic-scale circumglobal wave train whose amplitude peaks at the North Pacific and North Atlantic storm-track region. The feedback of synoptic-scale to planetary-scale disturbances manifests itself as a zonal wavenumber-2 structure driven by synoptic-eddy momentum fluxes. This wavenumber-2 structure locally enhances the East Asian trough and western Europe ridge of the wavenumber-1 planetary-scale disturbance that actively modulates the activity of synoptic-scale disturbances. The winter-mean amplitude of the actively interacting disturbances are characterized by pronounced fluctuations across interannual to decadal time scales.
Abstract
Predicting the timing and location of thunderstorms (“convection”) allows for preventive actions that can save both lives and property. We have applied U-nets, a deep-learning-based type of neural network, to forecast convection on a grid at lead times up to 120 min. The goal is to make skillful forecasts with only present and past satellite data as predictors. Specifically, predictors are multispectral brightness-temperature images from the Himawari-8 satellite, while targets (ground truth) are provided by weather radars in Taiwan. U-nets are becoming popular in atmospheric science due to their advantages for gridded prediction. Furthermore, we use three novel approaches to advance U-nets in atmospheric science. First, we compare three architectures—vanilla, temporal, and U-net++—and find that vanilla U-nets are best for this task. Second, we train U-nets with the fractions skill score, which is spatially aware, as the loss function. Third, because we do not have adequate ground truth over the full Himawari-8 domain, we train the U-nets with small radar-centered patches, then apply trained U-nets to the full domain. Also, we find that the best predictions are given by U-nets trained with satellite data from multiple lag times, not only the present. We evaluate U-nets in detail—by time of day, month, and geographic location—and compare them to persistence models. The U-nets outperform persistence at lead times ≥ 60 min, and at all lead times the U-nets provide a more realistic climatology than persistence. Our code is available publicly.
Abstract
Predicting the timing and location of thunderstorms (“convection”) allows for preventive actions that can save both lives and property. We have applied U-nets, a deep-learning-based type of neural network, to forecast convection on a grid at lead times up to 120 min. The goal is to make skillful forecasts with only present and past satellite data as predictors. Specifically, predictors are multispectral brightness-temperature images from the Himawari-8 satellite, while targets (ground truth) are provided by weather radars in Taiwan. U-nets are becoming popular in atmospheric science due to their advantages for gridded prediction. Furthermore, we use three novel approaches to advance U-nets in atmospheric science. First, we compare three architectures—vanilla, temporal, and U-net++—and find that vanilla U-nets are best for this task. Second, we train U-nets with the fractions skill score, which is spatially aware, as the loss function. Third, because we do not have adequate ground truth over the full Himawari-8 domain, we train the U-nets with small radar-centered patches, then apply trained U-nets to the full domain. Also, we find that the best predictions are given by U-nets trained with satellite data from multiple lag times, not only the present. We evaluate U-nets in detail—by time of day, month, and geographic location—and compare them to persistence models. The U-nets outperform persistence at lead times ≥ 60 min, and at all lead times the U-nets provide a more realistic climatology than persistence. Our code is available publicly.
Using Deep Learning to Emulate and Accelerate a Radiative Transfer ModelAbstract
This paper describes the development of U-net++ models, a type of neural network that performs deep learning, to emulate the shortwave Rapid Radiative Transfer Model (RRTM). The goal is to emulate the RRTM accurately in a small fraction of the computing time, creating a U-net++ that could be used as a parameterization in numerical weather prediction (NWP). Target variables are surface downwelling flux, top-of-atmosphere upwelling flux (
Abstract
This paper describes the development of U-net++ models, a type of neural network that performs deep learning, to emulate the shortwave Rapid Radiative Transfer Model (RRTM). The goal is to emulate the RRTM accurately in a small fraction of the computing time, creating a U-net++ that could be used as a parameterization in numerical weather prediction (NWP). Target variables are surface downwelling flux, top-of-atmosphere upwelling flux (
Abstract
Topological data analysis (TDA) is a tool from data science and mathematics that is beginning to make waves in environmental science. In this work, we seek to provide an intuitive and understandable introduction to a tool from TDA that is particularly useful for the analysis of imagery, namely, persistent homology. We briefly discuss the theoretical background but focus primarily on understanding the output of this tool and discussing what information it can glean. To this end, we frame our discussion around a guiding example of classifying satellite images from the sugar, fish, flower, and gravel dataset produced for the study of mesoscale organization of clouds by Rasp et al. We demonstrate how persistent homology and its vectorization, persistence landscapes, can be used in a workflow with a simple machine learning algorithm to obtain good results, and we explore in detail how we can explain this behavior in terms of image-level features. One of the core strengths of persistent homology is how interpretable it can be, so throughout this paper we discuss not just the patterns we find but why those results are to be expected given what we know about the theory of persistent homology. Our goal is that readers of this paper will leave with a better understanding of TDA and persistent homology, will be able to identify problems and datasets of their own for which persistent homology could be helpful, and will gain an understanding of the results they obtain from applying the included GitHub example code.
Significance Statement
Information such as the geometric structure and texture of image data can greatly support the inference of the physical state of an observed Earth system, for example, in remote sensing to determine whether wildfires are active or to identify local climate zones. Persistent homology is a branch of topological data analysis that allows one to extract such information in an interpretable way—unlike black-box methods like deep neural networks. The purpose of this paper is to explain in an intuitive manner what persistent homology is and how researchers in environmental science can use it to create interpretable models. We demonstrate the approach to identify certain cloud patterns from satellite imagery and find that the resulting model is indeed interpretable.
Abstract
Topological data analysis (TDA) is a tool from data science and mathematics that is beginning to make waves in environmental science. In this work, we seek to provide an intuitive and understandable introduction to a tool from TDA that is particularly useful for the analysis of imagery, namely, persistent homology. We briefly discuss the theoretical background but focus primarily on understanding the output of this tool and discussing what information it can glean. To this end, we frame our discussion around a guiding example of classifying satellite images from the sugar, fish, flower, and gravel dataset produced for the study of mesoscale organization of clouds by Rasp et al. We demonstrate how persistent homology and its vectorization, persistence landscapes, can be used in a workflow with a simple machine learning algorithm to obtain good results, and we explore in detail how we can explain this behavior in terms of image-level features. One of the core strengths of persistent homology is how interpretable it can be, so throughout this paper we discuss not just the patterns we find but why those results are to be expected given what we know about the theory of persistent homology. Our goal is that readers of this paper will leave with a better understanding of TDA and persistent homology, will be able to identify problems and datasets of their own for which persistent homology could be helpful, and will gain an understanding of the results they obtain from applying the included GitHub example code.
Significance Statement
Information such as the geometric structure and texture of image data can greatly support the inference of the physical state of an observed Earth system, for example, in remote sensing to determine whether wildfires are active or to identify local climate zones. Persistent homology is a branch of topological data analysis that allows one to extract such information in an interpretable way—unlike black-box methods like deep neural networks. The purpose of this paper is to explain in an intuitive manner what persistent homology is and how researchers in environmental science can use it to create interpretable models. We demonstrate the approach to identify certain cloud patterns from satellite imagery and find that the resulting model is indeed interpretable.
Abstract
The objective of this research is to develop techniques for assimilating GOES-R series observations in precipitating scenes for the purpose of improving short-term convective-scale forecasts of high-impact weather hazards. Whereas one approach is radiance assimilation, the information content of GOES-R radiances from its Advanced Baseline Imager saturates in precipitating scenes, and radiance assimilation does not make use of lightning observations from the GOES Lightning Mapper. Here, a convolutional neural network (CNN) is developed to transform GOES-R radiances and lightning into synthetic radar reflectivity fields to make use of existing radar assimilation techniques. We find that the ability of CNNs to utilize spatial context is essential for this application and offers breakthrough improvement in skill compared to traditional pixel-by-pixel based approaches. To understand the improved performance, we use a novel analysis method that combines several techniques, each providing different insights into the network’s reasoning. Channel-withholding experiments and spatial information–withholding experiments are used to show that the CNN achieves skill at high reflectivity values from the information content in radiance gradients and the presence of lightning. The attribution method, layerwise relevance propagation, demonstrates that the CNN uses radiance and lightning information synergistically, where lightning helps the CNN focus on which neighboring locations are most important. Synthetic inputs are used to quantify the sensitivity to radiance gradients, showing that sharper gradients produce a stronger response in predicted reflectivity. Lightning observations are found to be uniquely valuable for their ability to pinpoint locations of strong radar echoes.
Abstract
The objective of this research is to develop techniques for assimilating GOES-R series observations in precipitating scenes for the purpose of improving short-term convective-scale forecasts of high-impact weather hazards. Whereas one approach is radiance assimilation, the information content of GOES-R radiances from its Advanced Baseline Imager saturates in precipitating scenes, and radiance assimilation does not make use of lightning observations from the GOES Lightning Mapper. Here, a convolutional neural network (CNN) is developed to transform GOES-R radiances and lightning into synthetic radar reflectivity fields to make use of existing radar assimilation techniques. We find that the ability of CNNs to utilize spatial context is essential for this application and offers breakthrough improvement in skill compared to traditional pixel-by-pixel based approaches. To understand the improved performance, we use a novel analysis method that combines several techniques, each providing different insights into the network’s reasoning. Channel-withholding experiments and spatial information–withholding experiments are used to show that the CNN achieves skill at high reflectivity values from the information content in radiance gradients and the presence of lightning. The attribution method, layerwise relevance propagation, demonstrates that the CNN uses radiance and lightning information synergistically, where lightning helps the CNN focus on which neighboring locations are most important. Synthetic inputs are used to quantify the sensitivity to radiance gradients, showing that sharper gradients produce a stronger response in predicted reflectivity. Lightning observations are found to be uniquely valuable for their ability to pinpoint locations of strong radar echoes.
