Search

The definition of anomaly detection is the identification of an unexpected event. Real-time detection of extreme events such as wildfires, cyclones, or floods using satellite data has become crucial for disaster management. Although several earth-observing satellites provide information about disasters, satellites in the geostationary orbit provide data at intervals as frequent as every minute, effectively creating a video from space. There are many techniques that have been proposed to identify anomalies in surveillance videos; however, the available datasets do not have dynamic behavior, so we discuss an anomaly framework that can work on very high-frequency datasets to find very fast-moving anomalies. In this work, we present a diffusion model which does not need any motion component to capture the fast-moving anomalies and outperforms the other baseline methods.

Earth observing satellites carrying multi-spectral sensors are widely used to monitor the physical and biological states of the atmosphere, land, and oceans. These satellites have different vantage points above the earth and different spectral imaging bands resulting in inconsistent imagery from one to another. This presents challenges in building downstream applications. What if we could generate synthetic bands for existing satellites from the union of all domains? We tackle the problem of generating synthetic spectral imagery for multispectral sensors as an unsupervised image-to-image translation problem with partial labels and introduce a novel shared spectral reconstruction loss. Simulated experiments performed by dropping one or more spectral bands show that cross-domain reconstruction outperforms measurements obtained from a second vantage point. On a downstream cloud detection task, we show that generating synthetic bands with our model improves segmentation performance beyond our baseline. Our proposed approach enables synchronization of multispectral data and provides a basis for more homogeneous remote sensing datasets.

Recent progress in quantum algorithms and hardware indicates the potential importance of quantum computing in the near future. However, finding suitable application areas remains an active area of research. Quantum machine learning is touted as a potential approach to demonstrate quantum advantage within both the gate-model and the adiabatic schemes. For instance, the Quantum-assisted Variational Autoencoder has been proposed as a quantum enhancement to the discrete VAE. We extend on previous work and study the real-world applicability of a QVAE by presenting a proof-of-concept for similarity search in large-scale high-dimensional datasets. While exact and fast similarity search algorithms are available for low dimensional datasets, scaling to high-dimensional data is non-trivial. We show how to construct a space-efficient search index based on the latent space representation of a QVAE. Our experiments show a correlation between the Hamming distance in the embedded space and the Euclidean distance in the original space on the Moderate Resolution Imaging Spectroradiometer (MODIS) dataset. Further, we find real-world speedups compared to linear search and demonstrate memory-efficient scaling to half a billion data points.

Satellite image classification is a challenging problem that lies at the crossroads of remote sensing, computer vision, and machine learning. Due to the high variability inherent in satellite data, most of the current object classification approaches are not suitable for handling satellite datasets. The progress of satellite image analytics has also been inhibited by the lack of a single labeled high-resolution dataset with multiple class labels. In a preliminary version of this work, we introduced two new high resolution satellite imagery datasets (SAT-4 and SAT-6) and proposed DeepSat framework for classification based on "handcrafted" features and a deep belief network (DBN). The present paper is an extended version, we present an end-to-end framework leveraging an improved architecture that augments a convolutional neural network (CNN) with handcrafted features (instead of using DBN-based architecture) for classification. Our framework, having access to fused spatial information obtained from handcrafted features as well as CNN feature maps, have achieved accuracies of 99.90% and 99.84% respectively, on SAT-4 and SAT-6, surpassing all the other state-of-the-art results. A statistical analysis based on Distribution Separability Criterion substantiates the robustness of our approach in learning better representations for satellite imagery., Comment: This is an Accepted Manuscript of an article published by Taylor & Francis Group in Remote Sensing Letters. arXiv admin note: text overlap with arXiv:1509.03602

Numerical models based on physics represent the state-of-the-art in earth system modeling and comprise our best tools for generating insights and predictions. Despite rapid growth in computational power, the perceived need for higher model resolutions overwhelms the latest-generation computers, reducing the ability of modelers to generate simulations for understanding parameter sensitivities and characterizing variability and uncertainty. Thus, surrogate models are often developed to capture the essential attributes of the full-blown numerical models. Recent successes of machine learning methods, especially deep learning, across many disciplines offer the possibility that complex nonlinear connectionist representations may be able to capture the underlying complex structures and nonlinear processes in earth systems. A difficult test for deep learning-based emulation, which refers to function approximation of numerical models, is to understand whether they can be comparable to traditional forms of surrogate models in terms of computational efficiency while simultaneously reproducing model results in a credible manner. A deep learning emulation that passes this test may be expected to perform even better than simple models with respect to capturing complex processes and spatiotemporal dependencies. Here we examine, with a case study in satellite-based remote sensing, the hypothesis that deep learning approaches can credibly represent the simulations from a surrogate model with comparable computational efficiency. Our results are encouraging in that the deep learning emulation reproduces the results with acceptable accuracy and often even faster performance. We discuss the broader implications of our results in light of the pace of improvements in high-performance implementations of deep learning as well as the growing desire for higher-resolution simulations in the earth sciences., Comment: 6 pages, 8 figures, 4 tables

Applications of satellite data in areas such as weather tracking and modeling, ecosystem monitoring, wildfire detection, and land-cover change are heavily dependent on the trade-offs to spatial, spectral and temporal resolutions of observations. In weather tracking, high-frequency temporal observations are critical and used to improve forecasts, study severe events, and extract atmospheric motion, among others. However, while the current generation of geostationary satellites have hemispheric coverage at 10-15 minute intervals, higher temporal frequency observations are ideal for studying mesoscale severe weather events. In this work, we apply a task specific optical flow approach to temporal up-sampling using deep convolutional neural networks. We apply this technique to 16-bands of GOES-R/Advanced Baseline Imager mesoscale dataset to temporally enhance full disk hemispheric snapshots of different spatial resolutions from 15 minutes to 1 minute. Experiments show the effectiveness of task specific optical flow and multi-scale blocks for interpolating high-frequency severe weather events relative to bilinear and global optical flow baselines. Lastly, we demonstrate strong performance in capturing variability during a convective precipitation events.

Machine learning has proven to be useful in classification and segmentation of images. In this paper, we evaluate a training methodology for pixel-wise segmentation on high resolution satellite images using progressive growing of generative adversarial networks. We apply our model to segmenting building rooftops and compare these results to conventional methods for rooftop segmentation. We present our findings using the SpaceNet version 2 dataset. Progressive GAN training achieved a test accuracy of 93% compared to 89% for traditional GAN training., Comment: Accepted too and presented at DMESS 2018 as part of IEEE ICDM 2018

Deep Learning (DL) methods have been transforming computer vision with innovative adaptations to other domains including climate change. For DL to pervade Science and Engineering (S&E) applications where risk management is a core component, well-characterized uncertainty estimates must accompany predictions. However, S&E observations and model-simulations often follow heavily skewed distributions and are not well modeled with DL approaches, since they usually optimize a Gaussian, or Euclidean, likelihood loss. Recent developments in Bayesian Deep Learning (BDL), which attempts to capture uncertainties from noisy observations, aleatoric, and from unknown model parameters, epistemic, provide us a foundation. Here we present a discrete-continuous BDL model with Gaussian and lognormal likelihoods for uncertainty quantification (UQ). We demonstrate the approach by developing UQ estimates on `DeepSD', a super-resolution based DL model for Statistical Downscaling (SD) in climate applied to precipitation, which follows an extremely skewed distribution. We find that the discrete-continuous models outperform a basic Gaussian distribution in terms of predictive accuracy and uncertainty calibration. Furthermore, we find that the lognormal distribution, which can handle skewed distributions, produces quality uncertainty estimates at the extremes. Such results may be important across S&E, as well as other domains such as finance and economics, where extremes are often of significant interest. Furthermore, to our knowledge, this is the first UQ model in SD where both aleatoric and epistemic uncertainties are characterized., Comment: 10 Pages

The impacts of climate change are felt by most critical systems, such as infrastructure, ecological systems, and power-plants. However, contemporary Earth System Models (ESM) are run at spatial resolutions too coarse for assessing effects this localized. Local scale projections can be obtained using statistical downscaling, a technique which uses historical climate observations to learn a low-resolution to high-resolution mapping. Depending on statistical modeling choices, downscaled projections have been shown to vary significantly terms of accuracy and reliability. The spatio-temporal nature of the climate system motivates the adaptation of super-resolution image processing techniques to statistical downscaling. In our work, we present DeepSD, a generalized stacked super resolution convolutional neural network (SRCNN) framework for statistical downscaling of climate variables. DeepSD augments SRCNN with multi-scale input channels to maximize predictability in statistical downscaling. We provide a comparison with Bias Correction Spatial Disaggregation as well as three Automated-Statistical Downscaling approaches in downscaling daily precipitation from 1 degree (~100km) to 1/8 degrees (~12.5km) over the Continental United States. Furthermore, a framework using the NASA Earth Exchange (NEX) platform is discussed for downscaling more than 20 ESM models with multiple emission scenarios., Comment: 9 pages, 5 Figures, 2 Tables

We investigate the use of Deep Neural Networks for the classification of image datasets where texture features are important for generating class-conditional discriminative representations. To this end, we first derive the size of the feature space for some standard textural features extracted from the input dataset and then use the theory of Vapnik-Chervonenkis dimension to show that hand-crafted feature extraction creates low-dimensional representations which help in reducing the overall excess error rate. As a corollary to this analysis, we derive for the first time upper bounds on the VC dimension of Convolutional Neural Network as well as Dropout and Dropconnect networks and the relation between excess error rate of Dropout and Dropconnect networks. The concept of intrinsic dimension is used to validate the intuition that texture-based datasets are inherently higher dimensional as compared to handwritten digits or other object recognition datasets and hence more difficult to be shattered by neural networks. We then derive the mean distance from the centroid to the nearest and farthest sampling points in an n-dimensional manifold and show that the Relative Contrast of the sample data vanishes as dimensionality of the underlying vector space tends to infinity., Comment: Accepted in International Joint Conference on Neural Networks, IJCNN 2016

Satellite image classification is a challenging problem that lies at the crossroads of remote sensing, computer vision, and machine learning. Due to the high variability inherent in satellite data, most of the current object classification approaches are not suitable for handling satellite datasets. The progress of satellite image analytics has also been inhibited by the lack of a single labeled high-resolution dataset with multiple class labels. The contributions of this paper are twofold - (1) first, we present two new satellite datasets called SAT-4 and SAT-6, and (2) then, we propose a classification framework that extracts features from an input image, normalizes them and feeds the normalized feature vectors to a Deep Belief Network for classification. On the SAT-4 dataset, our best network produces a classification accuracy of 97.95% and outperforms three state-of-the-art object recognition algorithms, namely - Deep Belief Networks, Convolutional Neural Networks and Stacked Denoising Autoencoders by ~11%. On SAT-6, it produces a classification accuracy of 93.9% and outperforms the other algorithms by ~15%. Comparative studies with a Random Forest classifier show the advantage of an unsupervised learning approach over traditional supervised learning techniques. A statistical analysis based on Distribution Separability Criterion and Intrinsic Dimensionality Estimation substantiates the effectiveness of our approach in learning better representations for satellite imagery., Comment: Paper was accepted at ACM SIGSPATIAL 2015

Learning sparse feature representations is a useful instrument for solving an unsupervised learning problem. In this paper, we present three labeled handwritten digit datasets, collectively called n-MNIST. Then, we propose a novel framework for the classification of handwritten digits that learns sparse representations using probabilistic quadtrees and Deep Belief Nets. On the MNIST and n-MNIST datasets, our framework shows promising results and significantly outperforms traditional Deep Belief Networks., Comment: Published in the European Symposium on Artificial Neural Networks, ESANN 2015

The latest generation of geostationary satellites carry sensors such as the Advanced Baseline Imager (GOES-16/17) and the Advanced Himawari Imager (Himawari-8/9) that closely mimic the spatial and spectral characteristics of widely used polar orbiting sensors such as EOS/MODIS. More importantly, they provide observations at 1-5-15 minute intervals, instead of twice a day from MODIS, offering unprecedented opportunities for monitoring large parts of the Earth. In addition to serving the needs of weather forecasting, these observations offer new and exciting opportunities in managing solar power, fighting wildfires, and tracking air pollution. Creation of actionable information in near real-time from these data streams is a challenge that is best addressed through collaborative efforts among the industry, academia and government agencies.

In this on-going work, we report our efforts of building App Store at OpenNEX, which aims to provide a data analytics software search engine to help researchers find reusable software components to facilitate the development of their own algorithms.

GeoNEX is a collaborative project by scientists from NASA, NOAA, JAXA, and other organizations around the world with the purpose of generating a suite of Earth-monitoring products using data streams from the latest geostationary (GEO) sensors including the GOES-16/17 ABI and the Himawari-8/9 AHI. An accurate and consistent top-of-atmosphere (TOA) reflectance product, in particular the bidirectional reflectance factor (BRF), is the starting point in the scientific processing chain. We describe the main considerations and corresponding algorithms in generating the GeoNEX TOA BRF product. First, a special advantage of geostationary data streams is their high temporal resolution (~10 minutes per full-disk scan), providing a key source of information for many downstream products. To fully utilize this high temporal frequency demands a high georegistration accuracy for every acquired image. Our analysis shows that there can be substantial georegistration uncertainties in both GOES and Himawari L1b data which we addressed by implementing a phase-based correction algorithm to remove residual errors. Second, geostationary sensors have distinct illumination-view geometry features in that the solar angle changes for every pixel. Therefore, to accurately derive a BRF requires a solar position algorithm and the estimation of the pixel-wise acquisition time within an uncertainty of 10 seconds. Third, we discuss the measures we adopted to check and correct residual radiometric calibration issues of individual sensors to enable time-series analysis as well as the cross calibration between different satellite sensors (including those from low-Earth orbit). Finally, we also explain the rationale for the choice of the global grid/tile system of the GeoNEX TOA BRF product.

We have generated provisional Himawari-8 AHI surface reflectance (SR) product for land and vegetation monitoring. The Himawari-8 AHI surface reflectance product is part of our GeoNEX land products, which integrate level 2 and higher remote sensing data from a set of geostationary satellite sensors (i.e. GOES-16, -17 ABI, Himawari-8 AHI, FY4-A AGRI, and MTG-I). Adapted Multiangle Implementation of Atmospheric Correction (MAIAC) algorithm is used to process time series Himawari-8 AHI observations. Himawari-8 AHI SR provides gridded and tiled land SR in 1-km resolution with high frequency (every 10 minutes during daylight time). There are three subdatasets: 1) retrieved atmospheric properties (e.g. column water vapor at 0.86 m, aerosol optical depth at 0.47m and 0.51m); 2) spectral (AHI bands 1-6) surface reflectance, kernels of RTLS BRDF model; 3)spectral BRDF kernel weights, and extensive quality assurance flags. The evaluation results show that Himawari-8 AHI data yield much more valid pixels in a single day in the characterization of land surface, when compare to NASA flagship satellite MODIS Terra/Aqua. This observation frequency and resolution of geostationary data should allow for using continuous ecosystem monitoring in diurnal studies at continental scale. Initial evaluations indicate a stable Himawari-8 AHI land SR product.

Along with scientific applications, Geostationary imagery is often used to learn about weather patterns through true color visualizations. NOAA/NASA's GOES-R series of satellites uses the advanced baseline imager with 16-bands which, unlike previous generations, does not include the green wavelength (500-565 nm) and hence cannot directly generate true color images. However, Himawari, Japan's geostationary satellite, uses a similar 16-band advanced Himawari imager that does include a green band (but missing cirrus). In this work, we show how transfer learning with convolutional neural networks can be applied across satellites to generate "virtual sensors". We apply this approach to generate a green band for GOES-16 and present near true color images.

Applications in areas such as weather tracking and modeling, ecosystem monitoring, wildfire detection, and land-cover change are heavily dependent on spatial and temporal resolutions of satellite observations. However, there are typically trade-offs between spatial and temporal resolutions in dataset selection. For instance, geostationary weather tracking satellites are designed to take snapshots many times throughout the day but sensor hardware limits data collection. In this work we tackle this limitation, developing a method for temporal upsampling of multi-spectral satellite imagery using optical flow video interpolation deep convolutional neural networks. The presented model, extends Super SloMo (SSM) from single optical flow estimates to multichannel where flows are computed per band. We apply this technique on 8 multi-spectral bands of NOAA/NASA's GOES-16 mesoscale dataset to temporally enhance full disk hemispheric snapshots from 15 minutes to 1 minute. Through extensive experimentation, we show SSM vastly outperforms the linear interpolation baseline and that multichannel optical flows improves performance on GOES-16. A visual analysis of optical flow vectors clearly identifies hurricanes and large-scale atmospheric dynamics. Furthermore, we discuss challenges and open questions related to optical flow and temporal interpolation of multispectral geostationary satellite imagery.

The emergence of a new generation of geostationary satellite sensors provides land andatmosphere monitoring capabilities similar to MODIS and VIIRS with far greater temporal resolution (5-15 minutes). However, processing such large volume, highly dynamic datasets requires computing capabilities that (1) better support data access and knowledge discovery for scientists; (2) provide resources to enable real-time processing for emergency response (wildfire, smoke, dust, etc.); and (3) provide reliable and scalable services for the broader user community. This paper presents an implementation of GeoNEX (Geostationary NASA-NOAA Earth Exchange) services that integrate scientific algorithms with Amazon Web Services (AWS) to provide near realtime monitoring (~5 minute latency) capability in a hybrid cloud-computing environment. It offers a user-friendly, manageable and extendable interface and benefits from the scalability provided by Amazon Web Services. Four use cases are presented to illustrate how to (1) search and access geostationary data; (2) configure computing infrastructure to enable near real-time processing; (3) disseminate and utilize research results, visualizations, and animations to concurrent users; and (4) use a Jupyter Notebook-like interface for data exploration and rapid prototyping. As an example of (3), the Wildfire Automated Biomass Burning Algorithm (WF_ABBA) was implemented on GOES-16 and -17 data to produce an active fire map every 5 minutes over the conterminous US. Details of the implementation strategies, architectures, and challenges of the use cases are discussed.

New generation geostationary satellites make reflectance observations available at a continental scale with unprecedented spatiotemporal resolution and spectral range. Generating Earth monitoring products from these observations requires retrieval of the basic parameter, surface reflectance (SR), by atmospheric correction (AC). Algorithms for atmospheric correction, including Multi-Angle Implementation of Atmospheric Correction (MAIAC), are adapted for each sensor and are too computationally complex to be run in real time, relying instead on look-up tables with precomputed values. Machine learning methods, including convolutional neural networks, have demonstrated performance in learning complex, nonlinear mappings and extracting insight from high-dimensional remote sensing data. In this work, we present a deep learning emulator of MAIAC to retrieve both SR and cloud products. Using this adaptation of deep learning-based emulation to remote sensing, we demonstrate stable SR retrieval over a variety of land covers and viewing conditions and accurate cloud detection. Further, a comparison of computation time suggests emulation as a compelling alternative for expensive physical simulation, especially for applications benefited by near-real time data, such as agricultural management and disaster response.

The latest generation of geostationary satellites carry sensors such as the Advanced Baseline Imager (GOES-16/17) and the Advanced Himawari Imager (Himawari-8/9) that closely mimic the spatial and spectral characteristics of MODIS and VIIRS, useful for monitoring land surface conditions. The NASA Earth Exchange (NEX) team at Ames Research Center has embarked on a collaborative effort among scientists from NASA and NOAA exploring the feasibility of producing operational land surface products similar to those from MODIS/VIIRS. The team built a processing pipeline called GEONEX that is capable of converting raw geostationary data into routine products of Fires, surface reflectances, vegetation indices, LAI/FPAR, ET and GPP/NPP using algorithms adapted from both NASA/EOS and NOAA/GOES-R programs. The GEONEX pipeline will begin to produce provisional data products to be consumed by external collaborators and the academic community. In order to better inform and introduce the GEONEX products to the science community, the provisional products shall be distributed from the NAS data portal, located at data.nas.nasa.gov, and simple webpage at www.nasa.gov/geonex has been deployed, which describes any algorithms used in deriving the products, user manuals and data file information. We will also update the status of the data processing, on the website and provide links to the latest datasets, and use a geonex mailing list, using lists.nasa.gov.

The latest generation of geostationary satellites carry sensors such as the Advanced Baseline Imager (GOES-16/17) and the Advanced Himawari Imager (Himawari-8/9) that closely mimic the spatial and spectral characteristics of MODIS and VIIRS, useful for monitoring land surface conditions. The NASA Earth Exchange (NEX) team at Ames Research Center has embarked on a collaborative effort among scientists from NASA and NOAA exploring the feasibility of producing operational land surface products similar to those from MODIS/VIIRS. The team built a processing pipeline called GeoNEX that is capable of converting raw geostationary data into routine products of Fires, surface reflectances, vegetation indices, LAI/FPAR, ET and GPP/NPP using algorithms adapted from both NASA/EOS and NOAA/GOES-R programs. The GeoNEX pipeline has been deployed on Amazon Web Services cloud platform and it currently leverages near-realtime geostationary data hosted in AWS public datasets under a NOAA-AWS agreement. Initial analyses of various products from ABI/AHI sensors suggest that they are comparable to those from MODIS in representing the spatio-temporal dynamics of land conditions. Cloud computing offers a variety of options for deploying the GeoNEX pipeline including choice CPUs, storage media, and automation. By making the GEONEX pipeline available on the cloud, we hope to engage a broad community of Earth scientists from around the world in utilizing this new source of data for Earth monitoring.

The latest generation of geostationary satellites carry sensors such as the Advanced Baseline Imager (GOES-16/17) and the Advanced Himawari Imager (Himawari-8/9) that closely mimic the spatial and spectral characteristics of MODIS and VIIRS, useful for monitoring land surface conditions. The NASA Earth Exchange (NEX) team at Ames Research Center has embarked on a collaborative effort among scientists from NASA and NOAA exploring the feasibility of producing operational land surface products similar to those from MODIS/VIIRS. The team built a processing pipeline called GEONEX that is capable of converting raw geostationary data into routine products of Fires, surface reflectances, vegetation indices, LAI/FPAR, ET and GPP/NPP using algorithms adapted from both NASA/EOS and NOAA/GOES-R programs. The GEONEX pipeline has been deployed on Amazon Web Services cloud platform and it currently leverages near-realtime geostationary data hosted in AWS public datasets under a NOAA-AWS agreement. Initial analyses of various products from ABI/AHI sensors suggest that they are comparable to those from MODIS in representing the spatio-temporal dynamics of land conditions. Cloud computing offers a variety of options for deploying the GEONEX pipeline including choice CPUs, storage media, and automation. By making the GEONEX pipeline available on the cloud, we hope to engage a broad community of Earth scientists from around the world in utilizing this new source of data for Earth monitoring.