Search NASA⌕ Search

SEARCH · Search NASA

Results for “Observational Networks”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 703 records · Page 39

Neural Network Analysis on the NOvA Experiment

NOνA is collaboration of 180 scientists and engineers from 28 institutions which plans to study neutrino oscillations using the existing NuMI neutrino beam at Fermilab. The NOνA experiment is designed to search for oscillations of muon neutrinos to electron neutrinos by comparing the electron neutrino event rate measured at the Fermilab site with the electron neutrino event rate measured at a location just south of International Falls, MN 810 kilometers distant from Fermilab. If oscillations occur, the far site will see the appearance of electrons in the muon neutrino beam produced at Fermilab. The presence of background radiation obscures the desired particles and trails to be observed. Using neural network analysis, the goal of the project was to implement machine learning to automate the removal of background radiation to render pixel maps of the particle trajectories.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Network design for quantifying urban CO<sub>2</sub> emissions: assessing trade-offs between precision and network density

The majority of anthropogenic CO 2 emissions are attributable to urban areas. While the emissions from urban electricity generation often occur in locations remote from consumption, many of the other emissions occur within the city limits. Evaluating the effectiveness of strategies for controlling these emissions depends on our ability to observe urban CO 2 emissions and attribute them to specific activities. Cost-effective strategies for doing so have yet to be described. Here we characterize the ability of a prototype measurement network, modeled after the Berkeley Atmospheric CO 2 Observation Network (BEACO 2 N) in California's Bay Area, in combination with an inverse model based on the coupled Weather Research and Forecasting/Stochastic Time-Inverted Lagrangian Transport (WRF-STILT) to improve our understanding of urban emissions. The pseudo-measurement network includes 34 sites at roughly 2 km spacing covering an area of roughly 400 km 2 . The model uses an hourly 1 × 1 km 2 emission inventory and 1 × 1 km 2 meteorological calculations. We perform an ensemble of Bayesian atmospheric inversions to sample the combined effects of uncertainties of the pseudo-measurements and the model. We vary the estimates of the combined uncertainty of the pseudo-observations and model over a range of 20 to 0.005 ppm and vary the number of sites from 1 to 34. We use these inversions to develop statistical models that estimate the efficacy of the combined model–observing system in reducing uncertainty in CO 2 emissions. We examine uncertainty in estimated CO 2 fluxes on the urban scale, as well as for sources embedded within the city such as a line source (e.g., a highway) or a point source (e.g., emissions from the stacks of small industrial facilities). Using our inversion framework, we find that a dense network with moderate precision is the preferred setup for estimating area, line, and point sources from a combined uncertainty and cost perspective. The dense network considered here (modeled after the BEACO 2 N network with an assumed mismatch error of 1 ppm at an hourly temporal resolution) could estimate weekly CO 2 emissions from an urban region with less than 5 % error, given our characterization of the combined observation and model uncertainty.

54 ENVIRONMENTAL SCIENCES↗

Geometric adjustment of the South American satellite densification (PC-1000) network

Reduced normal equations were computed from observation data and combined with reduced normal equations of other satellite networks to provide station coordinates from a single least square adjustment. Terrestrial data, which include base-lines, heights, and survey coordinates, provide the necessary relative position constraints between collocated stations of two satellite networks. Survey information regarding the observation stations is summarized, and constraints used in the solution are given. Geoidal undulations are computed by using the formula and constants shown.

Mueller, I. I.↗

Collaborative Clustering for Sensor Networks

Traditionally, nodes in a sensor network simply collect data and then pass it on to a centralized node that archives, distributes, and possibly analyzes the data. However, analysis at the individual nodes could enable faster detection of anomalies or other interesting events, as well as faster responses such as sending out alerts or increasing the data collection rate. There is an additional opportunity for increased performance if individual nodes can communicate directly with their neighbors. Previously, a method was developed by which machine learning classification algorithms could collaborate to achieve high performance autonomously (without requiring human intervention). This method worked for supervised learning algorithms, in which labeled data is used to train models. The learners collaborated by exchanging labels describing the data. The new advance enables clustering algorithms, which do not use labeled data, to also collaborate. This is achieved by defining a new language for collaboration that uses pair-wise constraints to encode useful information for other learners. These constraints specify that two items must, or cannot, be placed into the same cluster. Previous work has shown that clustering with these constraints (in isolation) already improves performance. In the problem formulation, each learner resides at a different node in the sensor network and makes observations (collects data) independently of the other learners. Each learner clusters its data and then selects a pair of items about which it is uncertain and uses them to query its neighbors. The resulting feedback (a must and cannot constraint from each neighbor) is combined by the learner into a consensus constraint, and it then reclusters its data while incorporating the new constraint. A strategy was also proposed for cleaning the resulting constraint sets, which may contain conflicting constraints; this improves performance significantly. This approach has been applied to collaborative clustering of seismic and infrasonic data collected by the Mount Erebus Volcano Observatory in Antarctica. Previous approaches to distributed clustering cannot readily be applied in a sensor network setting, because they assume that each node has the same view of the data set. A view is the set of features used to represent each object. When a single data set is partitioned across several computational nodes, distributed clustering works; all objects have the same view. But when the data is collected from different locations, using different sensors, a more flexible approach is needed. This approach instead operates in situations where the data collected at each node has a different view (e.g., seismic vs. infrasonic sensors), but they observe the same events. This enables them to exchange information about the likely cluster membership relations between objects, even if they do not use the same features to represent the objects.

Wagstaff. Loro :/↗

Imprint of Anthropogenic Sources and Soil Removal on the Surface Concentration of H2 in the Contiguous US

Hydrogen (H2) is experiencing renewed interest throughout the world as a low carbon fuel alternative or complement to fossil fuels. Significant uncertainties remain regarding the environmental impact of increasing H2 usage, in part due to gaps in our understanding of the H2 atmospheric budget, including the H2 release from industrial activities and the H2 soil removal, the most important sink of H2. This study focuses on H2 dry air mole fractions measured by the NOAA Global Monitoring Laboratory in discrete ambient air samples collected every few days at sites located in the contiguous United States between 2010 and 2022. We take advantage of the long-term observations from this network to study the regional distribution of H2 sources using the potential source contribution function (PSCF). We find that H2 PSCF is consistent with a large anthropogenic source of atmospheric H2 from the incomplete combustion of fossil fuel. However, episodic air masses with high H2 dry air mole fractions (>700 ppb) recorded at some sampling locations in the Western and Southeastern US are not associated with elevated CO, suggesting significant noncombustion anthropogenic or geological sources of H2. Air masses depleted in H2 are recorded in rural or remote continental sites. Our analysis of observations from a tall tower in NE Colorado suggests that the degree of H2 depletion not only reflects differences in soil exposure but also the regional inhibition of the soil sink under arid conditions.

Paulot, Fabien↗

The Connection between Model Performance on the CCMVal Transport Diagnostics and Simulated Sensitivity of Ozone to Chlorine Change

The chemistry climate models that contributed simulations for past and future ozone evolution to the 2010 Scientific Assessment of Ozone Depletion were subject to extensive evaluation by the SPARC (Stratospheric Processes and their Role in Climate) CCMVal (Chemistry-Climate Model Validation) activity. The sensitivity of ozone to changes in composition and climate varies among the models, but the relationship between these variations and the model evaluations of CCMVal is not obvious. We have learned that the transport evaluation can be used to interpret the comparisons between observed and simulated columns of chlorine reservoirs, hydrochloric acid (HCl) and chlorine nitrate (ClONO2); these comparisons were part of the CCMVal evaluation of chemistry. The simulations with best performance on the transport diagnostics most faithfully reproduce the evolution and seasonal variation of the chlorine reservoirs as observed at NDACC (Network for Detection of Atmospheric Composition Change) stations (NyAlesund 78.9N, Kiruna 67.8N, Harestua 60.2N, Jungfraujoch 46.6N, Toronto 43.6N, Kitt Peak 31.9N, Izana 28.3N, Mauna Loa 19.5N, Lauder 45S and Arrival Heights 77.8S). In the simulations, the HCl in the lower stratosphere depends on total inorganic chlorine (Cly) and partitioning between HCl and ClON02. Total inorganic chlorine depends on the fractional release of chlorine from source gases, and ratio of ClON02 to HCl is inversely dependent on methane and varies quadratically with ozone. Simulated HCl from various models may agree with observations even though Cly is in error, partitioning is in error, or both. Simulated ozone sensitivity to chlorine is shown to be greater for models that produce larger values of chlorine nitrate for background chlorine levels, and vice versa. Comparisons with the NDACC data show why the models with 'best' transport have similar sensitivity to chlorine change. The realistic evolution of the simulated HCl and ClONO2 columns suggests realistic levels of Cly in the lower atmosphere. In addition, the wide range values for the sensitivity of ozone to chlorine obtained from the CCMVal simulations is explained by the wide range in lower atmospheric columns of ClONO2 and the concomitant wide range of levels for chlorine monoxide.

Douglass, Anne↗

Optimizing radar scan strategies for tracking isolated deep convection using observing system simulation experiments

Abstract. Optimizing radar observation strategies is one of the most important considerations in pre-field campaign periods. This is especially true for isolated convective clouds that typically evolve faster than the observations captured by operational radar networks. This study investigates uncertainties in radar observations of the evolution of the microphysical and dynamical properties of isolated deep convective clouds developing in clean and polluted environments. It aims to optimize the radar observation strategy for deep convection through the use of high-spatiotemporal cloud-resolving model simulations, which resolve the evolution of individual convective cells every 1 min, coupled with a radar simulator and a cell tracking algorithm. The radar simulation settings are based on the Tracking Aerosol Convection Interactions ExpeRiment (TRACER) and Experiment of Sea Breeze Convection, Aerosols, Precipitation and Environment (ESCAPE) field campaigns held in the Houston, TX, area but are generalizable to other field campaigns focusing on isolated deep convection. Our analysis produces the following four outcomes. First, a 5–7 m s−1 median difference in maximum updrafts of tracked cells is shown between the clean and polluted simulations in the early stages of the cloud lifetimes. This demonstrates the importance of obtaining accurate estimates of vertical velocity from observations if aerosol impacts are to be properly resolved. Second, tracking of individual cells and using vertical cross section scanning every minute capture the evolution of precipitation particle number concentration and size represented by polarimetric observables better than the operational radar observations that update the volume scan every 5 min. This approach also improves multi-Doppler radar updraft retrievals above 5 km above ground level for regions with updraft velocities greater than 10 m s−1. Third, we propose an optimized strategy composed of cell tracking by quick (1–2 min) vertical cross section scans from more than one radar in addition to the operational volume scans. We also propose the use of a single-RHI (range height indicator) updraft retrieval technique for cells close to the radars, for which multi-Doppler radar retrievals are still challenging. Finally, increasing the number of deep convective cells sampled by such observations better represents the median maximum updraft evolution with sample sizes of more than 10 deep cells, which decreases the error associated with sampling the true population to less than 3 m s−1.

54 ENVIRONMENTAL SCIENCES↗

Simulating the density of organic species in the atmosphere of Titan with a coupled ion-neutral photochemical model

We introduce a one-dimensional coupled ion-neutral photochemical kinetics and diffusion model to study the atmospheric composition of Titan in light of new theoretical kinetics calculations and scientific findings from the Cassini–Huygens mission. The model extends from the surface to the exobase. The atmospheric background, boundary conditions, vertical transport and aerosol opacity are all constrained by the Cassini–Huygens observations. The chemical network includes reactions between hydrocarbons, nitrogen and oxygen bearing species. It takes into account neutrals and both positive and negative ions with masses extending up to 116 and 74 u, respectively. We incorporate high-resolution isotopic photoabsorption and photodissociation cross sections for N 2 as well as new photodissociation branching ratios for CH 4 and C 2 H 2 . Ab initio transition state theory calculations are performed in order to estimate the rate coefficients and products for critical reactions. Critical reactions of production and loss for neutrals and ions are quantitatively assessed and thoroughly discussed. The vertical distributions of neutrals and ions predicted by the model generally reproduce observational data, suggesting that for the small species most chemical processes in Titan’s atmosphere and ionosphere are adequately described and understood; some differences are highlighted. Notable remaining issues include (i) the total positive ion density (essentially HCNH + ) in the upper ionosphere, (ii) the low mass negative ion densities (CN - , C 3 N - /C4H - ) in the upper atmosphere, and (iii) the minor oxygen-bearing species (CO 2 , H 2 O) density in the stratosphere. Pathways towards complex molecules and the impact of aerosols (UV shielding, atomic and molecular hydrogen budget, nitriles heterogeneous chemistry and condensation) are evaluated in the model, along with lifetimes and solar cycle variations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Synoptic conditions producing cirrus during the FIRE cirrus IFO

Although direct observations of cirrus clouds by the FIRE research aircraft were usually confined to the area of Intensive Field Observation (IFO) surface network, these cirrus were generally part of a more extensive zone of upper level cloudiness. It is these large scale patterns of cirriform cloud and their relationship to the corresponding synoptic environment which are the prime focus. Three conceptual models are presented and each of the individual cases are classified into one of these categories. Although the cases manifest significant differences in intensity and small scale structure, it is believed that they are best viewed in this unified context. The descriptions given are mostly qualitative, however, quantitative descriptions of the synoptic control and its relationship to cloud structure for all the IFO cases are summarized. The synoptic situations in which extensive cirriform clouds were observed are classified into three basic types: warm front cases, cold front cases, and closed low aloft cases. A simplified summary of each type of situation is presented.

Starr, David OC.↗

Overview of the New Version 3 NASA Micro-Pulse Lidar Network (MPLNET) Automatic Precipitation Detection Algorithm

Precipitation modifies atmospheric column thermodynamics through the process of evaporation and serves as a proxy for latent heat modulation. For this reason, a correct precipitation parameterization (especially for low-intensity precipitation) within global scale models is crucial. In addition to improving our modeling of the hydrological cycle, this will reduce the associated uncertainty of global climate models in correctly forecasting future scenarios, and will enable the application of mitigation strategies. In this manuscript we present a proof of concept algorithm to automatically detect precipitation from lidar measurements obtained from the National Aeronautics and Space Administration Micropulse lidar network (MPLNET). The algorithm, once tested and validated against other remote sensing instruments, will be operationally implemented into the network to deliver a near real time (latency <1.5 h) rain masking variable that will be publicly available on MPLNET website as part of the new Version 3 data products. The methodology, based on an image processing technique, detects only light precipitation events (defined by intensity and duration) such as light rain, drizzle, and virga. During heavy rain events, the lidar signal is completely extinguished after a few meters in the precipitation or it is unusable because of water accumulated on the receiver optics. Results from the algorithm, in addition to filling a gap in light rain, drizzle, and virga detection by radars, are of particular interest for the scientific community as they help to fully characterize the aerosol cycle, from emission to deposition, as precipitation is a crucial meteorological phenomenon accelerating atmospheric aerosol removal through the scavenging effect. Algorithm results will also help the understanding of long term aerosol–cloud interactions, exploiting the multi-year database from several MPLNET permanent observational sites across the globe. The algorithm is also applicable to other lidar and/or ceilometer network infrastructures in the framework of the Global Aerosol Watch (GAW) aerosol lidar observation network (GALION).

Simone Lolli↗

A Neural Network Parametrization of Volumetric Cloud Fraction Profiles Using Satellite Observations and MERRA-2 Reanalysis Meteorological Data

Clouds play a crucial role in regulating the hydrologic cycle and Earth's radiative energy budget, yet they are often poorly represented in global climate models (GCMs). This study applies deep machine learning techniques to develop a physical parameterization of volumetric cloud fraction (VCF), the fraction of a 3-D grid volume occupied by clouds using satellite lidar-radar measurements. The neural network (NN) captures the complicated relationships between observed VCF profiles and collocated meteorological variables from MERRA-2 reanalysis data. Our results show that the NN model, particularly a sequence-to-sequence long short-term memory (LSTM) network with a sixfactor loss function, effectively learns the underlying cloud physical processes. The NN model outperforms MERRA-2 reanalysis in representing low-level clouds in tropical and subtropical regions and low- and middle-level clouds over midlatitude storm-track regions, and also improves VCF histograms. These improvements are reflected in the vertical distributions of zonally, meridionally, and globally averaged VCFs, geographic distributions of low-, middle-, and high-level clouds, and seasonal variations in monthly-mean VCF. Furthermore, the NN predictions effectively capture the El Niño-Southern Oscillation (ENSO) effects and other interannual variations. The NN parameterization is further evaluated through a sensitivity analysis, in which a single predictor is perturbed at a time. This reveals that relative humidity (RH) is the dominant factor influencing variations in globally averaged VCF at low and middle altitudes, followed by temperature. At higher altitudes, temperature becomes the primary driver of VCF through its effect on RH. Changes in wind components had minimal impact on globally averaged VCF.

Shan Zeng↗

The Graph Laplacian and the Dynamics of Complex Networks [Slides]

In this talk, we explore the structure of networks from a spectral graph-theoretic perspective by analyzing the properties of the Laplacian matrix associated with the graph induced by a network. We will see how the eigenvalues of the graph Laplacian relate to the underlying network structure and dynamics and provides insight into a phenomenon frequently observed in real world networks - the emergence of collective behavior from purely local interactions seen in the coordinated motion of animals and phase transitions in biological networks, to name a few.

97 MATHEMATICS AND COMPUTING↗

Observing the Global Water Cycle from Space

This paper presents an approach to measuring all major components of the water cycle from space. Key elements of the global water cycle are discussed in terms of the storage of water-in the ocean, air, cloud and precipitation, in soil, ground water, snow and ice, and in lakes and rivers, and in terms of the global fluxes of water between these reservoirs. Approaches to measuring or otherwise evaluating the global water cycle are presented, and the limitations on known accuracy for many components of the water cycle are discussed, as are the characteristic spatial and temporal scales of the different water cycle components. Using these observational requirements for a global water cycle observing system, an approach to measuring the global water cycle from space is developed. The capabilities of various active and passive microwave instruments are discussed, as is the potential of supporting measurements from other sources. Examples of space observational systems, including TRMM/GPM precipitation measurement, cloud radars, soil moisture, sea surface salinity, temperature and humidity profiling, other measurement approaches and assimilation of the microwave and other data into interpretative computer models are discussed to develop the observational possibilities. The selection of orbits is then addressed, for orbit selection and antenna size/beamwidth considerations determine the sampling characteristics for satellite measurement systems. These considerations dictate a particular set of measurement possibilities, which are then matched to the observational sampling requirements based on the science. The results define a network of satellite instrumentation systems, many in low Earth orbit, a few in geostationary orbit, and all tied together through a sampling network that feeds the observations into a data-assimilative computer model.

Hildebrand, P. H.↗

Global Precipitation Measurement (GPM) Mission: Overview and Status

The Global Precipitation Measurement (GPM) Mission is an international satellite mission specifically designed to unify and advance precipitation measurements from a constellation of research and operational microwave sensors. NASA and JAXA will deploy a Core Observatory in 2014 to serve as a reference satellite to unify precipitation measurements from the constellation of sensors. The GPM Core Observatory will carry a Ku/Ka-band Dual-frequency Precipitation Radar (DPR) and a conical-scanning multi-channel (10-183 GHz) GPM Microwave Radiometer (GMI). The DPR will be the first dual-frequency radar in space to provide not only measurements of 3-D precipitation structures but also quantitative information on microphysical properties of precipitating particles. The DPR and GMI measurements will together provide a database that relates vertical hydrometeor profiles to multi-frequency microwave radiances over a variety of environmental conditions across the globe. This combined database will be used as a common transfer standard for improving the accuracy and consistency of precipitation retrievals from all constellation radiometers. For global coverage, GPM relies on existing satellite programs and new mission opportunities from a consortium of partners through bilateral agreements with either NASA or JAXA. Each constellation member may have its unique scientific or operational objectives but contributes microwave observations to GPM for the generation and dissemination of unified global precipitation data products. In addition to the DPR and GMI on the Core Observatory, the baseline GPM constellation consists of the following sensors: (1) Special Sensor Microwave Imager/Sounder (SSMIS) instruments on the U.S. Defense Meteorological Satellite Program (DMSP) satellites, (2) the Advanced Microwave Scanning Radiometer-2 (AMSR-2) on the GCOM-W1 satellite of JAXA, (3) the Multi-Frequency Microwave Scanning Radiometer (MADRAS) and the multi-channel microwave humidity sounder (SAPHIR) on the French-Indian MeghaTropiques satellite, (4) the Microwave Humidity Sounder (MHS) on the National Oceanic and Atmospheric Administration (NOAA)-19, (5) MHS instruments on MetOp satellites launched by the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT), (6) the Advanced Technology Microwave Sounder (ATMS) on the National Polar-orbiting Operational Environmental Satellite System (NPOESS) Preparatory Project (NPP), and (7) ATMS instruments on the NOAA-NASA Joint Polar Satellite System (JPSS) satellites. Data from Chinese and Russian microwave radiometers may also become available through international collaboration under the auspices of the Committee on Earth Observation Satellites (CEOS) and Group on Earth Observations (GEO). The current generation of global rainfall products combines observations from a network of uncoordinated satellite missions using a variety of merging techniques. GPM will provide "next-generation" precipitation products characterized by: (1) more accurate instantaneous precipitation estimate (especially for light rain and cold-season solid precipitation), (2) intercalibrated microwave brightness temperatures from constellation radiometers within a consistent framework, and (3) unified precipitation retrievals from constellation radiometers using a common a priori hydrometeor database constrained by combined radar/radiometer measurements provided by the GPM Core Observatory. GPM is a science mission with integrated applications goals. GPM will provide a key measurement to improve understanding of global water cycle variability and freshwater availability in a changing climate. The DPR and GMI measurements will offer insights into 3-dimensional structures of hurricanes and midlatitude storms, microphysical properties of precipitating particles, and latent heat associated with precipitation processes. The GPM mission will also make data available in near realtime (within 3 hours of observations) forocietal applications ranging from position fixes of storm centers, numerical weather prediction, flood forecasting, freshwater management, landslide warning, crop prediction, to tracking of water-borne diseases. An overview of the GPM mission design, retrieval strategy, ground validation activities, and international science collaboration will be presented.

Hou, Arthur Y.↗

Collaborative Supervised Learning for Sensor Networks

Collaboration methods for distributed machine-learning algorithms involve the specification of communication protocols for the learners, which can query other learners and/or broadcast their findings preemptively. Each learner incorporates information from its neighbors into its own training set, and they are thereby able to bootstrap each other to higher performance. Each learner resides at a different node in the sensor network and makes observations (collects data) independently of the other learners. After being seeded with an initial labeled training set, each learner proceeds to learn in an iterative fashion. New data is collected and classified. The learner can then either broadcast its most confident classifications for use by other learners, or can query neighbors for their classifications of its least confident items. As such, collaborative learning combines elements of both passive (broadcast) and active (query) learning. It also uses ideas from ensemble learning to combine the multiple responses to a given query into a single useful label. This approach has been evaluated against current non-collaborative alternatives, including training a single classifier and deploying it at all nodes with no further learning possible, and permitting learners to learn from their own most confident judgments, absent interaction with their neighbors. On several data sets, it has been consistently found that active collaboration is the best strategy for a distributed learner network. The main advantages include the ability for learning to take place autonomously by collaboration rather than by requiring intervention from an oracle (usually human), and also the ability to learn in a distributed environment, permitting decisions to be made in situ and to yield faster response time.

Wagstaff, Kiri L.↗

Observations of Leonids 2009 by the Tajikistan Fireball Network

The fireball network in Tajikistan has operated since 2009. Five stations of the network covering the territory of near eleven thousands square kilometers are equipped with all-sky cameras with the Zeiss Distagon "fish-eye" objectives and by digital SLR cameras Nikon with the Nikkor "fish-eye" objectives. Observations of the Leonid activity in 2009 were carried out during November 13-21. In this period, 16 Leonid fireballs have been photographed. As a result of astrometric and photometric reductions, the precise data including atmospheric trajectories, velocities, orbits, light curves, photometric masses and densities were determined for 10 fireballs. The radiant positions during the maximum night suggest that the majority of the fireball activity was caused by the annual stream component with only minor contribution from the 1466 trail. According to the PE criterion, the majority of Leonid fireballs belonged to the most fragile and weak fireball group IIIB. However, one detected Leonid belonged to the fireball group I. This is the first detection of an anomalously strong Leonid individual.

Borovicka, J.↗

How Well Do Satellite AOD Observations Represent the Spatial and Temporal Variability of PM2.5 Concentration for the United States?

Due to their extensive spatial coverage, satellite Aerosol Optical Depth (AOD) observations have been widely used to estimate and predict surface PM2.5 concentrations. While most previous studies have focused on establishing relationships between collocated, hourly or daily AOD and PM2.5 measurements, in this study, we instead focus on the comparison of the large-scale spatial and temporal variability between satellite AOD and PM2.5 using monthly mean measurements. A newly developed spectral analysis technique e Combined Maximum Covariance Analysis (CMCA) is applied to Moderate Resolution Imaging Spectroradiometer (MODIS), Multi-angle Imaging Spectroradiometer (MISR), Sea-viewing Wide Field-of-view Sensor (SeaWiFS) and Ozone Monitoring Instrument (OMI) AOD datasets and Environmental Protection Agency (EPA) PM2.5 data, in order to extract and compare the dominant modes of variability. Results indicate that AOD and PM2.5 agree well in terms of interannual variability. An overall decrease is found in both AOD and PM2.5 across the United States, with the strongest signal over the eastern US. With respect to seasonality, good agreement is found only for Eastern US, while for Central and Western US, AOD and PM2.5 seasonal cycles are largely different or even reversed. These results are verified using Aerosol Robotic Network (AERONET) AOD observations and differences between satellite and AERONET are also examined. MODIS and MISR appear to have the best agreement with AERONET. In order to explain the disagreement between AOD and PM2.5 seasonality, we further use Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) extinction profile data to investigate the effect of two possible contributing factors, namely aerosol vertical distribution and cloud-free sampling. We find that seasonal changes in aerosol vertical distribution, due to the seasonally varying mixing height, is the primary cause for the AOD and PM2.5 seasonal discrepancy, in particular, the low AOD but high PM2.5 observed during the winter season for Central and Western US. In addition, cloud-free sampling by passive sensors also induces some bias in AOD seasonality, especially for the Western US, where the largest seasonal change in cloud fraction is found. The seasonal agreement between low level (below 500 m AGL), all sky CALIOP AOD and PM2.5 is significantly better than column AOD from MODIS, MISR, SeaWiFS and OMI. In particular, the correlation between low level, all sky AOD and PM2.5 seasonal cycles increases to above 0.7 for Central and Western US, as opposed to near zero or negative correlation for column, clear sky AOD. This result highlights the importance of accounting for the seasonally varying aerosol profiles and cloud-free sampling bias when using column AOD measurements to infer surface PM2.5 concentrations.

MODIS (radiometry)↗

Improving Quality of Observational Streaming Medical Data by Using Long Short-Term Memory Networks (LSTMs)

We present an exploration of the encoder-decoder structured Long Short-Term Memory Network (LSTM) as a detector of the anomalous missing observations in streaming medical data by using the difference between the LSTM-reconstructed and observed values as the anomaly detector. We experiment with time-series data from bedside monitoring devices from the available Medical Information Mart for Intensive Care Database (MIMIC). Our results show that not only encoder-decoder LSTM approach works well for detecting the difference between anomalous and normal missing observations in streaming medical data, but also has an imputation potential for the missing data.

Bowie, Michael B.↗