Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “gradient boost machine”

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.

17 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS2021

Large-Scale High-Resolution Coastal Mangrove Forests Mapping Across West Africa With Machine Learning Ensemble and Satellite Big Data

Coastal mangrove forests provide important ecosystem goods and services, including carbon sequestration, biodiversity conservation, and hazard mitigation. However, they are being destroyed at an alarming rate by human activities. To characterize mangrove forest changes, evaluate their impacts, and support relevant protection and restoration decision making, accurate and up-to-date mangrove extent mapping at large spatial scales is essential. Available large-scale mangrove extent data products use a single machine learning method commonly with 30 m Landsat imagery, and significant inconsistencies remain among these data products. With huge amounts of satellite data involved and the heterogeneity of land surface characteristics across large geographic areas, finding the most suitable method for large-scale high-resolution mangrove mapping is a challenge. The objective of this study is to evaluate the performance of a machine learning ensemble for mangrove forest mapping at 20 m spatial resolution across West Africa using Sentinel-2 (optical) and Sentinel-1 (radar) imagery. The machine learning ensemble integrates three commonly used machine learning methods in land cover and land use mapping, including Random Forest (RF), Gradient Boosting Machine (GBM), and Neural Network (NN). The cloud-based big geospatial data processing platform Google Earth Engine (GEE) was used for pre-processing Sentinel-2 and Sentinel-1 data. Extensive validation has demonstrated that the machine learning ensemble can generate mangrove extent maps at high accuracies for all study regions in West Africa (92%–99% Producer’s Accuracy, 98%–100% User’s Accuracy, 95%–99% Overall Accuracy). This is the first-time that mangrove extent has been mapped at a 20 m spatial resolution across West Africa. The machine learning ensemble has the potential to be applied to other regions of the world and is therefore capable of producing high-resolution mangrove extent maps at global scales periodically.

coastal environment↗
NASA NTRS2020

Building a landslide hazard indicator with machine learning and land surface models

The U.S. Pacific Northwest has a history of frequent and occasionally deadly landslides caused by various factors. Using a multivariate, machine-learning approach, we combined a Pacific Northwest Landslide Inventory with a 36-year gridded hydrologic dataset from the National Climate Assessment – Land Data Assimilation System to produce a landslide hazard indicator (LHI) on a daily 0.125-degree grid. The LHI identified where and when landslides were most probable over the years 1979–2016, addressing issues of bias and completeness that muddy the analysis of multi-decadal landslide inventories. The seasonal cycle was strong along the west coast, with a peak in the winter, but weaker east of the Cascade Range. This lagging indicator can fill gaps in the observational record to identify the seasonality of landslides over a large spatiotemporal domain and show how landslide hazard has responded to a changing climate.

XGBoost↗
NASA NTRSDate not supplied

Predicting Airport Runway Configurations for Decision-Support Using Supervised Learning

One of the most challenging tasks for air traffic controllers is runway configuration management (RCM). It deals with the optimal selection of runways to operate on (for arrivals and departures) based on traffic, surface wind speed, wind direction, other environmental variables, noise constraints, and several other airport-specific factors. It affects the efficiency of the National Airspace System (NAS) and both surface and airspace operations can benefit from better understanding future runway configurations. In this paper, we present a comprehensive implementation of predictive models for runway configuration estimation from large volumes of historical data. Specifically, operational data from two full years (2018 and 2019) is collected, analyzed, and fused together to build the data product used in this work. The data set differs from prior work in the field in terms of its scope, resolution, and variety of factors collected and considered. Meteorological data is collected from two different sources – current weather conditions from METAR (Meteorological Terminal Aviation Routine Weather Report) and forecast weather conditions from Localized Aviation MOS Program (LAMP). Operational data from the Federal Aviation Administration (FAA) Aviation System Performance Metrics (ASPM) related to scheduled and actual number of arrivals and departures, average taxi times, etc. are collected. NASA’s Sherlock Data Warehouse is used to identify critical information such as go-arounds, and other events that might impact RCM decision-making. All data is collected and aggregated over 15-minute intervals throughout the two years. This provides a resolution like the timescales that might be necessary for runway configuration management decision-making. A variety of supervised learning algorithms are tested including Support Vector Machine, Random Forest, Gradient Boosting, etc. including tuning of the model hyperparameters. The modeling process is applied and presented on two representative U.S. airports – Charlotte Douglas International Airport (KCLT) and Denver International Airport (KDEN). The two airports present different levels of complexity in terms of the total number of configurations used and provide a balanced perspective on the generalizability of the developed approach to other airports in the NAS. Initial results are promising (F1 score of 0.91 at KCLT and 0.83 at KDEN) for data in the test set. The final paper will contain a comprehensive comparison between different models and model building strategies as well as further refined results. Most important predictors for each airport will be identified along with a discussion and recommendations on adapting the framework to other scenarios.

Tejas G Puranik↗
NASA NTRS2021

Machine Learning to Predict Joint Performance in Epoxy Composites Based on Process Parameters

Polymer matrix composites are gaining popularity in the aerospace industry due to their high specific strength, fatigue properties, and processability. However, based on current FAA certification guidelines, manufacturers utilizing current state-of-the-art composites made with adhesive bonds commonly install redundant fasteners to guarantee the strength of these adhesively bonded composite parts. The number of fasteners in a single-aisle commercial transport aircraft is typically on the order of 105, which reduces manufacturing rate, increases cost tremendously, and reduces the advantage of the specific strength composites provide. Due to this, the Adhesive Free Bonding of Composites (AERoBOND) project at NASA Langley Research Center has developed a novel assembly process to manufacture complex composite parts without the use of adhesives and fasteners. However, optimization of the process is currently challenging due to the complex and interdependent process parameters. To assist with the optimization, four machine learning algorithms utilizing gradient boosting decision trees were created to provide predictions for the mechanical and characterization properties of the composite parts. Approximately 200 random states from each algorithm were tested, and the models from each state were isolated and analyzed based on their accuracy, a validation process, and their feature importance. This analysis concluded that the models created from the machine learning algorithms could accelerate a parametric study for the AERoBOND process by rapidly optimizing process parameters to achieve desired performance characteristics.

Brennen Michael Middleton↗
NASA NTRSDate not supplied

Machine Learning to Predict Joint Performance in Epoxy Composites Based on Process Parameters

Polymer matrix composites are gaining popularity in the aerospace industry due to their high specific strength, fatigue properties, and processability. However, based on current FAA certification guidelines, manufacturers utilizing current state-of-the art composites made with adhesive bonds commonly install redundant fasteners to guarantee the strength of these adhesively bonded composite parts.1,2 The number of fasteners in a single-aisle commercial transport aircraft is typically on the order of 105, which reduces manufacturing rate, increases cost tremendously, and reduces the advantage of the specific strength composites provide. Due to this, the Adhesive Free Bonding of Composites (AERoBOND) project at NASA Langley Research Center has developed a novel assembly process to manufacture complex composite parts without the use of adhesives and fasteners.1 However, optimization of the process is currently challenging due to the complex and interdependent process parameters. To assist with the optimization, four machine learning algorithms utilizing gradient boosting decision trees were created to provide predictions for the mechanical and characterization properties of the composite parts. Approximately 200 random states from each algorithm were tested, and the models from each state were isolated and analyzed based on their accuracy, a validation process, and their feature importance. This analysis concluded that the models created from the machine learning algorithms could accelerate a parametric study for the AERoBOND process by rapidly optimizing process parameters to achieve desired performance characteristics.

Brennen M Middleton↗
NASA NTRSDate not supplied

Uncertainty Quantification using Deep Ensembles for Decision Making in Cyber-Physical-Human Systems

In this paper and its companion, Differential Equation Approximation Using Gradient-Boosted Quantile Regression, Robison et al., we examine an approach to quantifying model uncertainty with the aim of increasing the trustworthiness of computational models in human-machine interactions. In Differential Equation Approximation Using Gradient-Boosted Quantile Regression, we focus on gradient-boosted decision trees, while in this one, we give more details about deep ensembles. Uncertainty quantification is crucial for building trustworthy autonomous decision-making agents in human-machine teams. There are two types of uncertainties: aleatoric and epistemic. The former is related to the inherent stochasticity (noise) of the process, whereas the latter is associated with the lack of knowledge or representation capability of models, such as neural networks. By lack of knowledge, we mean the model’s inability to accurately predict outputs for all possible inputs. The aleatory uncertainty can be estimated fairly easily with, for example, filters, whereas epistemic uncertainty is challenging to compute. This paper uses deep ensembles to quantify both aleatory and epistemic uncertainty. It can act as an uncertainty-aware surrogate transition model for decision-making frameworks. "Uncertainty-aware" means that the surrogate transition model should make predictions along with confidence in those predictions. In the context of decision-making, the transition models are ordinary differential equations (ODEs). Since ODEs can be simulated to make one-step or multi-step predictions, a good surrogate model for them should perform reasonably well in both modes. In a multi-step approach, the trajectory sampling method TS∞ was used to propagate uncertainty over multiple steps. The cartpole dynamical system was selected to demonstrate the ability of deep ensembles as good surrogate transition models for decision-making frameworks. The deep ensembles modeled the dynamics of cartpole ODEs and made uncertainty-aware predictions in single-step and multi-step transition modes.

CPH systems↗
NASA NTRS2022

A Comprehensive Machine Learning Study to Classify Precipitation Type over Land from Global Precipitation Measurement Microwave Imager (GPM-GMI) Measurements

Precipitation type is a key parameter used for better retrieval of precipitation characteristics as well as to understand the cloud–convection–precipitation coupling processes. Ice crystals and water droplets inherently exhibit different characteristics in different precipitation regimes (e.g., convection, stratiform), which reflect on satellite remote sensing measurements that help us distinguish them. The Global Precipitation Measurement (GPM) Core Observatory’s microwave imager (GMI) and dual-frequency precipitation radar (DPR) together provide ample information on global precipitation characteristics. As an active sensor, the DPR provides an accurate precipitation type assignment, while passive sensors such as the GMI are traditionally only used for empirical understanding of precipitation regimes. Using collocated precipitation type flags from the DPR as the “truth”, this paper employs machine learning (ML) models to train and test the predictability and accuracy of using passive GMI-only observations together with ancillary information from a reanalysis and GMI surface emissivity retrieval products. Out of six ML models, four simple ones (support vector machine, neural network, random forest, and gradient boosting) and the 1-D convolutional neural network (CNN) model are identified to produce 90–94% prediction accuracy globally for five types of precipitation (convective, stratiform, mixture, no precipitation, and other precipitation), which is much more robust than previous similar effort. One novelty of this work is to introduce data augmentation (subsampling and bootstrapping) to handle extremely unbalanced samples in each category. A careful evaluation of the impact matrices demonstrates that the polarization difference (PD), brightness temperature (Tc) and surface emissivity at high-frequency channels dominate the decision process, which is consistent with the physical understanding of polarized microwave radiative transfer over different surface types, as well as in snow and liquid clouds with different microphysical properties. Furthermore, the view-angle dependency artifact that the DPR’s precipitation flag bears with does not propagate into the conical-viewing GMI retrievals. This work provides a new and promising way for future physics-based ML retrieval algorithm development.

machine learning/artificial intelligence↗
NASA NTRSDate not supplied

Linking OH Variability to Observable Variables, Meteorology and Transport

The hydroxyl radical (OH) plays a vital role in tropospheric chemistry, as it provides the dominant sink for a multitude of pollutants and climate-relevant gases such as methane. Observational constraints on the global distribution and temporal variability of OH are limited, and models simulate a wide range of OH distributions. While OH itself has a short atmospheric lifetime, OH is photochemically coupled to longer-lived species that undergo atmospheric transport. Here, we investigate how much of the OH variability within and between models can be explained by differences in observable species to develop diagnostics for OH differences. We find that NO 2 and water vapor together explain much of the spatial and temporal variability in simulated OH, and we use satellite observations to identify biases in these variables. The OH response to ENSO also differs between models, and we investigate potential causes of these differences such as differences in convection or lightning NOx. We also explore the potential of idealized tracers to represent the OH distribution. Within a single model, meteorological variables such as humidity and idealized tracers of transport can explain a significant portion of the OH spatial variability. We use a Gradient Boosted Regression Trees, a type of machine learning, to account for non-linear relationships between OH and the input variables.

Meteorology↗
NASA NTRS2017

Prediction of Cognitive States During Flight Simulation Using Multimodal Psychophysiological Sensing

The Commercial Aviation Safety Team found the majority of recent international commercial aviation accidents attributable to loss of control inflight involved flight crew loss of airplane state awareness (ASA), and distraction was involved in all of them. Research on attention-related human performance limiting states (AHPLS) such as channelized attention, diverted attention, startle/surprise, and confirmation bias, has been recommended in a Safety Enhancement (SE) entitled "Training for Attention Management." To accomplish the detection of such cognitive and psychophysiological states, a broad suite of sensors was implemented to simultaneously measure their physiological markers during a high fidelity flight simulation human subject study. Twenty-four pilot participants were asked to wear the sensors while they performed benchmark tasks and motion-based flight scenarios designed to induce AHPLS. Pattern classification was employed to predict the occurrence of AHPLS during flight simulation also designed to induce those states. Classifier training data were collected during performance of the benchmark tasks. Multimodal classification was performed, using pre-processed electroencephalography, galvanic skin response, electrocardiogram, and respiration signals as input features. A combination of one, some or all modalities were used. Extreme gradient boosting, random forest and two support vector machine classifiers were implemented. The best accuracy for each modality-classifier combination is reported. Results using a select set of features and using the full set of available features are presented. Further, results are presented for training one classifier with the combined features and for training multiple classifiers with features from each modality separately. Using the select set of features and combined training, multistate prediction accuracy averaged 0.64 +/- 0.14 across thirteen participants and was significantly higher than that for the separate training case. These results support the goal of demonstrating simultaneous real-time classification of multiple states using multiple sensing modalities in high fidelity flight simulators. This detection is intended to support and inform training methods under development to mitigate the loss of ASA and thus reduce accidents and incidents.

Harrivel, Angela R.↗
NASA NTRS2022

Hybrid data-driven and model-informed online tool wear detection in milling machines

Precision machining tool wear is responsible for low product throughput and quality. Monitoring the tool wear online is vital to prevent degradation in machining quality. However, direct real-time tool wear measurement is not practical. This paper presents residual-based anomaly detection models, combining a hybrid model comprised of a physics-based model and a data-driven model (a decision tree or a neural network) to predict signals of interest (e.g., power or forces) under nominal conditions, followed by Page’s cumulative sum test for detecting tool wear on-line using the computer numerical control machine measurements. The most informative features are ranked using dynamic programming and its approximation variants from real-time measurements and machine settings, such as the width of cut, depth of cut, feed rate and spindle speed, that serve as inputs to the predictive models. The baseline nominal model is incrementally updated with experimental data via a gradient boosted adaptation model to generate the residuals that account for discrepancies between the actual machine data under normal conditions and the baseline nominal model predictions. The hybrid model is validated against 20 Mazak milling machine experimental tests and one Haas run-to-failure experiment. The proposed anomaly detector is applied to synthetic data from simulations of the physics-based model at different operating conditions, measurement noise levels, and tool wear levels, and the methods were able to achieve an overall 92% accuracy in data with 1% noise. The anomaly detection methods based on hybrid model reduced the false alarms of either the data-driven or physical-based models alone, and are found to be capable of good online detection of tool wear.

Online anomaly detection↗
NASA NTRS2024

Machine Learning for the Prediction of Local Asteroid Damages

Risk assessment studies of local asteroid hazards traditionally simulate the physics of meteors with engineering models tailored to analyze tens-of-millions of scenarios. However, these simplified approaches still need to solve time-dependent ODEs to model the entry process and the resulting ground damage. With a computational cost of O(0.01 CPU.s) per scenario, simulating these large numbers of potential entry conditions in risk assessment studies can take several days on local computers. To improve computational efficiency, we propose in this paper an orthogonal approach based on machine learning models to predict the size of damaged areas given a list of entry parameters. We train 5 machine learning methods and compare the predictions to the outputs of the PAIR model, first only with primitive entry condition variables, and then with more advanced features. We find that complex models like neural networks are well-suited to estimate blast hazards, while simpler linear models can accurately assess thermal damage. For both types of hazards, the radii of damaged areas can be predicted with around 10% average errors and a coefficient of determination (R2) of 0.99. The CPU time is decreased by a factor O(10 3 ) compared to the PAIR model, which enables the simulation of millions of scenarios in minutes, on a local computer. We then use the same machine learning approaches for a classification task where the models are trained to predict if an asteroid will produce a given level of damage. Results show that complex models like the gradient boosting classifier and the neural network can perform this task with 98% accuracy. Beyond surrogate models, we finally incorporate the machine learning algorithms to the state-of-the-art Shapley sensitivity analysis and present a ranking of the entry parameters based on their contributions to ground damages.

SMD↗
NASA NTRS2020

Advancing Methodologies for Applying Machine Learning and Evaluating Spatiotemporal Models of Fine Particulate Matter (PM 2.5 ) Using Satellite Data Over Large Regions

Reconstructing the distribution of fine particulate matter (PM 2.5 ) in space and time, even far from ground monitoring sites, is an important exposure science contribution to epidemiologic analyses of PM 2.5 health impacts. Flexible statistical methods for prediction have demonstrated the integration of satellite observations with other predictors, yet these algorithms are susceptible to overfitting the spatiotemporal structure of the training datasets. We present a new approach for predicting PM 2.5 using machine-learning methods and evaluating prediction models for the goal of making predictions where they were not previously available. We apply extreme gradient boosting (XGBoost) modeling to predict daily PM 2.5 on a 1 x 1 km 2 resolution for a 13 state region in the Northeastern USA for the years 2000–2015 using satellite-derived aerosol optical depth and implement a recursive feature selection to develop a parsimonious model. We demonstrate excellent predictions of withheld observations but also contrast an RMSE of 3.11 μg/m 3 in our spatial cross-validation withholding nearby sites versus an overfit RMSE of 2.10 μg/m 3 using a more conventional random ten-fold splitting of the dataset. As the field of exposure science moves forward with the use of advanced machine-learning approaches for spatiotemporal modeling of air pollutants, our results show the importance of addressing data leakage in training, overfitting to spatiotemporal structure, and the impact of the predominance of ground monitoring sites in dense urban sub-networks on model evaluation. The strengths of our resultant modeling approach for exposure in epidemiologic studies of PM 2.5 include improved efficiency, parsimony, and interpretability with robust validation while still accommodating complex spatiotemporal relationships.

air pollution↗
NASA NTRS2023

Reconstructing PM 2.5 Data Record for the Kathmandu Valley Using a Machine Learning Model

This paper presents a method for reconstructing the historical hourly concentrations of particulate matter 2.5 (PM2.5) over the Kathmandu Valley from 1980 to the present. The method uses a machine learning model that is trained using PM2.5 readings from US Embassy (Phora Durbar) as a ground truth, and the meteorological data from Modern-Era Retrospective Analysis for Research and Applications v2 (MERRA2) as input. The Extreme Gradient Boosting (XGBoost) model acquires a credible 10-fold cross-validation (CV) score of ~83.4%, an r2-score of ~84%, a Root Mean Square Error (RMSE) of ~15.82 µg/m3, and a Mean Absolute Error (MAE) of ~10.27 µg/m3. Further demonstrating the model's applicability to years other than those for which truth values are unavailable, the multiple cross-test with an unseen data set offered r2-scores for 2018, 2019, and 2020 ranging from 56% to 67%. The model-predicted data agrees with true values and indicates that MERRA2 underestimates PM2.5 over the region. It strongly agrees with ground-based evidence showing substantially higher mass concentrations in the dry pre- and post-monsoon seasons than in the monsoon months. It also shows a strong anti-correlation between PM2.5 concentration and humidity. The results also demonstrate that none of the years fulfilled the annual mean air quality index (AQI) standards set by the World Health Organization (WHO).

machine learning↗
NASA NTRS2021

Decoding the effects of synonymous variants

Synonymous single nucleotide variants (sSNVs) are common in the human genome but are often overlooked. However, sSNVs can have significant biological impact and may lead to disease. Existing computational methods for evaluating the effect of sSNVs suffer from the lack of gold-standard training/evaluation data and exhibit over-reliance on sequence conservation signals. We developed synVep (synonymous Variant effect predictor), a machine learning-based method that overcomes both of these limitations. Our training data was a combination of variants reported by gnomAD (observed) and those unreported, but possible in the human genome (generated). We used positive-unlabeled learning to purify the generated variant set of any likely unobservable variants. We then trained two sequential extreme gradient boosting models to identify subsets of the remaining variants putatively enriched and depleted in effect. Our method attained 90% precision/recall on a previously unseen set of variants. Furthermore, although synVep does not explicitly use conservation, its scores correlated with evolutionary distances between orthologs in cross-species variation analysis. synVep was also able to differentiate pathogenic vs. benign variants, as well as splice-site disrupting variants (SDV) vs. non-SDVs. Thus, synVep provides an important improvement in annotation of sSNVs, allowing users to focus on variants that most likely harbor effects.

Zishuo Zeng↗
NASA NTRS2019

Accelerated Simulation of Air Pollution Using NVIDIA RAPIDS

Atmospheric chemistry models are a central tool to study and forecast the impact of air pollution on the environment, vegetation, and human health. However, the numerical simulation of chemical kinetics is computationally expensive due to the stiffness of the system of ordinary differential equations that describes atmospheric chemistry. Here we present an alternative approach to the computation of atmospheric chemistry based on machine learning. Our training data set is produced using the NASA Goddard Earth Observing System (GEOS) model with GEOS-Chem chemistry, run on the NASA Center for Climate Simulation (NCCS) Discover supercomputing cluster on 384 Intel Xeon Haswell cores. This model spends more than 50% of total run time on solving atmospheric chemistry. The data set contains as input features the air pollution concentrations before solving the differential equations, together with some key physical parameters such as temperature and sun intensity. As target variables we define the air pollution concentrations after solving the differential equations. Using Dask-cuDF and Dask-XGBoost on the NVIDIA RAPIDS platform on 8 Tesla V100 GPUs, we generate from this training set gradient boosted decision tree models that can reproduce the simulation of chemical kinetics. We do this on the NCCS Advanced Data Analytics Platform (ADAPT) science cloud environment. Our application takes full advantage of recent advances in Dask-XGBoost, such as multi-node and multi-GPU scaling for distributed training with large data sets. The increase in training data size enabled by this is critical to capture the full range of chemical environments encountered across the globe and all annual seasons.The boosted tree models offer good predictability and show many of the features of the full chemistry reference simulation. Further improvements can be achieved through mass balance considerations and by accounting for error correlations. We incorporate the boosted tree models into the GEOS reference model using XGBoost's C API. This enables a seamless integration of the GPU trained models into GEOS-Chem, which is written in Fortran and optimized for use in a massively parallel CPU environment. We show the benefits of this approach and discuss the potential speedup of this machine learning accelerated atmospheric chemistry model.

Keller, Christoph A.↗
NASA NTRS2023

Large-Scale Groundwater Monitoring in Brazil Assisted With Satellite-Based Artificial Intelligence Techniques

Here, we develop and test an artificial intelligence (AI)-based approach to monitor major Brazilian aquifers. The approach combines Gravity Recovery and Climate Experiment (GRACE) data and ground-based hydrogeological measurements from Brazil’s Integrated Groundwater Monitoring Network at hundreds of wells distributed in twelve aquifers across the country. We tested model ensembles based on three AI approaches: Extreme Gradient Boost, Light Gradient Boosting Model and CatBoost, followed by a Linear Regression (LR) step. The approach is further boosted with wavelet and seasonal decomposition processes applied to GRACE data. To determine the AI-based model’s sensitivity to data availability, we propose four experiments combining hydrogeological measurements from different aquifers. Groundwater storage estimates from the Global Land Data Assimilation System (GLDAS) are used as benchmark. A sensitivity analysis shows that the LR-based model ensemble is the best suited and to reproduce groundwater storage change in all studied Brazilian aquifers. Results show that the proposed approach outperforms GLDAS in all experiments, with an RMSE value of 2.68cm for the experiment that covers all monitored wells in Brazil. GLDAS resulted in RMSE=6.76cm. Using our AI model outputs, we quantified the groundwater storage change of two major aquifers, Urucuia and Bauru-Caiuá, over the past two decades: -31km 3 and -6km 3 , respectively. Water loss is driven by a prolonged drought across most of the country and intensification of groundwater pumping for irrigation. This study demonstrates that combining satellite data and AI can be a cost-effective alternative to monitor poorly equipped aquifers at the continental scale, with possible global replicability.

GRACE↗
NASA NTRSDate not supplied

A Novel Machine Learning Method for Surface PM2.5 Estimations from Geostationary Satellites

Particulate matter (PM) with a diameter of less or equal to 2.5 μm, known as PM , affects human health as it penetrates the respiratory system. The Environmental Protection Agency (EPA) measures the atmospheric concentration of PM using air quality monitors stationed throughout the Continental United States (CONUS). Such measurements are points on a spatial domain and therefore, might not be representative of the air quality at nearby areas considering that the composition of the atmosphere is highly variable from place to place. Satellite based AOD permits a spatially uniform means of estimating PM and new geostationary satellites provide high temporal and spatial resolution estimation of AOD. However, the concentration of PM is non-linearly dependent on other atmospheric parameters that include relative humidity, temperature, and height of the planetary boundary layer. This information may be estimated at similar spatial and temporal resolutions as AOD from numerical modeling such as from the National Oceanic and Atmospheric Administration’s (NOAA) High Resolution Rapid Refresh (HRRR) model which resolves near real-time atmospheric conditions over the CONUS. The estimation of PM concentration is a multi-parametric problem that considers the effect of temporal dependencies among the different parameters. Deep learning approaches are appropriate for such complex estimation problems as they intrinsically capture relations among multiple non-linear parameters. This study compares deep-learning methods to traditional regression analysis to demonstrate the capabilities of these methods in predicting PM2.5 concentrations. Additionally, a novel ensemble learning approach is employed to identify scientific processes that could further improve the estimation of PM concentration. Utilizing Long Short-Term Memory (LSTM) neural networks, which are suitable for multivariate time series estimation problems as they are capable of learning long-term dependencies, individual models are created for each EPA station and trained on the aforementioned dataset collocated over each station. Individual station models are merged if the model's performance is improved by reducing the root mean squared error (RMSE) metric. This ensemble training method ultimately reduces the RMSE value. Evaluation of these results provide insights into physical processes and related observable parameters that may contribute to PM concentrations. Identified parameters evaluated to be statistically different between the merged and unmerged models are expected to improve overall performance. These new parameters are then utilized for reevaluation of the deep learning methods with an extreme gradient boosting model with an RMSE of 5.5 providing the best results.

George Priftis↗