scholarly journals Calibration of Probabilistic Quantitative Precipitation Forecasts with an Artificial Neural Network

2007 ◽  
Vol 22 (6) ◽  
pp. 1287-1303 ◽  
Author(s):  
Huiling Yuan ◽  
Xiaogang Gao ◽  
Steven L. Mullen ◽  
Soroosh Sorooshian ◽  
Jun Du ◽  
...  
Keyword(s):  
Neural Network ◽  
Great Basin ◽  
Joint Probability ◽  
Stage Iv ◽  
Skill Score ◽  
Precipitation Forecast ◽  
Brier Score ◽  
Joint Probability Distribution ◽  
Feed Forward Neural Network ◽  
Precipitation Thresholds

Abstract A feed-forward neural network is configured to calibrate the bias of a high-resolution probabilistic quantitative precipitation forecast (PQPF) produced by a 12-km version of the NCEP Regional Spectral Model (RSM) ensemble forecast system. Twice-daily forecasts during the 2002–2003 cool season (1 November–31 March, inclusive) are run over four U.S. Geological Survey (USGS) hydrologic unit regions of the southwest United States. Calibration is performed via a cross-validation procedure, where four months are used for training and the excluded month is used for testing. The PQPFs before and after the calibration over a hydrological unit region are evaluated by comparing the joint probability distribution of forecasts and observations. Verification is performed on the 4-km stage IV grid, which is used as “truth.” The calibration procedure improves the Brier score (BrS), conditional bias (reliability) and forecast skill, such as the Brier skill score (BrSS) and the ranked probability skill score (RPSS), relative to the sample frequency for all geographic regions and most precipitation thresholds. However, the procedure degrades the resolution of the PQPFs by systematically producing more forecasts with low nonzero forecast probabilities that drive the forecast distribution closer to the climatology of the training sample. The problem of degrading the resolution is most severe over the Colorado River basin and the Great Basin for relatively high precipitation thresholds where the sample of observed events is relatively small.

scholarly journals A Statistical Model for the Uncertainty Analysis of Satellite Precipitation Products

2015 ◽  
Vol 16 (5) ◽  
pp. 2101-2117 ◽  
Author(s):  
Sepideh Sarachi ◽  
Kuo-lin Hsu ◽  
Soroosh Sorooshian
Keyword(s):  
Southeastern United States ◽  
Joint Probability ◽  
Stage Iv ◽  
Joint Probability Distribution ◽  
Precipitation Estimation ◽  
Distribution Parameters ◽  
Generalized Normal Distribution ◽  
High Degree ◽  
Temporal Coverage ◽  
Hydrological Applications

Abstract Earth-observing satellites provide a method to measure precipitation from space with good spatial and temporal coverage, but these estimates have a high degree of uncertainty associated with them. Understanding and quantifying the uncertainty of the satellite estimates can be very beneficial when using these precipitation products in hydrological applications. In this study, the generalized normal distribution (GND) model is used to model the uncertainty of the Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks (PERSIANN) precipitation product. The stage IV Multisensor Precipitation Estimator (radar-based product) was used as the reference measurement. The distribution parameters of the GND model are further extended across various rainfall rates and spatial and temporal resolutions. The GND model is calibrated for an area of 5° × 5° over the southeastern United States for both summer and winter seasons from 2004 to 2009. The GND model is used to represent the joint probability distribution of satellite (PERSIANN) and radar (stage IV) rainfall. The method is further investigated for the period of 2006–08 over the Illinois watershed south of Siloam Springs, Arkansas. Results show that, using the proposed method, the estimation of the precipitation is improved in terms of percent bias and root-mean-square error.

scholarly journals Machine Learning Applications to Kronian Magnetospheric Reconnection Classification

2021 ◽  
Vol 7 ◽  
Author(s):  
Tadhg M. Garton ◽  
Caitriona M. Jackman ◽  
Andrew W. Smith ◽  
Kiley L. Yeakel ◽  
Shane A. Maloney ◽  
...  
Keyword(s):  
Neural Network ◽  
Magnetic Field ◽  
Magnetic Reconnection ◽  
Network Model ◽  
Neural Network Model ◽  
Skill Score ◽  
Feed Forward Neural Network ◽  
The North ◽  
Machine Learning Applications ◽  
High Level

The products of magnetic reconnection in Saturn’s magnetotail are identified in magnetometer observations primarily through characteristic deviations in the north–south component of the magnetic field. These magnetic deflections are caused by traveling plasma structures created during reconnection rapidly passing over the observing spacecraft. Identification of these signatures have long been performed by eye, and more recently through semi-automated methods, however these methods are often limited through a required human verification step. Here, we present a fully automated, supervised learning, feed forward neural network model to identify evidence of reconnection in the Kronian magnetosphere with the three magnetic field components observed by the Cassini spacecraft in Kronocentric radial–theta–phi coordinates as input. This model is constructed from a catalog of reconnection events which covers three years of observations with a total of 2093 classified events, categorized into plasmoids, traveling compression regions and dipolarizations. This neural network model is capable of rapidly identifying reconnection events in large time-span Cassini datasets, tested against the full year 2010 with a high level of accuracy (87%), true skill score (0.76), and Heidke skill score (0.73). From this model, a full cataloging and examination of magnetic reconnection events in the Kronian magnetosphere across Cassini's near Saturn lifetime is now possible.

scholarly journals A fusion of data science and feed-forward neural network-based modelling of COVID-19 outbreak forecasting in IRAQ

2021 ◽  
Vol 118 ◽  
pp. 103766
Author(s):  
Ahmed J. Aljaaf ◽  
Thakir M. Mohsin ◽  
Dhiya Al-Jumeily ◽  
Mohamed Alloghani
Keyword(s):  
Neural Network ◽  
Data Science ◽  
Feed Forward Neural Network ◽  
Feed Forward

scholarly journals Graph neural networks for an accurate and interpretable prediction of the properties of polycrystalline materials

2021 ◽  
Vol 7 (1) ◽  
Author(s):  
Minyi Dai ◽  
Mehmet F. Demirel ◽  
Yingyu Liang ◽  
Jia-Mian Hu
Keyword(s):  
Neural Network ◽  
Network Architecture ◽  
Polycrystalline Materials ◽  
Feed Forward Neural Network ◽  
Microscopic Interactions ◽  
Physical Interactions ◽  
Graph Neural Networks ◽  
Individual Grains ◽  
Physical Features ◽  
Machine Learning Models

AbstractVarious machine learning models have been used to predict the properties of polycrystalline materials, but none of them directly consider the physical interactions among neighboring grains despite such microscopic interactions critically determining macroscopic material properties. Here, we develop a graph neural network (GNN) model for obtaining an embedding of polycrystalline microstructure which incorporates not only the physical features of individual grains but also their interactions. The embedding is then linked to the target property using a feed-forward neural network. Using the magnetostriction of polycrystalline Tb0.3Dy0.7Fe2 alloys as an example, we show that a single GNN model with fixed network architecture and hyperparameters allows for a low prediction error of ~10% over a group of remarkably different microstructures as well as quantifying the importance of each feature in each grain of a microstructure to its magnetostriction. Such a microstructure-graph-based GNN model, therefore, enables an accurate and interpretable prediction of the properties of polycrystalline materials.

scholarly journals Probability elicitation using geostatistics in hydrocarbon exploration

2021 ◽  
Author(s):  
André Luís Morosov ◽  
Reidar Brumer Bratvold
Keyword(s):  
Value Creation ◽  
Decision Process ◽  
Conceptual Models ◽  
Joint Probability ◽  
Decision Makers ◽  
Hydrocarbon Exploration ◽  
Joint Probability Distribution ◽  
Proof Of Concept ◽  
Geological Information ◽  
Decision Supporting

AbstractThe exploratory phase of a hydrocarbon field is a period when decision-supporting information is scarce while the drilling stakes are high. Each new prospect drilled brings more knowledge about the area and might reveal reserves, hence choosing such prospect is essential for value creation. Drilling decisions must be made under uncertainty as the available geological information is limited and probability elicitation from geoscience experts is key in this process. This work proposes a novel use of geostatistics to help experts elicit geological probabilities more objectively, especially useful during the exploratory phase. The approach is simpler, more consistent with geologic knowledge, more comfortable for geoscientists to use and, more comprehensive for decision-makers to follow when compared to traditional methods. It is also flexible by working with any amount and type of information available. The workflow takes as input conceptual models describing the geology and uses geostatistics to generate spatial variability of geological properties in the vicinity of potential drilling prospects. The output is stochastic realizations which are processed into a joint probability distribution (JPD) containing all conditional probabilities of the process. Input models are interactively changed until the JPD satisfactory represents the expert’s beliefs. A 2D, yet realistic, implementation of the workflow is used as a proof of concept, demonstrating that even simple modeling might suffice for decision-making support. Derivative versions of the JPD are created and their effect on the decision process of selecting the drilling sequence is assessed. The findings from the method application suggest ways to define the input parameters by observing how they affect the JPD and the decision process.

Sign in / Sign up

Export Citation Format

Share Document