Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed algorithm”

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 577 records · Page 32

Optimizing High Performance Markov Clustering for Pre-Exascale Architectures

HipMCL is a high-performance distributed memory implementation of the popular Markov Cluster Algorithm (MCL) and can cluster large-scale networks within hours using a few thousand CPU-equipped nodes. It relies on sparse matrix computations and heavily makes use of the sparse matrix-sparse matrix multiplication kernel (SpGEMM). The existing parallel algorithms in HipMCL are not scalable to Exascale architectures, both due to their communication costs dominating the runtime at large concurrencies and also due to their inability to take advantage of accelerators that are increasingly popular. In this work, we systematically remove scalability and performance bottlenecks of HipMCL. We enable GPUs by performing the expensive expansion phase of the MCL algorithm on GPU. Additionally, we propose a CPU-GPU joint distributed SpGEMM algorithm called pipelined Sparse SUMMA and integrate a probabilistic memory requirement estimator that is fast and accurate. Furthermore, we develop a new merging algorithm for the incremental processing of partial results produced by the GPUs, which improves the overlap efficiency and the peak memory usage. We also integrate a recent and faster algorithm for performing SpGEMM on CPUs. We validate our new algorithms and optimizations with extensive evaluations. With the enabling of the GPUs and integration of new algorithms, HipMCL is up to 12.4x faster, being able to cluster a network with 70 million proteins and 68 billion connections just under 15 minutes using 1024 nodes of ORNL's Summit supercomputer.

97 MATHEMATICS AND COMPUTING↗

Tests of Radar Rainfall Retrieval Algorithms

The NASA/JPL Airborne Rain Mapping Radar (ARMAR) operates at 14 GHz. ARMAR flew on the NASA DC-8 aircraft during Tropical Ocean Global Atmosphere (TOGA) Coupled Ocean Atmosphere Response Experiment (COARE), collecting data in oceanic mesoscale convective systems, similar to those now being observed by the Tropical Rainfall Measuring Mission (TRMM) satellite, which includes a 14-GHz precipitation radar. Several algorithms for retrieving rain rate from downward looking radars are in existence. These can be categorized as deterministic and stochastic. Deterministic algorithms use the path integrated attenuation (PIA), measured by the surface reference technique, as a constraint. One deterministic algorithm corrects the attenuation-rainfall (k-R) relation, while another corrects the reflectivity rainfall (ZR) relation. Stochastic algorithms apply an Extended Kalman Filter to the reflectivity profile. One employs radar reflectivity only; the other additionally uses the PIA. We find that the stochastic algorithm with PIA is the most robust algorithm with regard to incorrect assumptions about the drop-size distribution (DSD). The deterministic algorithm that uses the PIA to adjust the Z-R relation is also fairly robust and produces rain rates similar to the stochastic algorithm that uses the PIA. The deterministic algorithm that adjusts only the k-R relation and the stochastic radar-only algorithm are more sensitive to assumptions about the DSD. It is likely that they underestimate convective rainfall, especially if the DSD is erroneously assumed to be appropriate for stratiform rain conditions. The underestimation is illustrated in the diagram. The algorithm labeled D IS initially assumes a DSD that is appropriate for stratiform. rain, while the rain is most likely convective. The PIA constraint causes the k-R relation to be adjusted, resulting in a much lower rain rate than the other algorithms. Additional information is contained in the original.

Durden, Stephen L.↗

Distributed-Memory Computing With the Langley Aerothermodynamic Upwind Relaxation Algorithm (LAURA)

The Langley Aerothermodynamic Upwind Relaxation Algorithm (LAURA), a Navier-Stokes solver, has been modified for use in a parallel, distributed-memory environment using the Message-Passing Interface (MPI) standard. A standard domain decomposition strategy is used in which the computational domain is divided into subdomains with each subdomain assigned to a processor. Performance is examined on dedicated parallel machines and a network of desktop workstations. The effect of domain decomposition and frequency of boundary updates on performance and convergence is also examined for several realistic configurations and conditions typical of large-scale computational fluid dynamic analysis.

Riley, Christopher J.↗

Mueller-matrix mapping of biological tissues in differential diagnosis of optical anisotropy mechanisms of protein networks

We report a model of Mueller-matrix description of optical anisotropy of protein networks in biological tissues with allowance for the linear birefringence and dichroism. The model is used to construct the reconstruction algorithms of coordinate distributions of phase shifts and the linear dichroism coefficient. In the statistical analysis of such distributions, we have found the objective criteria of differentiation between benign and malignant tissues of the female reproductive system. From the standpoint of evidence-based medicine, we have determined the operating characteristics (sensitivity, specificity and accuracy) of the Mueller-matrix reconstruction method of optical anisotropy parameters and demonstrated its effectiveness in the differentiation of benign and malignant tumours. (laser applications and other topics in quantum electronics)

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Management of Pulmonary Arterial Hypertension in the Pediatric Patient

Purpose of Review: Pediatric pulmonary arterial hypertension (PAH) is associated with significant morbidity and mortality. Herein we review the diagnosis and classification for pediatric PAH and detail the current therapeutic options available for use in the pediatric PAH population. Recent Findings: Classification and treatment of pediatric PAH is guided by adult criteria and treatment algorithms, yet the distribution of factors contributing to PAH in children differs significantly from that seen in adults. It is necessary to understand these differences in order to appropriately tailor therapy to the needs of the child or adolescent. An expanding array of targeted PAH drugs are now approved for use in adults, and many of these drugs are used “off-label” to treat children and adolescents with PAH. Use of these novel therapies has coincided with marked improvement in outcomes, suggesting significant benefit. However, because most of these drugs have not been studied in rigorous randomized, controlled trials in children, it is critical that physicians understand their mechanisms of action, potential benefits, and safety profiles. Summary: Pediatric PAH outcomes have improved substantially in the modern era, coinciding with the “off-label” use of targeted PAH drugs in children and adolescents. Ideally, care should be provided at centers with specialized expertise in the diagnosis and treatment of pediatric PAH by providers who understand the appropriate diagnostic algorithms, classification schemes, and treatment approaches.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Distributed Minimal Residual (DMR) method for acceleration of iterative algorithms

A new method for enhancing the convergence rate of iterative algorithms for the numerical integration of systems of partial differential equations was developed. It is termed the Distributed Minimal Residual (DMR) method and it is based on general Krylov subspace methods. The DMR method differs from the Krylov subspace methods by the fact that the iterative acceleration factors are different from equation to equation in the system. At the same time, the DMR method can be viewed as an incomplete Newton iteration method. The DMR method was applied to Euler equations of gas dynamics and incompressible Navier-Stokes equations. All numerical test cases were obtained using either explicit four stage Runge-Kutta or Euler implicit time integration. The formulation for the DMR method is general in nature and can be applied to explicit and implicit iterative algorithms for arbitrary systems of partial differential equations.

Lee, Seungsoo↗

Localization of Ad-Hoc Lunar Constellations in Communication Failure Modes for Distributed Spacecraft Autonomy

As lunar missions increase in complexity inspired by NASA’s Artemis Program, they will require reliable and sufficient capability of the Position, Navigation, and Timing (PNT) system to support their scientific objectives. In addition, NASA's Commercial Lunar Payload Services (CLPS) program initiates the proliferation of public and private exploration partnerships using small satellites from commercial and private organizations, expanding traditionally confined low Earth orbit to be used for missions beyond geosynchronous orbit (Zucherman et al., 2022). Therefore, the Lunar PNT system is also required to provide navigation services compatible with the smaller platforms being sent by the public and private sectors, like CubeSats. However, traditional approaches to deep space missions’ navigation based on ground radio facilities have difficulties in providing sufficient support for the increasing number of users and communication at a distance from the Earth (Kaplev et al., 2022). In particular, the existing Lunar navigation technologies such as weak signal global positioning system (GPS) and deep space network (DSN) are not able to ensure operations of the upcoming small-scale Lunar missions due to their limitations in localization performance as well as capacity aspects. Another way to provide Lunar PNT service is to create a dedicated Lunar global navigation satellite system (GNSS) constellation, like GNSS systems on Earth. Space agencies like NASA, ESA, and JAXA are now developing the lunar communications relay and navigation systems (LCRNS) and Lunar navigation satellite systems (LNSS). In their systems, satellites will be deployed in moon orbits to provide the communication, positioning, navigation, and timing (CPNT) service at the lunar south pole region where the Artemis base camp will be expected (Murata et al., 2022). Meanwhile, common challenges considered in lunar PNT research arise from poor geometry of the terrestrial GNSS satellites when seen from the lunar user, highly perturbed lunar orbits, and limitations in power, size, and cost of the equipment on lunar satellites (Iiyama et al., 2023). It is also not clear if there will be enough Lunar users to support the cost and resources this would require as the Low-cost surface missions may not be able to support the large power, mass, and weight requirements that these navigation solutions entail (Niemoeller et al., 2022). As an alternative, existing Lunar science and exploration assets could be used to create a low-cost, autonomous, ad-hoc, and on-demand mission-centric Lunar PNT swarm capable of providing PNT services to these low-cost lunar missions (Hagenau et al., 2021). Introducing the non-dedicated and ad-hoc Lunar navigation constellation gives a way to provide PNT services on-demand. The non-dedicated swarm assets of Lunar constellations are designed to localize themselves with minimal interaction with Earth by adding cooperative autonomous localization to lunar missions, freeing up valuable bandwidth and ground segment resources. An autonomous localization of Lunar constellations is based on the concept of the decentralized PNT system with a distributed extended Kalman filter (DEKF) approach to state estimation for minimal onboard operating costs. In the distributed data processing algorithm, computation is broken down and assigned to each satellite, resulting in a considerably decreased computational amount while maintaining the accuracy of the orbit ephemeris and clock offsets as the result of centralized data processing (Wen et al., 2019). The DEKF requires spacecraft to perform two-way ranging operations with each other to communicate simultaneously, leveraging neighbor two-way intersatellite link (ISL) measurements such as pseudoranges to, and relative velocities between, visible satellites as sensor values (Frank et al., 2021). The Lunar autonomous PNT simulation (LAPS) demonstrated the feasibility of orbital asset localization among ad-hoc Lunar small-sat constellations based on the DEKF in Hagenau et al. (2021) and evaluated the matching algorithm proposed by Frank et al. (2021) in scheduling position estimation updates. In previous papers, all assets and measurements are assumed to be always available without consideration of the impact of intermittent and permanent communication failure. This study presents localization performance with increasing levels of network degradation for swarm assets and users to demonstrate the robustness of the decentralized Lunar PNT service in more realistic scenarios. Main issues arising from communication failure include spacecraft permanent or transient loss, antenna failures, message delays, etc. We tested four possible reasons for network degradation for 7 days in 21 satellites frozen with an altitude of 5500 km, evenly spaced around 3 circular, 40 inclination orbital planes where each spacecraft has two directional antennas. As anchor nodes with an independent estimate of their position are required in the DEKF approach, two ground nodes in each pole and one node in the gateway were implemented in the simulation. First, the most probable failure scenario involves the loss of a single spacecraft due to solar interference and technical malfunctions of the assets. Losing the availability of a single spacecraft means losing the two-way ISL measurement of the asset in the DEKF update. In order to provide the best possible quality of PNT service with limited time and resources, the distributed Lunar constellations must schedule the communication activities. The scheduler leverages mixed-integer linear programming (MILP) for the coordination and scheduling of the desired “as-needed” localization service (Niemoeller et al., 2022). We assume the scheduler has completely excluded the spacecraft information before the DEKF update in the failure scenario. When a random spacecraft has been turned off at a specific time, the robustness of the autonomous Lunar PNT system is evaluated. The simulation results give an 11.5% degradation in median position accuracy compared to the idealized performance excluding the asset loss. Second, a large number of assets may vanish due to major hardware problems or meteor strikes around the moon. A multiple spacecraft loss can degrade the localization performance very fast by losing the communication ability to do cross-plane measurements and in-plane measurements in a 3-plane constellation. When the matching-based scheduler is aware of ISL availability, we investigate a large number of in-plane and cross-plane asset vanishments both in close proximity and equally spaced throughout the orbital plane. According to the simulations, the loss of in-plane measurements gives 40.2% degradation while cross-plane measurements degrade 50.5% of asset localization performance among available assets. Therefore, it is concluded that cross-plane measurements are more important in improving the position estimation accuracy. Third, spacecraft failure information can be lost due to the internal message delay, resulting in the DEKF update scheduler to solve the matching problem with unavailable assets. The DEKF update cycle is comprised of network setup, communication, and computations where a global broadcast network and a 2-way ISL network setup take 6 minutes in total (Frank et al., 2021). Once the broadcast network successfully transmits and receives information, a random spacecraft may lose its availability right before solving the matching problem. This means the matching solution is no longer optimal, resulting in degradation in the localization performance. A numerical assessment shows the matching-based scheduler with knowing failure holds 11.5% of position accuracy degradation, whereas the scheduler without knowing failure gives 34% degraded localization performance without asset loss. Fourth, a transient loss of a single or multiple spacecraft may occur due to their antenna outages. After losing the two-way ISL availability for a few DEKF update cycles, the availability of spacecraft can easily be recovered as their states have been independently updated using measurements from anchor nodes. It is likely that the longer failure will result in worse localization performance. We have tested the transient failure of a random single asset for 30 min in the simulation, which is losing 3 update cycles in the DEKF system. From the simulation results, the position accuracy has been degraded to 4.84% which is better than the degraded localization performance of 11.5% from the permanent loss scenario among available assets. In conclusion, the autonomous Lunar PNT system based on the DEKF approach shows the ability to maintain resilience and robustness in the possible communication failure scenarios, ensuring that localization accuracy is preserved across various network degradation and outages. Future studies on investigating user localization performance near the South Pole and the broadcast network system will be continued in the following months.

Yeji Kim↗

Bayesian seismic inversion based on rock-physics prior modeling for the joint estimation of acoustic impedance, porosity and lithofacies

We propose a Bayesian approach for seismic inversion to estimate acoustic impedance, porosity and lithofacies within the reservoir conditioned to post-stack seismic and well data. The link between elastic and petrophysical properties is given by a joint prior distribution for the logarithm of impedance and porosity, based on a rock-physics model. The well conditioning is performed through a background model obtained by well log interpolation. Two different approaches are presented: in the first approach, the prior is defined by a single Gaussian distribution, whereas in the second approach it is defined by a Gaussian mixture to represent the well data multimodal distribution and link the Gaussian components to different geological lithofacies. The forward model is based on a linearized convolutional model. For the single Gaussian case, we obtain an analytical expression for the posterior distribution, resulting in a fast algorithm to compute the solution of the inverse problem, i.e. the posterior distribution of acoustic impedance and porosity as well as the facies probability given the observed data. For the Gaussian mixture prior, it is not possible to obtain the distributions analytically, hence we propose a Gibbs algorithm to perform the posterior sampling and obtain several reservoir model realizations, allowing an uncertainty analysis of the estimated properties and lithofacies. Both methodologies are applied to a real seismic dataset with three wells to obtain 3D models of acoustic impedance, porosity and lithofacies. The methodologies are validated through a blind well test and compared to a standard Bayesian inversion approach. Using the probability of the reservoir lithofacies, we also compute a 3D isosurface probability model of the main oil reservoir in the studied field.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Comparison of Dose Distributions With TG-43 and Collapsed Cone Convolution Algorithms Applied to Accelerated Partial Breast Irradiation Patient Plans

Purpose: To compare the treatment plans for accelerated partial breast irradiation calculated by the new commercially available collapsed cone convolution (CCC) and current standard TG-43–based algorithms for 50 patients treated at our institution with either a Strut-Adjusted Volume Implant (SAVI) or Contura device. Methods and Materials: We recalculated target coverage, volume of highly dosed normal tissue, and dose to organs at risk (ribs, skin, and lung) with each algorithm. For 1 case an artificial air pocket was added to simulate 10% nonconformance. We performed a Wilcoxon signed rank test to determine the median differences in the clinical indices V90, V95, V100, V150, V200, and highest-dosed 0.1 cm{sup 3} and 1.0 cm{sup 3} of rib, skin, and lung between the two algorithms. Results: The CCC algorithm calculated lower values on average for all dose-volume histogram parameters. Across the entire patient cohort, the median difference in the clinical indices calculated by the 2 algorithms was <10% for dose to organs at risk, <5% for target volume coverage (V90, V95, and V100), and <4 cm{sup 3} for dose to normal breast tissue (V150 and V200). No discernable difference was seen in the nonconformance case. Conclusions: We found that on average over our patient population CCC calculated (<10%) lower doses than TG-43. These results should inform clinicians as they prepare for the transition to heterogeneous dose calculation algorithms and determine whether clinical tolerance limits warrant modification.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

WE-AB-207B-07: Dose Cloud: Generating “Big Data” for Radiation Therapy Treatment Plan Optimization Research

Purpose: To provide a tool to generate large sets of realistic virtual patient geometries and beamlet doses for treatment optimization research. This tool enables countless studies exploring the fundamental interplay between patient geometry, objective functions, weight selections, and achievable dose distributions for various algorithms and modalities. Methods: Generating realistic virtual patient geometries requires a small set of real patient data. We developed a normalized patient shape model (PSM) which captures organ and target contours in a correspondence-preserving manner. Using PSM-processed data, we perform principal component analysis (PCA) to extract major modes of variation from the population. These PCA modes can be shared without exposing patient information. The modes are re-combined with different weights to produce sets of realistic virtual patient contours. Because virtual patients lack imaging information, we developed a shape-based dose calculation (SBD) relying on the assumption that the region inside the body contour is water. SBD utilizes a 2D fluence-convolved scatter kernel, derived from Monte Carlo simulations, and can compute both full dose for a given set of fluence maps, or produce a dose matrix (dose per fluence pixel) for many modalities. Combining the shape model with SBD provides the data needed for treatment plan optimization research. Results: We used PSM to capture organ and target contours for 96 prostate cases, extracted the first 20 PCA modes, and generated 2048 virtual patient shapes by randomly sampling mode scores. Nearly half of the shapes were thrown out for failing anatomical checks, the remaining 1124 were used in computing dose matrices via SBD and a standard 7-beam protocol. As a proof of concept, and to generate data for later study, we performed fluence map optimization emphasizing PTV coverage. Conclusions: We successfully developed and tested a tool for creating customizable sets of virtual patients suitable for large-scale radiation therapy optimization research.

60 APPLIED LIFE SCIENCES↗

Post-extreme-event restoration using linear topological constraints and DER scheduling to enhance distribution system resilience

In this paper, a post-extreme-event restoration (PEER) algorithm is proposed to improve distribution system resilience. Linear topological constraints are proposed to ensure radial topology after N-k contingencies, possibly in multiple islands. The approach is made comprehensive by considering dispatchable distributed energy resources (DERs), non-dispatchable DERs, and demand responses, as well as on-load tap changers (OLTCs) and shunt capacitors. The goal is to minimize the accumulative expense caused by load reduction payment or penalty, as well as DER operation cost. As a result, the overall system will survive longer with higher resilience during an extreme event. To verify the effectiveness of the PEER algorithm, we proposed a resilience evaluation algorithm using Monte Carlo simulation (MCS) with reduced scenarios. This is based on a probabilistic model for generating random scenarios which consider the uncertainty of line faults and solar irradiance. Combined with the proposed PEER algorithm, this reduced-scenario MCS can evaluate the expected energy not served (EENS) which is an essential index for distribution system resilience. Case studies of the IEEE 33-bus and 123-bus test systems validate the proposed algorithm in reducing EENS.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Fully relativistic lattice Boltzmann algorithm

Starting from the Maxwell-Juettner equilibrium distribution, we develop a relativistic lattice Boltzmann (LB) algorithm capable of handling ultrarelativistic systems with flat, but expanding, spacetimes. The algorithm is validated through simulations of a quark-gluon plasma, yielding excellent agreement with hydrodynamic simulations. The present scheme opens the possibility of transferring the recognized computational advantages of lattice kinetic theory to the context of both weakly and ultrarelativistic systems.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Cyber Physical Grid-Interactive Distributed Energy Resources Control for VPP Dispatch and Regulation

This paper presents a cyber-physical algorithm for grid-interactive Distributed Energy Resource (DER) control to enable two features of Virtual Power Plants (VPPs) dispatch and grid voltage regulation, considering the communication and security impacts. We first formulate the DER dispatch problem as a real-time, iterative, and grid-interactive DER control problem. Thereafter, we consider a probabilistic traffic model to characterize packet delays and loss in a communication network, and study how the delays enter the process of information exchange among the grid measurement units, local DER controllers and the grid control center that coordinately execute this dispatch algorithm. Finally, a strategy combining delay threshold and modified message update rules is proposed to immune the asynchrony resulting from the communications network traffic and it avoids possible numerical instabilities and sensitivities of the tracking and regulation capabilities of this DER control algorithm. By implementing the proposed cyber-physical algorithm on the modified IEEE 37-node system, our preliminary results exhibit that the uncertainties of the underlying communications infrastructure must be considered for the VPP tracking and regulation capabilities of any DER in a generic Cyber-Physical System (CPS), because the delayed voltage measurements in the uplink/bi-link cases result in the off-track in VPP dispatch and jittery in voltage regulation.

cyber-physical algorithm↗

Radiation image reconstruction and uncertainty quantification using a Gaussian process prior

We propose a complete framework for Bayesian image reconstruction and uncertainty quantification based on a Gaussian process prior (GPP) to overcome limitations of maximum likelihood expectation maximization (ML-EM) image reconstruction algorithm. The prior distribution is constructed with a zero-mean Gaussian process (GP) with a choice of a covariance function, and a link function is used to map the Gaussian process to an image. Unlike many other maximum a posteriori approaches, our method offers highly interpretable hyperparamters that are selected automatically with the empirical Bayes method. Furthermore, the GP covariance function can be modified to incorporate a priori structural priors, enabling multi-modality imaging or contextual data fusion. Lastly, we illustrate that our approach lends itself to Bayesian uncertainty quantification techniques, such as the preconditioned Crank–Nicolson method and the Laplace approximation. The proposed framework is general and can be employed in most radiation image reconstruction problems, and we demonstrate it with simulated free-moving single detector radiation source imaging scenarios. We compare the reconstruction results from GPP and ML-EM, and show that the proposed method can significantly improve the image quality over ML-EM, all the while providing greater understanding of the source distribution via the uncertainty quantification capability. Furthermore, significant improvement of the image quality by incorporating a structural prior is illustrated.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Generalized Metropolis acceptance criterion for hybrid non-equilibrium molecular dynamics—Monte Carlo simulations

A family of hybrid simulation methods that combines the advantages of Monte Carlo (MC) with the strengths of classical molecular dynamics (MD) consists in carrying out short non-equilibrium MD (neMD) trajectories to generate new configurations that are subsequently accepted or rejected via an MC process. In the simplest case where a deterministic dynamic propagator is used to generate the neMD trajectories, the familiar Metropolis acceptance criterion based on the change in the total energy ΔE, min[1, exp( − βΔE)], guarantees that the hybrid algorithm will yield the equilibrium Boltzmann distribution. However, the functional form of the acceptance probability is more complex when the non-equilibrium switching process is generated via a non-deterministic stochastic dissipative propagator coupled to a heat bath. Here, we clarify the conditions under which the Metropolis criterion remains valid to rigorously yield a proper equilibrium Boltzmann distribution within hybrid neMD-MC algorithm.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A moment-convergence method for stochastic analysis of biochemical reaction networks

Traditional moment-closure methods need to assume that high-order cumulants of a probability distribution approximate to zero. However, this strong assumption is not satisfied for many biochemical reaction networks. Here, we introduce convergent moments (defined in mathematics as the coefficients in the Taylor expansion of the probability-generating function at some point) to overcome this drawback of the moment-closure methods. As such, we develop a new analysis method for stochastic chemical kinetics. This method provides an accurate approximation for the master probability equation (MPE). In particular, the connection between low-order convergent moments and rate constants can be more easily derived in terms of explicit and analytical forms, allowing insights that would be difficult to obtain through direct simulation or manipulation of the MPE. In addition, it provides an accurate and efficient way to compute steady-state or transient probability distribution, avoiding the algorithmic difficulty associated with stiffness of the MPE due to large differences in sizes of rate constants. Applications of the method to several systems reveal nontrivial stochastic mechanisms of gene expression dynamics, e.g., intrinsic fluctuations can induce transient bimodality and amplify transient signals, and slow switching between promoter states can increase fluctuations in spatially heterogeneous signals. The overall approach has broad applications in modeling, analysis, and computation of complex biochemical networks with intrinsic noise.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A linear merging methodology for high-resolution precipitation products using spatiotemporal regression

Currently, the only viable option for a global precipitation product is the merger of several precipitation products from different modalities. In this article, we develop a linear merging methodology based on spatiotemporal regression. Four highresolution precipitation products (HRPPs), obtained through methods including the Climate Prediction Center's Morphing (CMORPH), Geostationary Operational Environmental Satellite-Based Auto-Estimator (GOES-AE), GOES-Based Hydro-Estimator (GOES-HE) and Self-Calibrating Multivariate Precipitation Retrieval (SCAMPR) algorithms, are used in this study. The merged data are evaluated against the Arkansas Red Basin River Forecast Center's (ABRFC's) ground-based rainfall product. The evaluation is performed using the Heidke skill score (HSS) for four seasons, from summer 2007 to spring 2008, and for two different rainfall detection thresholds. It is shown that the merged data outperform all the other products in seven out of eight cases. A key innovation of this machine learning method is that only 6% of the validation data are used for the initial training. The sensitivity of the algorithm to location, distribution of training data, selection of input data sets and seasons is also analysed and presented.

Turlapaty, Anish C↗