Search

Google Vizier has performed millions of optimizations and accelerated numerous research and production systems at Google, demonstrating the success of Bayesian optimization as a large-scale service. Over multiple years, its algorithm has been improved considerably, through the collective experiences of numerous research efforts and user feedback. In this technical report, we discuss the implementation details and design choices of the current default algorithm provided by Open Source Vizier. Our experiments on standardized benchmarks reveal its robustness and versatility against well-established industry baselines on multiple practical modes., Comment: Google DeepMind Technical Report. Code can be found in https://github.com/google/vizier

This study explores the suitability of semi-insulating compounds, specifically (Cd,Mn)Te and (Cd,Mn)(Te,Se), as materials for room temperature X-ray and gamma-ray detectors. These compounds were grown using the Bridgman method, known for its efficient growth rate. The investigation aims to compare their crystal structure, mechanical properties, optical characteristics, and radiation detection capabilities. The addition of selenium to (Cd,Mn)Te increased the compound's hardness. However, (Cd,Mn)(Te,Se) exhibited one order of magnitude higher etch pit density compared to (Cd,Mn)Te. Photoluminescence analysis at low temperatures revealed the presence of defect states in both materials, characterized by shallow and deep donor-acceptor pair transitions (DAP). Annealing in cadmium vapors effectively eliminated DAP luminescence in (Cd,Mn)Te but not in (Cd,Mn)(Te,Se). Spectroscopic performance assessments indicated that the (Cd,Mn)Te detector outperformed the (Cd,Mn)(Te,Se) detector in responding to a Co-57 source. The reduced performance in the latter case may be attributed to either the presence of a deep trap related to deep DAP luminescence, minimally affected by annealing, or the dominant presence of block-like structures in the samples, as indicated by X-ray diffraction measurements. The block-like structures in (Cd,Mn)(Te,Se) showed ten times larger misorientation angles compared to the (Cd,Mn)Te crystals. (Cd,Mn)Te crystal revealed excellent single crystal properties, demonstrated by narrower omega scan widths. The study also highlights the influence of grain boundaries and twins on crystal structure quality. In our opinion, Bridgman-grown (Cd,Mn)Te shows greater promise as a material for X-ray and gamma-ray detectors compared to (Cd,Mn)(Te,Se)., Comment: 33 pages, 11 figures

Increases in wildfire activity and the resulting impacts have prompted the development of high-resolution wildfire behavior models for forecasting fire spread. Recent progress in using satellites to detect fire locations further provides the opportunity to use measurements to improve fire spread forecasts from numerical models through data assimilation. This work develops a method for inferring the history of a wildfire from satellite measurements, providing the necessary information to initialize coupled atmosphere-wildfire models from a measured wildfire state in a physics-informed approach. The fire arrival time, which is the time the fire reaches a given spatial location, acts as a succinct representation of the history of a wildfire. In this work, a conditional Wasserstein Generative Adversarial Network (cWGAN), trained with WRF-SFIRE simulations, is used to infer the fire arrival time from satellite active fire data. The cWGAN is used to produce samples of likely fire arrival times from the conditional distribution of arrival times given satellite active fire detections. Samples produced by the cWGAN are further used to assess the uncertainty of predictions. The cWGAN is tested on four California wildfires occurring between 2020 and 2022, and predictions for fire extent are compared against high resolution airborne infrared measurements. Further, the predicted ignition times are compared with reported ignition times. An average Sorensen's coefficient of 0.81 for the fire perimeters and an average ignition time error of 32 minutes suggest that the method is highly accurate.

The current fuel moisture content (FMC) subsystems in WRF-SFIRE and its workflow system WRFx use a time-lag differential equation model with assimilation of data from FMC sensors on Remote Automated Weather Stations (RAWS) by the extended augmented Kalman filter. But the quality of the result is constrained by the limitations of the model and of the Kalman filter. We observe that the data flow in a system consisting of a model and the Kalman filter can be interpreted to be the same as the data flow in a recurrent neural network (RNN). Thus, instead of building more sophisticated models and data assimilation methods, we want to train a RNN to approximate the dynamics of the response of the FMC sensor to a time series of environmental data. Because standard AI approaches did not converge to reasonable solutions, we pre-train the RNN with special initial weights devised to turn it into a numerical solver of the differential equation. We then allow the AI training machinery to optimize the RNN weights to fit the data better. We illustrate the method on an example of a time series of 10h-FMC from RAWS and weather data from the Real-Time Mesoscale Analysis (RTMA)., Comment: 4 pages, 4 figures. Seminar on Numerical Analysis SNA'23, Ostrava, Czech Republic, January 23-27, 2023

Vizier is the de-facto blackbox and hyperparameter optimization service across Google, having optimized some of Google's largest products and research efforts. To operate at the scale of tuning thousands of users' critical systems, Google Vizier solved key design challenges in providing multiple different features, while remaining fully fault-tolerant. In this paper, we introduce Open Source (OSS) Vizier, a standalone Python-based interface for blackbox optimization and research, based on the Google-internal Vizier infrastructure and framework. OSS Vizier provides an API capable of defining and solving a wide variety of optimization problems, including multi-metric, early stopping, transfer learning, and conditional search. Furthermore, it is designed to be a distributed system that assures reliability, and allows multiple parallel evaluations of the user's objective function. The flexible RPC-based infrastructure allows users to access OSS Vizier from binaries written in any language. OSS Vizier also provides a back-end ("Pythia") API that gives algorithm authors a way to interface new algorithms with the core OSS Vizier system. OSS Vizier is available at https://github.com/google/vizier., Comment: Published as a conference paper for the systems track at the 1st International Conference on Automated Machine Learning (AutoML-Conf 2022). Code can be found at https://github.com/google/vizier

Meta-learning hyperparameter optimization (HPO) algorithms from prior experiments is a promising approach to improve optimization efficiency over objective functions from a similar distribution. However, existing methods are restricted to learning from experiments sharing the same set of hyperparameters. In this paper, we introduce the OptFormer, the first text-based Transformer HPO framework that provides a universal end-to-end interface for jointly learning policy and function prediction when trained on vast tuning data from the wild, such as Google's Vizier database, one of the world's largest HPO datasets. Our extensive experiments demonstrate that the OptFormer can simultaneously imitate at least 7 different HPO algorithms, which can be further improved via its function uncertainty estimates. Compared to a Gaussian Process, the OptFormer also learns a robust prior distribution for hyperparameter response functions, and can thereby provide more accurate and better calibrated predictions. This work paves the path to future extensions for training a Transformer-based model as a general HPO optimizer., Comment: Published as a conference paper in Neural Information Processing Systems (NeurIPS) 2022. Code can be found in https://github.com/google-research/optformer and Google AI Blog can be found in https://ai.googleblog.com/2022/08/optformer-towards-universal.html

We present a simple finite element formulation of mass-consistent approximation, and a fast multigrid iterative method with adaptive semicoarsening, which maintains the convergence of the iterations over a range of grids and penalty coefficients. The method is designed to run in each time step of WRF-SFIRE and replace the interpolation from the atmospheric mesh to the fire mesh by a diagnostic model of the air flow over a fine-scale terrain., Comment: 4 pages, Seminar on Numerical Analysis - SNA'21, Ostrava, Czech Republic, January 25--28, 2021

Predicting changes in sea ice cover is critical for shipping, ecosystem monitoring, and climate modeling. Current sea ice models, however, predict more ice than is observed in the Arctic, and less in the Antarctic. Improving the fit of these physics-based models to observations is challenging because the models are expensive to run, and therefore expensive to optimize. Here, we construct a machine learning surrogate that emulates the effect of changing model physics on forecasts of sea ice area from the Los Alamos Sea Ice Model (CICE). We use the surrogate model to investigate the sensitivity of CICE to changes in the parameters governing: ice's ridging and albedo; snow's albedo, aging, and thermal conductivity; the effect of meltwater on albedo; and the effect of ponds on albedo. We find that CICE's sensitivity to these model parameters differs between hemispheres. We propose that future sea ice modelers separate the snow conductivity and snow grain size distributions on a seasonal and inter-hemispheric basis, and we recommend optimal values of these parameters. This will make it possible to make models that fit observations of both Arctic and Antarctic sea ice more closely. These results demonstrate that important aspects of the behavior of a leading sea ice model can be captured by a relatively simple support vector regression surrogate model, and that this surrogate dramatically increases the ease of tuning the full simulation., Comment: 6 pages

A dry decade in the Navajo Nation has killed vegetation, dessicated soils, and released once-stable sand into the wind. This sand now covers one-third of the Nation's land, threatening roads, gardens and hundreds of homes. Many arid regions have similar problems: global warming has increased dune movement across farmland in Namibia and Angola, and the southwestern US. Current dune models, unfortunately, do not scale well enough to provide useful forecasts for the $\sim$5\% of land surfaces covered by mobile sand. We test the ability of two deep learning algorithms, a GAN and a CNN, to model the motion of sand dunes. The models are trained on simulated data from community-standard cellular automaton model of sand dunes. Preliminary results show the GAN producing reasonable forward predictions of dune migration at ten million times the speed of the existing model., Comment: Workshop on Tackling climate change with machine learning at NeurIPS. Vancouver, Canada, December 2019

Zeroth-order optimization is the process of minimizing an objective $f(x)$, given oracle access to evaluations at adaptively chosen inputs $x$. In this paper, we present two simple yet powerful GradientLess Descent (GLD) algorithms that do not rely on an underlying gradient estimate and are numerically stable. We analyze our algorithm from a novel geometric perspective and present a novel analysis that shows convergence within an $\epsilon$-ball of the optimum in $O(kQ\log(n)\log(R/\epsilon))$ evaluations, for any monotone transform of a smooth and strongly convex objective with latent dimension $k < n$, where the input dimension is $n$, $R$ is the diameter of the input space and $Q$ is the condition number. Our rates are the first of its kind to be both 1) poly-logarithmically dependent on dimensionality and 2) invariant under monotone transformations. We further leverage our geometric perspective to show that our analysis is optimal. Both monotone invariance and its ability to utilize a low latent dimensionality are key to the empirical success of our algorithms, as demonstrated on BBOB and MuJoCo benchmarks., Comment: 11 main pages, 26 total pages

Climate change is one of the greatest challenges facing humanity, and we, as machine learning experts, may wonder how we can help. Here we describe how machine learning can be a powerful tool in reducing greenhouse gas emissions and helping society adapt to a changing climate. From smart grids to disaster management, we identify high impact problems where existing gaps can be filled by machine learning, in collaboration with other fields. Our recommendations encompass exciting research questions as well as promising business opportunities. We call on the machine learning community to join the global effort against climate change., Comment: For additional resources, please visit the website that accompanies this paper: https://www.climatechange.ai/

We analyze the accuracy of traffic simulations metamodels based on neural networks and gradient boosting models (LightGBM), applied to traffic optimization as fitness functions of genetic algorithms. Our metamodels approximate outcomes of traffic simulations (the total time of waiting on a red signal) taking as an input different traffic signal settings, in order to efficiently find (sub)optimal settings. Their accuracy was proven to be very good on randomly selected test sets, but it turned out that the accuracy may drop in case of settings expected (according to genetic algorithms) to be close to local optima, which makes the traffic optimization process more difficult. In this work, we investigate 16 different metamodels and 20 settings of genetic algorithms, in order to understand what are the reasons of this phenomenon, what is its scale, how it can be mitigated and what can be potentially done to design better real-time traffic optimization methods., Comment: Presented at NeurIPS 2018 Workshop "Machine Learning for Intelligent Transportation Systems" https://sites.google.com/site/nips2018mlits

Data likelihood of fire detection is the probability of the observed detection outcome given the state of the fire spread model. We derive fire detection likelihood of satellite data as a function of the fire arrival time on the model grid. The data likelihood is constructed by a combination of the burn model, the logistic regression of the active fires detections, and the Gaussian distribution of the geolocation error. The use of the data likelihood is then demonstrated by an estimation of the ignition point of a wildland fire by the maximization of the likelihood of MODIS and VIIRS data over multiple possible ignition points., Comment: 12 pages, 6 figures; VIII International Conference on Forest Fire Research, Coimbra, Portugal, November 2018

Observational data collected during experiments, such as the planned Fire and Smoke Model Evaluation Experiment (FASMEE), are critical for progressing and transitioning coupled fire-atmosphere models like WRF-SFIRE and WRF-SFIRE-CHEM into operational use. Historical meteorological data, representing typical weather conditions for the anticipated burn locations and times, have been processed to initialize and run a set of simulations representing the planned experimental burns. Based on an analysis of these numerical simulations, this paper provides recommendations on the experimental setup that include the ignition procedures, size and duration of the burns, and optimal sensor placement. New techniques are developed to initialize coupled fire-atmosphere simulations with weather conditions typical of the planned burn locations and time of the year. Analysis of variation and sensitivity analysis of simulation design to model parameters by repeated Latin Hypercube Sampling are used to assess the locations of the sensors. The simulations provide the locations of the measurements that maximize the expected variation of the sensor outputs with the model parameters., Comment: 35 pages, 4 tables, 28 figures

Assimilation of data into a fire-spread model is formulated as an optimization problem. The level set equation, which relates the fire arrival time and the rate of spread, is allowed to be satisfied only approximately, and we minimize a norm of the residual. Previous methods based on modification of the fire arrival time either used an additive correction to the fire arrival time, or made a position correction. Unlike additive fire arrival time corrections, the new method respects the dependence of the fire rate of spread on diurnal changes of fuel moisture and on weather changes, and, unlike position corrections, it respects the dependence of the fire spread on fuels and terrain as well. The method is used to interpolate the fire arrival time between two perimeters by imposing the fire arrival time at the perimeters as constraints., Comment: 13 pages, 6 figures, ICCS 2018. Minor update

A new Thurstonian rating scale model uses a variable decision rule (VDR) that incorporates three previously formulated, distinct decision rules. The model includes probabilities for choosing each rule, along with Gaussian representation and criterion densities. Numerical optimisation techniques were validated through demonstrating that the model fits simulated data tightly. For simulations with 400 trials per stimulus (tps), useful information emerged about the generating parameters. However, larger experiments (e.g. 4000 tps) proved desirable for better recovery of generating parameters and to support trustworthy choices between competing models by the Akaike Information Criterion. In reanalyses of experiments by others, the VDR model explained most of the data better than did classical signal detection theory models., Comment: Contains source code as an attachment to this PDF, in files 2015-12-02_speechresearch.tgz and Rosner_Kochanski_2016.tgz

Currently available satellite active fire detection products from the VIIRS and MODIS instruments on polar-orbiting satellites produce detection squares in arbitrary locations. There is no global fire/no fire map, no detection under cloud cover, false negatives are common, and the detection squares are much coarser than the resolution of a fire behavior model. Consequently, current active fire satellite detection products should be used to improve fire modeling in a statistical sense only, rather than as a direct input. We describe a new data assimilation method for active fire detection, based on a modification of the fire arrival time to simultaneously minimize the difference from the forecast fire arrival time and maximize the likelihood of the fire detection data. This method is inspired by contour detection methods used in computer vision, and it can be cast as a Bayesian inverse problem technique, or a generalized Tikhonov regularization. After the new fire arrival time on the whole simulation domain is found, the model can be re-run from a time in the past using the new fire arrival time to generate the heat fluxes and to spin up the atmospheric model until the satellite overpass time, when the coupled simulation continues from the modified state., Comment: 9 pages, VII International Conference on Forest Fire Research, Coimbra, Portugal, November 17-20, 2014

Fuel moisture has a major influence on the behavior of wildland fires and is an important underlying factor in fire risk assessment. We propose a method to assimilate dead fuel moisture content observations from remote automated weather stations (RAWS) into a time-lag fuel moisture model. RAWS are spatially sparse and a mechanism is needed to estimate fuel moisture content at locations potentially distant from observational stations. This is arranged using a trend surface model (TSM), which allows us to account for the effects of topography and atmospheric state on the spatial variability of fuel moisture content. At each location of interest, the TSM provides a pseudo-observation, which is assimilated via Kalman filtering. The method is tested with the time-lag fuel moisture model in the coupled weather-fire code WRF-SFIRE on 10-hr fuel moisture content observations from Colorado RAWS in 2013. We show using leave-one-out testing that the TSM compares favorably with inverse squared distance interpolation as used in the Wildland Fire Assessment System. Finally, we demonstrate that the data assimilation method is able to improve fuel moisture content estimates in unobserved fuel classes., Comment: 28 pages

In this study, we describe how WRF-Sfire is coupled with WRF-Chem to construct WRFSC, an integrated forecast system for wildfire and smoke prediction. The integrated forecast system has the advantage of not requiring a simple plume-rise model and assumptions about the size and heat release from the fire in order to determine fire emissions into the atmosphere. With WRF-Sfire, wildfire spread, plume and plume-top heights are predicted directly, at every WRF timestep, providing comprehensive meteorology and fire emissions to the chemical transport model WRF-Chem. Evaluation of WRFSC was based on comparisons between available observations to the results of two WRFSC simulations. The study found overall good agreement between forecasted and observed fire spread and smoke transport for the Witch-Guejito fire. Also the simulated PM2.5 (fine particulate matter) peak concentrations matched the observations. However, the NO and ozone levels were underestimated in the simulations and the peak concentrations were mistimed. Determining the terminal or plume-top height is one of the most important aspects of simulating wildfire plume transport, and the study found overall good agreement between simulated and observed plume-top heights, with some (10% or less) underestimation by the simulations. One of the most promising results of the study was the agreement between passive-tracer modeled plume-top heights for the Barker Canyon fire simulation and observations. This simulation took only 13h, with the first 24h forecast ready in almost 3h, making it a possible operational tool for providing emission profiles for external chemical transport models., Comment: 28 pages, 12 figures

Fuel moisture is a major influence on the behavior of wildland fires and an important underlying factor in fire risk. We present a method to assimilate spatially sparse fuel moisture observations from remote automatic weather stations (RAWS) into the moisture model in WRF-SFIRE. WRF-SFIRE is a coupled atmospheric and fire behavior model which simulates the evolution of fuel moisture in idealized fuel species based on atmospheric state. The proposed method uses a modified trend surface model to estimate the fuel moisture field and its uncertainty based on currently available observations. At each grid point of WRF-SFIRE, this information is combined with the model forecast using a nonlinear Kalman filter, leading to an updated estimate of fuel moisture. We demonstrate the effectiveness of the method with tests in two real-world situations: a region in Southern California, where two large Santa Ana fires occurred recently, and on a domain enclosing Colorado., Comment: 4th Fire Behavior and Fuels Conference Raleigh, North Carolina, USA, February 18 - 22, 2013

A single-mirror small-size (1M-SST) Davies-Cotton telescope has been proposed for the southern observatory of the Cherenkov Telescope Array (CTA) by a consortium of scientific institutions from Poland, Switzerland, and Germany. The telescope has a 4 m diameter reflector and will be equipped with a fully digital camera based on Geiger avalanche photodiodes (APDs). Such a design is particularly interesting for CTA because it represents a very simple, reliable, and cheap solution for a SST. Here we present the design and the characteristics of the mechanical structure of the 1M-SST telescope and its drive system. We also discuss the results of a finite element method analysis in order to demonstrate the conformance of the design with the CTA specifications and scientific objectives. In addition, we report on the current status of the construction of a prototype telescope structure at the Institute of Nuclear Physics PAS in Krakow., Comment: In Proceedings of the 33rd International Cosmic Ray Conference (ICRC2013), Rio de Janeiro (Brazil). All CTA contributions at arXiv:1307.2232

The Cherenkov Telescope Array (CTA) is the next generation very high energy gamma-ray observatory. It will consist of three classes of telescopes, of large, medium and small sizes. The small telescopes, of 4 m diameter, will be dedicated to the observations of the highest energy gamma-rays, above several TeV. We present the technical characteristics of a single mirror, 4 m diameter, Davies-Cotton telescope for the CTA and the performance of the sub-array consisting of the telescopes of this type. The telescope will be equipped with a fully digital camera based on custom made, hexagonal Geiger-mode avalanche photodiodes. The development of cameras based on such devices is an RnD since traditionally photomultipliers are used. The photodiodes are now being characterized at various institutions of the CTA Consortium. Glass mirrors will be used, although an alternative is being considered: composite mirrors that could be adopted if they meet the project requirements. We present a design of the telescope structure, its components and results of the numerical simulations of the telescope performance., Comment: In Proceedings of the 33rd International Cosmic Ray Conference (ICRC2013), Rio de Janeiro (Brazil). All CTA contributions at arXiv:1307.2232

Atmospheric pollution regulations have emerged as a dominant obstacle to prescribed burns. Thus, forecasting the pollution caused by wildland fires has acquired high importance. WRF and SFIRE model wildland fire spread in a two-way interaction with the atmosphere. The surface heat flux from the fire causes strong updrafts, which in turn change the winds and affect the fire spread. Fire emissions, estimated from the burning organic matter, are inserted in every time step into WRF-Chem tracers at the lowest atmospheric layer. The buoyancy caused by the fire then naturally simulates plume dynamics, and the chemical transport in WRF-Chem provides a forecast of the pollution spread. We discuss the choice of wood burning models and compatible chemical transport models in WRF-Chem, and demonstrate the results on case studies., Comment: Proceedings of 4th Fire Behavior and Fuels Conference, February 18-22, 2013, Raleigh, North Carolina, USA

We describe two recent additions to WRF coupled with a fire spread model. Fire propagation is strongly dependent on fuel moisture, which in turn depends on the history of the atmosphere. We have implemented a equilibrium time-lag model of fuel moisture driven by WRF variables. The code allows the user to specify fuel parameters, with the defaults calibrated to the Canadian fire danger rating system for 10-hour fuel. The moisture model can run coupled with the atmosphere-fire model, or offline from WRF output to equilibrate the moisture over a period of time and to provide initial moisture conditions for a coupled atmosphere-fire-moisture simulation. The fire model also inserts smoke tracers into WRF-Chem to model the transport of fire emissions. The coupled model is available from OpenWFM.org. An earlier version of the fire model coupled with atmosphere is a part of WRF release., Comment: 2012 WRF Users Workshop, Boulder, CO, 25-29 June 2012, paper P51, 5 pages. Corrected a typo

We introduce several useful utilities in development for the creation and analysis of real wildland fire simulations using WRF and SFIRE. These utilities exist as standalone programs and scripts as well as extensions to other well known software. Python web scrapers automate the process of downloading and preprocessing atmospheric and surface data from common sources. Other scripts simplify the domain setup by creating parameter files automatically. Integration with Google Earth allows users to explore the simulation in a 3D environment along with real surface imagery. Postprocessing scripts provide the user with a number of output data formats compatible with many commonly used visualization suites allowing for the creation of high quality 3D renderings. As a whole, these improvements build toward a unified web application that brings a sophisticated wildland fire modeling environment to scientists and users alike., Comment: Submitted to proceedings of IGARSS 2012, 4 papers, 1 figure

This study uses in-situ measurements collected during the FireFlux field experiment to evaluate and improve the performance of coupled atmosphere-fire model WRF-Sfire. The simulation by WRF-Sfire of the experimental burn shows that WRF-Sfire is capable of providing realistic head fire rate-of-spread and the vertical temperature structure of the fire plume, and, up to 10 m above ground level, fire-induced surface flow and vertical velocities within the plume. The model captured the changes in wind speed and direction before, during, and after fire front passage, along with arrival times of wind speed, temperature, and updraft maximae, at the two instrumented flux towers used in FireFlux. The model overestimated vertical velocities and underestimated horizontal wind speeds measured at tower heights above the 10 m, and it is hypothesized that the limited model resolution over estimated the fire front depth, leading to too high a heat release and, subsequently, too strong an updraft. However, on the whole, WRF-Sfire fire plume behavior is consistent with the interpretation of FireFlux observations. The study suggests optimal experimental pre-planning, design, and execution of future field campaigns that are needed for further coupled atmosphere-fire model development and evaluation.

Phonology typically describes speech in terms of discrete signs like features. The field of intonational phonology uses discrete accents to describe intonation and prosody. But, are such representations useful? The results of mimicry experiments indicate that discrete signs are not a useful representation of the shape of intonation contours. Human behaviour seems to be better represented by a attractors where memory retains substantial fine detail about an utterance. There is no evidence that discrete abstract representations that might be formed that have an effect on the speech that is subsequently produced. This paper also discusses conditions under which a discrete phonology can arise from an attractor model and why - for intonation - attractors can be inferred without the implying a discrete phonology., Comment: 36 pages, plus extra figures

We present a methodology to change the state of the Weather Research Forecasting (WRF) model coupled with the fire spread code SFIRE, based on Rothermel's formula and the level set method, and with a fuel moisture model. The fire perimeter in the model changes in response to data while the model is running. However, the atmosphere state takes time to develop in response to the forcing by the heat flux from the fire. Therefore, an artificial fire history is created from an earlier fire perimeter to the new perimeter, and replayed with the proper heat fluxes to allow the atmosphere state to adjust. The method is an extension of an earlier method to start the coupled fire model from a developed fire perimeter rather than an ignition point. The level set method is also used to identify parameters of the simulation, such as the spread rate and the fuel moisture. The coupled model is available from openwfm.org, and it extends the WRF-Fire code in WRF release., Comment: ICCS 2012, 10 pages; corrected some DOI typesetting in the references

There are many wildfire behaviors of increasing relevance that are outside the forecast capabilities of even the most sophisticated operational fire spread and fire behavior model. The limitations of the operational models are due primarily to their inability to represent coupled fire-atmosphere interactions. Coupled wildfire-atmosphere models are physics-based fluid-dynamical prognostic models of wildfire spread and behavior that attempt an almost complete representation of fire-atmosphere interactions. This level of fidelity however means that these models cannot be used operationally. The reason is that, despite ever increasing computational resources, the complexity and range of processes and scales (1 mm to 100 km) involved in this modeling approach make computational costs prohibitively expensive. In this study we propose an intermediate approach. A physics-based coupled atmosphere-fire model is used to resolve the large-scale and local weather as well as the atmosphere-fire interactions, while combustion is represented simply using an existing operational surface fire behavior model. This model combination strikes a balance between fidelity and speed of execution. The feasibility of this approach is examined based on an analysis of a numerical simulation of two very large Santa Ana fires using WRF-Sfire, a coupled atmosphere-fire model available at the Open Wildland Fire Modeling Community (OpenWFM.org); an earlier version is available as WRF-Fire in WRF release. The study demonstrates that a wind and fire spread forecast of reasonable accuracy was obtained at an execution speed that would have made real-time wildfire forecasting of this event possible., Comment: Exploring the Mega-fire Reality 2011, November 14-17, 2011, Florida State University Conference Center, Tallahassee, Florida. Added full author names

The ability to forecast grass fire spread could be of a great importance for agencies making decisions about prescribed burns. However, the usefulness of the models used for fire-spread predictions is limited by the time required for completing the coupled atmosphere-fire simulations. In this study we analyze the sensitivity of a coupled model with respect to the vertical resolution of the atmospheric grid and the resolution of fire mesh that both affect computational performance of the model. Based on the observations of the plume properties recorded during the FireFlux experiment (Clements et al., 2007), we try to establish the optimal model configuration that provides realistic results for the least computational expense., Comment: AMS, Ninth Symposium on Fire and Forest Meteorology, Palm Springs 2011

We present an overview of a modeling environment, consisting of a coupled atmosphere-wildfire model, utilities for visualization, data processing, and diagnostics, open source software repositories, and a community wiki. The fire model, called SFIRE, is based on a fire-spread model, implemented by the level-set method, and it is coupled with the Weather Research Forecasting (WRF) model. A version with a subset of the features is distributed with WRF 3.3 as WRF-Fire. In each time step, the fire module takes the wind as input and returns the latent and sensible heat fluxes. The software architecture uses WRF parallel infrastructure for massively parallel computing. Recent features of the code include interpolation from an ideal logarithmic wind profile for nonhomogeneous fuels and ignition from a fire perimeter with an atmosphere and fire spin-up. Real runs use online sources for fuel maps, fine-scale topography, and meteorological data, and can run faster than real time. Visualization pathways allow generating images and animations in many packages, including VisTrails, VAPOR, MayaVi, and Paraview, as well as output to Google Earth. The environment is available from openwfm.org. New diagnostic variables were added to the code recently, including a new kind of fireline intensity, which takes into account also the speed of burning, unlike Byram's fireline intensity., Comment: 12 pages; Ninth Symposium on Fire and Forest Meteorology, Palm Springs, CA, October 2011

The current WRF-Fire model starts the fire from a given ignition point at a given time. We want to start the model from a given fire perimeter at a given time instead. However, the fuel balance and the state of the atmosphere depend on the history of the fire. The purpose of this work is to create an approximate artificial history of the fire based on the given fire perimeter and time and an approximate ignition point and time. Replaying the fire history then establishes a reasonable fuel balance and outputs heat fluxes into the atmospheric model, which allow the atmospheric circulation to develop. Then the coupled atmosphere-fire model takes over. In this preliminary investigation, the ignition times in the fire area are calculated based on the distance from the ignition point to the perimeter, assuming that the perimeter is convex or star-shaped. Simulation results for an ideal example show that the fire can continue in a natural way from the perimeter. Possible extensions include algorithms for more general perimeters and running the fire model backwards in time from the perimeter to create a more realistic history. The model used extends WRF-Fire and it is available from openwfm.org., Comment: 12th Annual WRF Users' Workshop, Boulder, CO, June 2011

We use a coupled atmosphere-fire model to simulate a fire that occurred on August 14--17, 2009, in the Harmanli region, Bulgaria. Data was obtained from GIS and satellites imagery, and from standard atmospheric data sources. Fuel data was classified in the 13 Anderson categories. For correct fire behavior, the spatial resolution of the models needed to be fine enough to resolve the essential micrometeorological effects. The simulation results are compared to available incident data. The code runs faster than real time on a cluster. The model is available from openwfm.org and it extends WRF-Fire from WRF 3.3 release., Comment: 8 pages, 2 tables, 5 figures. 8th International Conference on Large-Scale Scientific Computations, June 6-10, 2011, Sozopol, Bulgaria

We describe the physical model, numerical algorithms, and software structure of WRF-Fire. WRF-Fire consists of a fire-spread model, implemented by the level-set method, coupled with the Weather Research and Forecasting model. In every time step, the fire model inputs the surface wind, which drives the fire, and outputs the heat flux from the fire into the atmosphere, which in turn influences the atmosphere. The level-set method allows submesh representation of the burning region and flexible implementation of various ignition modes. WRF-Fire is distributed as a part of WRF and it uses the WRF parallel infrastructure for parallel computing., Comment: Version 3.3, 41 pages, 2 tables, 12 figures. As published in Discussions, under review for Geoscientific Model Development

We describe the coupled atmosphere-wildfire model WRF-Fire, which is distributed as a part of WRF. The fire module is based on a fire-spread model, implemented by the level-set method. In each time step, the fire module takes the wind as input and returns the latent and sensible heat fluxes. We report on the software architecture and features of the software., Comment: 91st American Meteorological Society Annual Meeting, Seattle, WA, January 2011

The paper presents methods for evaluating the accuracy of alignments between transcriptions and audio recordings. The methods have been applied to the Spoken British National Corpus, which is an extensive and varied corpus of natural unscripted speech. Early results show good agreement with human ratings of alignment accuracy. The methods also provide an indication of the location of likely alignment problems; this should allow efficient manual examination of large corpora. Automatic checking of such alignments is crucial when analysing any very large corpus, since even the best current speech alignment systems will occasionally make serious errors. The methods described here use a hybrid approach based on statistics of the speech signal itself, statistics of the labels being evaluated, and statistics linking the two., Comment: Four pages, 3 figures. Presented at "New Tools and Methods for Very-Large-Scale Phonetics Research", University of Pennsylvania, January 28-31, 2011

In this paper, we ask what properties makes a large corpus more or less useful. We suggest that size, by itself, should not be the ultimate goal of building a corpus. Large-scale corpora are considered desirable because they offer statistical stability and rich variation. But this rich variation means more factors to control and evaluate, which can limit the advantages of size. We discuss the use of multi-channel data to complement large-scale speech corpora. Even though multi-channel data may limit the scale of a corpus (due to the complex and labor-intensive nature of data collection) they can offer information that allows us to tease apart various factors related to speech production.

A novel procedure is described for accelerating the convergence of Markov chain Monte Carlo computations. The algorithm uses an adaptive bootstrap technique to generate candidate steps in the Markov Chain. It is efficient for symmetric, convex probability distributions, similar to multivariate Gaussians, and it can be used for Bayesian estimation or for obtaining maximum likelihood solutions with confidence limits. As a test case, the Law of Categorical Judgment (Corrected) was fitted with the algorithm to data sets from simulated rating scale experiments. The correct parameters were recovered from practical-sized data sets simulated for Full Signal Detection Theory and its special cases of standard Signal Detection Theory and Complementary Signal Detection Theory.

The notion of the Trojan state of a Rydberg electron, introduced by I.Bialynicki-Birula, M.Kali\'nski, and J.H.Eberly (Phys. Rev. Lett. 73, 1777 (1994)) is extended to the case of the electromagnetic field quantized in acavity. The shape of the electronic wave packet describing the Trojan state is practically the same as in the previously studied externally driven system. The fluctuations of the quantized electromagnetic field around its classical value exhibit strong squeezing. The emergence of Trojan states in the cylindrically symmetrical system is attributed to spontaneous symmetry braking., Comment: 9 pages, 8 figures

An operational time of arrival is introduced using a realistic position and momentum measurement scheme. The phase space measurement involves the dynamics of a quantum particle probed by a measuring device. For such a measurement an operational positive operator valued measure in phase space is introduced and investigated. In such an operational formalism a quantum mechanical time operator is constructed and analyzed. A phase space time and energy uncertainty relation is derived., Comment: 23 pages, 5 figures, to appear in Phys. Rev. A

We construct a high resolution mass map of the z=0.39 cluster 0024+1654, based on parametric inversion of the associated gravitational lens. The lens creates eight well-resolved sub-images of a background galaxy, seen in deep imaging with HST. Excluding mass concentrations centered on visible galaxies, more than 98% of the remaining mass is represented by a smooth concentration of dark matter centered near the brightest cluster galaxies, with a 35 h^{-1} kpc soft core. The asymmetry in the mass distribution is <3% inside 107 ~h^{-1} kpc radius. The dark matter distribution we observe in CL0024 is far more smooth, symmetric, and nonsingular than in typical simulated clusters in either Omega=1 or Omega=0.3 CDM cosmologies. Integrated to 107 h^{-1} kpc radius, the rest-frame mass to light ratio is M/L_V = 276\pm 40 h (M/L_V)_solar., Comment: 16 pages, 4 figures (3 .jpg, 1 .ps), minor changes to make consistent with the final ApJL article. To appear in ApJL, May 8 1998

The eight-port homodyne detection apparatus is analyzed in the framework of the operational theory of quantum measurement. For an arbitrary quantum noise leaking through the unused port of the beam splitter, the positive operator valued measure and the corresponding operational homodyne observables are derived. It is shown that such an eight-port homodyne device can be used to construct the operational quantum trigonometry of an optical field. The quantum trigonometry and the corresponding phase space Wigner functions are derived for a signal field probed by a classical local oscillator and a squeezed vacuum in the unused port., Comment: 18 pages, uses Revtex and epsfig.sty, 4 Encapsulated Postscript figures, submitted to J. Mod. Opt

The concept of intrinsic and operational observables in quantum mechanics is introduced. It is argued that, in any realistic description of a quantum measurement that includes a detecting device, it is possible to construct from the statistics of the recorded raw data a set of operational quantities that correspond to the intrinsic quantum mechanical observable. Using the concept of the propensity and the associated operational positive operator valued measure (POVM) a general description of the operational algebra of quantum mechanical observables is derived for a wide class of realistic detection schemes. This general approach is illustrated by the example of an operational Malus measurement of the spin phases and by an analysis of the operational homodyne detection of the phase of an optical field with a squeezed vacuum in the unused ports., Comment: 8 pages, LaTex type, 2 Encapsulated Postscript figures, uses a4wide.sty, to appear in acta physica slovaca vol. 46 No. 3 (1996)

Optical variability in galaxies at high redshift is a tracer of evolution in AGN activity, and should provide a useful constraint on models of galaxy evolution, AGN structure, and cosmology. We studied optical variability in multiple deep CCD and photographic surveys of blank fields for galaxies with $B_j = 20 - 25$ mag. Weakly variable objects are far more common than strongly variable ones. For objects near $B_j = 22$, $0.74\% \pm 0.2 \%$ vary by 0.026~mag RMS or more, over a decade. This is small compared with previous claims based on photographic surveys, and also small compared with the fraction of bright quasars ($\approx 5\%$ at $B_j = 20$~mag) or Seyferts ($\approx 1-2\%$ for $B_j < 18$). The fraction of objects that vary increases slowly with magnitude. Detection probabilities and error rates were checked by simulations and statistical analysis of fluctuations of sample sky spots., Comment: AAS LaTeX file, full paper with figures available at http://www.astro.lsa.umich.edu:80/users/philf/ Accepted for publication in the AJ

In this paper we describe weighting techniques used for the optimal coaddition of CCD frames with differing characteristics. Optimal means maximum signal-to-noise (s/n) for stellar objects. We derive formulae for four applications: 1) object detection via matched filter, 2) object detection identical to DAOFIND, 3) aperture photometry, and 4) ALLSTAR profile-fitting photometry. We have included examples involving 21 frames for which either the sky brightness or image resolution varied by a factor of three. The gains in s/n were modest for most of the examples, except for DAOFIND detection with varying image resolution which exhibited a substantial s/n increase. Even though the only consideration was maximizing s/n, the image resolution was seen to improve for most of the variable resolution examples. Also discussed are empirical fits for the weighting and the availability of the program, WEIGHT, used to generate the weighting for the individual frames. Finally, we include appendices describing the effects of clipping algorithms and a scheme for star/galaxy and cosmic ray/star discrimination., Comment: 27 pages (uuencoded compressed postscript), 1993