Application of an evolutionary algorithm to the inverse parameter estimation of an individual-based model

Inverse parameter estimation of individual-based models (IBMs) is a research area which is still in its infancy, in a context where conventional statistical methods are not well suited to confront this type of models with data. In this paper, we propose an original evolutionary algorithm which is designed for the calibration of complex IBMs, i.e. characterized by high stochasticity, parameter uncertainty and numerous non-linear interactions between parameters and model output. Our algorithm corresponds to a variant of the population-based incremental learning (PBIL) genetic algorithm, with a specific “optimal individual” operator. The method is presented in detail and applied to the individual-based model OSMOSE. The performance of the algorithm is evaluated and estimated parameters are compared with an independent manual calibration. The results show that automated and convergent methods for inverse parameter estimation are a significant improvement to existing ad hoc methods for the calibration of IBMs.     

Comparing model averaging with other model selection strategies for benchmark dose estimation

Model averaging (MA) has been proposed as a method of accommodating model uncertainty when estimating risk. Although the use of MA is inherently appealing, little is known about its performance using general modeling conditions. We investigate the use of MA for estimating excess risk using a Monte Carlo simulation. Dichotomous response data are simulated under various assumed underlying dose–response curves, and nine dose–response models (from the USEPA Benchmark dose model suite) are fit to obtain both model specific and MA risk estimates. The benchmark dose estimates (BMDs) from the MA method, as well as estimates from other commonly selected models, e.g., best fitting model or the model resulting in the smallest BMD, are compared to the true benchmark dose value to better understand both bias and coverage behavior in the estimation procedure. The MA method has a small bias when estimating the BMD that is similar to the bias of BMD estimates derived from the assumed model. Further, when a broader range of models are included in the family of models considered in the MA process, the lower bound estimate provided coverage close to the nominal level, which is superior to the other strategies considered. This approach provides an alternative method for risk managers to estimate risk while incorporating model uncertainty.

A simple model for simulation of insect pheromone dispersion within forest canopies

Synthetic pheromones and other behavioral chemicals are used by land managers to prevent insect-caused tree mortality or crop failure in forest and agricultural systems. Currently, no method exists to continuously measure pheromone concentration or movement in real-time. To improve our understanding of pheromone fate and transport under different forest canopies, results from a set of surrogate pheromone (sulfur hexafluoride tracer) experimental trials were used to evaluate a simple, instantaneous, three-dimensional Lagrangian dispersion model. The model was designed to predict both instantaneous and time-averaged pheromone concentrations. Overall, the results from the model show simulated time-averaged arc maximum concentrations within a factor of two of the observed data. The model correctly matched the sharp peaks and narrow widths of the meandering plumes observed in the instantaneous data, however the magnitude of the instantaneous peaks was often under-estimated. This model and evaluation provide the basis for a tool that can be used to guide deployment of synthetic pheromones or other semiochemicals for monitoring, mass trapping, or disruption of mating or aggregation.     

Development of an empirical equation for the transverse dispersion coefficient in natural streams

In this study, a new empirical equation for the transverse dispersion coefficient has been developed based on the hydraulic and geometric parameters in natural streams using a regression technique. First, a total of 32 data sets in 16 streams were collected. Among those sets, 16 sets were used for deriving the new equation, and the other 16 sets were used for verifying the equation. Then, through dimensional analysis, it was found that the normalized transverse dispersion coefficient is associated with several parameters such as sinuosity, aspect ratio, and a friction term. The robust least square method was applied to estimate regression coefficients. The newly proposed equation was proven to be superior in explaining the dispersion characteristics of natural streams more precisely compared to the existing equations.     

珠江三角洲土地利用变化对特征大气污染物扩散的影响

在珠江三角洲两种下垫面条件下,应用CALPUFF大气污染扩散模式,对特征污染物SO2、SO42-的扩散进行数值模拟,探讨大规模土地利用变化,尤其是城镇建设用地增加,对珠江三角洲地区大气污染物扩散的影响,并通过对4个典型区污染物月均质量浓度变化特征分析,揭示土地利用变化对不同地区的污染物分布的影响机制。模拟结果表明：土地利用变化,尤其是城镇建设用地增加,不利于污染物扩散,污染源下风方向地区受影响较大,污染物质量浓度明显升高,SO2和 SO42-年均质量浓度分别增加14.07%和3.31%;受影响范围、变化幅度与污染源排污强度呈正相关,变化幅度亦与污染源距离远近呈负相关。土地利用变化后,尤其是城镇建设用地增加,四个典型区 SO2月均质量浓度都表现为升高趋势,且冬季 SO2质量浓度升高幅度最大,夏季升高幅度最小,临近污染源密集区的两个典型区SO2月均质量浓度分别增加33.6%和26.3%。土地利用变化不仅改变局地的污染扩散,也会对区域的污染扩散有一定影响,尤其对污染源分布密集区的大气污染物扩散影响强度最大。因此,建议人类在城市化建设过程中尽可能保留自然斑块,消除人工下垫面对污染物扩散的负面影响。     

基于MODIS数据的河南省冬小麦产量遥感估算模型

小麦是世界上最重要的粮食作物,小麦生产对中国的粮食保障起着十分重要的作用,及时、准确、大范围对小麦产量进行监测预报,对于农学经济发展和粮食政策制定具有极为重要的现实意义。对作物产量进行遥感监测的原理是建立在其遥感特征基础之上的,通过建立作物长势指标与遥感信息的定量关系,可实现对作物产量的监测预报。文章基于2009年MODIS遥感数据和气象数据,利用Arcgis和ENVI提取纯小麦像元,并提取纯小麦像元对应的NDVI、NPP和LAI,获取分县NDVI、NPP和LAI均值,利用统计软件对产量数据和分县遥感参数均值进行数据整理和分析,建立了河南省冬小麦产量估算模型。以往研究多采用遥感图像上某像元和地面调查点进行研究,具有很大的不确定性,文章以县为单位,对冬小麦平均单产和县域内冬小麦种植像元遥感参数的均值进行相关研究,提高了模型模拟精度。同时文章选用多种遥感参数和多项气象因子建立估产模型,避免了针对一个参数进行估产的局限性。在最佳时相的选择上,根据冯美辰（2010）以往的研究结果,从4月以后,5月8日和4月20Et植被指数和产量相关性最大,4月份之前冬小麦处于返青到拔节期,对产量来说还有很多不确定闪素,因此文章选用5月8El和4月20日进行冬小麦估产研究。结果表明,5月8日的估产模型优于4月20日,加入气象冈子的遥感气象估产模型优于只采用遥感参数进行估产的遥感模型。利用2010年产量数据对模型精度进行检验,遥感气象模型预测精度在70．2％N99．7％之间,平均精度为90．7％;遥感模型预测精度在68．1％到95．5％之间,平均精度为83．9％。表明遥感气象模型模拟精度更高,其精度可以满足大面积估产要求,可以对产量预报提供科学参考。     

环境归趋模型参数估计方法的比较

大量的研究表明,参数估计方法的可行性和有效性,决定着环境归趋模型的成功与否。在文献检索的基础上,简单介绍了环境归趋模型参数估计方法的类型和最优化原理,分析了各种方法的优点与不足,并初步探讨了环境归趋模型参数估计方法的研究趋势。     

Sensitivity of atmospheric dispersion simulations by HYSPLIT to the meteorological predictions from a meso-scale model

Mesoscale transport and dispersion of air pollutants from a few major point sources in the Mississippi Gulf coastal region is calculated using a coupled modeling system consisting of the atmospheric dynamical model WRF and the lagrangian particle model HYSPLIT. The sensitivity of the dispersion model results to the meteorological fields is studied by conducting an ensemble of simulations using the WRF model for the same dispersion case. Several parameterization schemes for the physical processes of boundary layer turbulence and land surface temperature/moisture prediction in WRF are used in various combinations to produce different meteorological members which are then used for dispersion simulation. The uncertainty in the simulated concentration probabilities to the meteorological model configurations and the ensemble mean are presented. The parameters used for determining the uncertainties include the wind fields, temperature, area of concentration and the levels of concentration. The results indicate that dispersion model results are influenced by the choices made in respect of the planetary boundary layer and land surface schemes in the mesoscale model to produce the meteorological forecast thereby leading to certain amount of uncertainty in the resultant concentrations. Results show that the specific choices made about the atmospheric model configuration can significantly after the simulated concentrations.     

An EM algorithm for capture-recapture estimation

Analysis of capture—recapture data often involves maximizing a complex likelihood function with many unknown parameters. Statistical inference based on selection of a proper model depends on successful attainment of this maximum. An EM algorithm is developed for obtaining maximum likelihood estimates of capture and survival probabilities conditional on first capture from standard capture—recapture data. The algorithm does not require the use of numerical derivatives which may improve precision and stability relative to other estimation schemes. The asymptotic covariance matrix of the estimated parameters can be obtained using the supplemented EM algorithm. The EM algorithm is compared to a more traditional Newton-Raphson algorithm with both a simulated and a real dataset. The two algorithms result in the same parameter estimates, but Newton-Raphson variance estimates depend on a numerically estimated Hessian matrix that is sensitive to step size choice.     

Static modeling of dynamic recreation behavior: Implications for prediction and welfare estimation

This paper examines the consequences of using a static model of recreation trip-taking behavior when the underlying decision problem is dynamic. Specifically we examine the implications for trip forecasting and welfare estimation using a panel dataset of Lake Michigan salmon anglers for the 1996 and 1997 fishing seasons. We derive and estimate both a structural dynamic model using Bellman's equation, and a reduced-form static model with trip probability expressions mimicking those of the dynamic model. We illustrate an inherent identification problem in the reduced-form model that creates biased welfare estimates, and we discuss the general implications of this for the interpretation of preference parameters in static models. We then use both models to simulate trip taking behavior and show that although their in-sample trip forecasts are similar, their welfare estimates and out-of-sample forecasts are quite different.     

Evaluation of aquatic rehabilitation technologies for polluted urban rivers and the case study of the Foshan Channel

An index system for evaluation of technologies for urban river rehabilitation was proposed and discussed. The index system includes indicators of cost, resources, environmental improvement, and social effects. The calculation method for an objective value of each index based on its attributes and weights was presented. The Foshan Channel, which is a seriously polluted, black and malodorous urban river in Foshan City, China, was selected as a case study. The values of the attributes and the weights of the indices for the Foshan Channel were determined. The technologies for the rehabilitation of the Foshan Channel were evaluated based on this index system. Finally, a rehabilitation scheme for the Foshan Channel was proposed.     

Improving inverse model fitting in trees—Anisotropy,multiplicative effects,and Bayes estimation

Model fitting for individual-based effects in forests has some problems. Because samples measuring the separate influence of each individual are rarely available, the measured value in the sample represents the influence of all surrounding individual trees. Therefore, it is helpful to build inverse models that use the spatial pattern of the variable as well as that of the source trees. For example, since seed dispersal is influenced by wind effects, a model is discussed describing anisotropic effects to ensure an unbiased estimate of the total fruit number. Further, we present a model describing the absorption of radiation by trees. In this case a multiplicative combination of individual effects yields the total effect. Our approach uses logarithmic transformations of the original data to model multiplicative combinations as sum of transformed single effects. For fitting model parameters we propose an approach based on Bayesian statistics, to ensure ecologically interpretable parameters.     

Evaluation of the QUIC-URB fast response urban wind model for a cubical building array and wide building street canyon

This paper describes the QUIC-URB fast response urban wind modeling tool and evaluates it against wind tunnel data for a 7 × 11 cubical building array and wide building street canyon. QUIC-URB is based on the Röckle diagnostic wind modeling strategy that rapidly produces spatially resolved wind fields in urban areas and can be used to drive urban dispersion models. Röckle-type models do not solve transport equations for momentum or energy; rather, they rely heavily on empirical parameterizations and mass conservation. In the model-experiment comparisons, we test two empirical building flow parameterizations within the QUIC-URB model: our implementation of the standard Röckle (SR) algorithms and a set of modified Röckle (MR) algorithms. The MR model attempts to build on the strengths of the SR model and introduces additional physically based, but simple parameterizations that significantly improve the results in most regions of the flow for both test cases. The MR model produces vortices in front of buildings, on rooftops and within street canyons that have velocities that compare much more favorably to the experimental results. We expect that these improvements in the wind field will result in improved dispersion calculations in built environments.     

Hierarchical Bayes estimation of hunting success rates

Although most post-season harvest surveys are conducted at the state level, the effective management of wildlife populations often requires estimates of hunting success rate, hunting pressure and harvest at the sub-area (such as management unit, regional, or county) level.Sample sizes for some sub-areas are often very small or even zero. Because of small sample sizes, estimates for small sub-areas often yield unacceptably large standard errors. In this article, a hierarchical Bayes model is used to estimate hunting success rates at the sub-area level from post-season harvest surveys. The computation is done by Gibbs sampling and adaptive rejection sampling techniques. The method is illustrated using data from the Missouri Turkey Hunting Survey 1994 Spring Season. The Bayesian estimates are close to the frequency estimates for the sub-areas with large sample sizes and more stable than the frequency estimates for those with small sample sizes. The Bayesian estimates will be more useful to wildlife biologists in estab-lishing hunting regulation on small sub-areas at no additional survey cost.     

A simple plankton model for the Oregon upwelling ecosystem: Sensitivity and validation against time-series ocean data

Simple plankton models serve as useful platforms for testing our understanding of the mechanisms underlying ecosystem dynamics. A simple, one-dimensional plankton model was developed to describe the dynamics of nitrate, ammonium, two phytoplankton size-classes, meso-zooplankton, and detritus in the Oregon upwelling ecosystem. Computational simplicity was maintained by linking the biological model to a one-dimensional, cross-shelf physical model driven by the daily coastal upwelling index. The model sacrificed resolution of regional-scale and along-shore (north to south) processes and assumed that seasonal productivity is primarily driven by local cross-shelf Ekman transport of surface waters and upwelling of nutrient-rich water from depth.Our goals were to see how well a simple plankton model could capture the general temporal and spatial dynamics of the system, test system sensitivity to alternate parameter set values, and observe system response to the effective scale of potential retention mechanisms. Model performance across the central Oregon shelf was evaluated against two years (2000-2001) of chlorophyll and copepod time-series observations. While the modeled meso-zooplankton biomass was close in scale to the observed copepod biomass, phytoplankton was overestimated relative to that inferred from the observed surface chlorophyll concentration. Inshore, the system was most sensitive to the nutrient uptake kinetics of diatom-size phytoplankton and to the functional grazing response of meso-zooplankton. Meso-zooplankton was more sensitive to alternate parameter values than was phytoplankton. Reduction of meso-zooplankton cross-shelf advection rates (crudely representing behavioral retention mechanisms) reduced the scale of model error relative to the observed seasonal mean inshore copepod biomass but had little effect of the modeled meso-zooplankton biomass offshore nor upon phytoplankton biomass across the entire shelf.     

Bayesian estimation of age-specific bird nest survival rates with categorical covariates

The populations of many North American landbirds are showing signs of declining. Gathering information on breeding productivity allows critical assessment of population performance and helps identify good habitat-management practices. He (Biometrics (2003) 59 962–973) proposed a Bayesian model to estimate the age-specific nest survival rates. The model allows irregular visiting schedule under the assumption that the observed nests have homogeneous nest survival. Because nest survival studies are often conducted in different sites and time periods, it is not realistic to assume homogeneous nest survival. In this paper, we extend He’s model by incorporating these factors as categorical covariates. The simulation results show that the Bayesian hierarchical model can produce satisfactory estimates on nest survival and capture different factor effects. Finally the model is applied to a Missouri red-winged blackbird data set.     

Impact of bias in predicted height on tree volume estimation: A case-study of intrinsic nonlinearity

Bias originating from intrinsic nonlinearity in nonlinear models is caused by excess curvature in the solution locus of parameter estimates derived from least squares procedures. Bias due to intrinsic nonlinearity varies according to sample size as well as model specification. This paper analyses consequences of fractionising data into smaller sub-samples. Based on measurements of stem diameter and total tree height from the first Danish national forest inventory, it is demonstrated how data splitting at random may cause the intrinsic nonlinear curvature to exceed the critical F-value. Application of a Taylor-series expansion shows that, for all practical purposes, the bias in predictions of individual tree volume (based on stem diameter and tree height) is negligible. To minimize residual variance, intrinsic curvature and, in turn, prediction bias, it is recommended that data be stratified according to site conditions, stand characteristics or other relevant criteria. Finally, the preferred model should exhibit close-to-linear behaviour.     

Hierarchical spatial modeling for estimation of population size

Estimation of population size has traditionally been viewed from a finite population sampling perspective. Typically, the objective is to obtain an estimate of the total population count of individuals within some region. Often, some stratification scheme is used to estimate counts on subregions, whereby the total count is obtained by aggregation with weights, say, proportional to the areas of the subregions. We offer an alternative to the finite population sampling approach for estimating population size. The method does not require that the subregions on which counts are available form a complete partition of the region of interest. In fact, we envision counts coming from areal units that are small relative to the entire study region and that the total area sampled is a very small proportion of the total study area. In extrapolating to the entire region, we might benefit from assuming that there is spatial structure to the counts. We implement this by modeling the intensity surface as a realization from a spatially correlated random process. In the case of multiple population or species counts, we use the linear model of coregionalization to specify a multivariate process which provides associated intensity surfaces hence association between counts within and across areal units. We illustrate the method of population size estimation with simulated data and with tree counts from a Southwestern pinyon-juniper woodland data set.     

Improved strategies for coverage estimation by using replicated line-intercept sampling

Coverage, i.e., the area covered by the target attribute in the study region, is a key parameter in many surveys. Coverage estimation is usually performed by adopting a replicated protocol based on line-intercept sampling coupled with a suitable linear homogeneous estimator. Since coverage is a parameter which may be interestingly represented as the integral of a suitable function, improved Monte Carlo strategies for implementing the replicated protocol are introduced in order to achieve estimators with small variance rates. In addition, new specific theoretical results on Monte Carlo integration methods are given to deal with the integrand functions arising in the special coverage estimation setting.