Hierarchical Bayesian space-time models

Space-time data are ubiquitous in the environmental sciences. Often, as is the case with atmo- spheric and oceanographic processes, these data contain many different scales of spatial and temporal variability. Such data are often non-stationary in space and time and may involve many observation/prediction locations. These factors can limit the effectiveness of traditional space- time statistical models and methods. In this article, we propose the use of hierarchical space-time models to achieve more flexible models and methods for the analysis of environmental data distributed in space and time. The first stage of the hierarchical model specifies a measurement- error process for the observational data in terms of some 'state' process. The second stage allows for site-specific time series models for this state variable. This stage includes large-scale (e.g. seasonal) variability plus a space-time dynamic process for the anomalies'. Much of our interest is with this anomaly proc ess. In the third stage, the parameters of these time series models, which are distributed in space, are themselves given a joint distribution with spatial dependence (Markov random fields). The Bayesian formulation is completed in the last two stages by speci- fying priors on parameters. We implement the model in a Markov chain Monte Carlo framework and apply it to an atmospheric data set of monthly maximum temperature.     

Evaluating the ecological benefits of wildfire by integrating fire and ecosystem simulation models

Fire managers are now realizing that wildfires can be beneficial because they can reduce hazardous fuels and restore fire-dominated ecosystems. A software tool that assesses potential beneficial and detrimental ecological effects from wildfire would be helpful to fire management. This paper presents a simulation platform called FLEAT (Fire and Landscape Ecology Assessment Tool) that integrates several existing landscape- and stand-level simulation models to compute an ecologically based measure that describes if a wildfire is moving the burning landscape towards or away from the historical range and variation of vegetation composition. FLEAT uses a fire effects model to simulate fire severity, which is then used to predict vegetation development for 1, 10, and 100 years into the future using a landscape simulation model. The landscape is then simulated for 5000 years using parameters derived from historical data to create an historical time series that is compared to the predicted landscape composition at year 1, 10, and 100 to compute a metric that describes their similarity to the simulated historical conditions. This tool is designed to be used in operational wildfire management using the LANDFIRE spatial database so that fire managers can decide how aggressively to suppress wildfires. Validation of fire severity predictions using field data from six wildfires revealed that while accuracy is moderate (30-60%), it is mostly dictated by the quality of GIS layers input to FLEAT. Predicted 1-year landscape compositions were only 8% accurate but this was because the LANDFIRE mapped pre-fire composition accuracy was low (21%). This platform can be integrated into current readily available software products to produce an operational tool for balancing benefits of wildfire with potential dangers.     

Linking 3D spatial models of fuels and fire: Effects of spatial heterogeneity on fire behavior

Crown fire endangers fire fighters and can have severe ecological consequences. Prediction of fire behavior in tree crowns is essential to informed decisions in fire management. Current methods used in fire management do not address variability in crown fuels. New mechanistic physics-based fire models address convective heat transfer with computational fluid dynamics (CFD) and can be used to model fire in heterogeneous crown fuels. However, the potential impacts of variability in crown fuels on fire behavior have not yet been explored. In this study we describe a new model, FUEL3D, which incorporates the pipe model theory (PMT) and a simple 3D recursive branching approach to model the distribution of fuel within individual tree crowns. FUEL3D uses forest inventory data as inputs, and stochastically retains geometric variability observed in field data. We investigate the effects of crown fuel heterogeneity on fire behavior with a CFD fire model by simulating fire under a homogeneous tree crown and a heterogeneous tree crown modeled with FUEL3D, using two different levels of surface fire intensity. Model output is used to estimate the probability of tree mortality, linking fire behavior and fire effects at the scale of an individual tree. We discovered that variability within a tree crown altered the timing, magnitude and dynamics of how fire burned through the crown; effects varied with surface fire intensity. In the lower surface fire intensity case, the heterogeneous tree crown barely ignited and would likely survive, while the homogeneous tree had nearly 80% fuel consumption and an order of magnitude difference in total net radiative heat transfer. In the higher surface fire intensity case, both cases burned readily. Differences for the homogeneous tree between the two surface fire intensity cases were minimal but were dramatic for the heterogeneous tree. These results suggest that heterogeneity within the crown causes more conditional, threshold-like interactions with fire. We conclude with discussion of implications for fire behavior modeling and fire ecology.     

Assessing the risk of ignition in the Russian far east within a modeling framework of fire threat.

The forests of high biological importance in the Russian Far East (RFE) have been experiencing increasing pressure from growing demands for natural resources under the changing economy of post-Soviet Russia. This pressure is further amplified by the rising threat of large and catastrophic fire occurrence, which threatens both the resources and the economic potential of the region. In this paper we introduce a conceptual Fire Threat Model (FTM) and use it to provide quantitative assessment of the risk of ignition in the Russian Far East. The remotely sensed data driven FTM is aimed at evaluating potential wildland fire occurrence and its impact and recovery potential for a given resource. This model is intended for use by resource managers to assist in assessing current levels of fire threat to a given resource, projecting the changes in fire threat under changing climate and land use, and evaluating the efficiency of various management approaches aimed at minimizing the fire impact. Risk of ignition (one of the major uncertainties within fire threat modeling) was analyzed using the MODIS active fire product. The risk of ignition in the RFE is shown to be highly variable in spatial and temporal domains. However, the number of ignition points is not directly proportional to the amount of fire occurrence in the area. Fire ignitions in the RFE are strongly linked to anthropogenic activity (transportation routes, settlements, and land use). An increase in the number of fire ignitions during summer months could be attributed to (1) disruption of the summer monsoons and subsequent changes in fire weather and (2) an increase in natural sources of fire ignitions.     

Conditioning uncertainty in ecological models: Assessing the impact of fire management strategies

A simple simulation model has been used to investigate whether large fires in Mediterranean regions are a result of extreme weather conditions or the cumulative effect of a policy of fire suppression over decades. The model reproduced the fire regime characteristics for a wide variety of regions of Mediterranean climate in California, France and Spain. The Generalised Likelihood Uncertainty Estimation (GLUE) methodology was used to assess the possibility of multiple model parameter sets being consistent with the available calibration data. The resulting set of behavioural models was used to assess uncertainty in the predictions. The results suggested that (1) for a given region, the total area burned is much the same whether suppression or prescribed fire policies are used or not; however fire suppression enhances fire intensity and prescribed burning reduces it; (2) the proportion of large fires can be reduced, but not eliminated, using prescribed fires, especially in areas which have the highest proportion of large fires.     

On the use of the advanced very high resolution radiometer for development of prognostic land surface phenology models

Regulation of interannual phenological variability is an important component of climate and ecological models. Prior phenological efforts using the advanced very high resolution radiometer (AVHRR) as a proxy of vegetation dynamics have often simulated spring events only or failed to simulate interannual variability. Our aim is to address these shortcomings and to use the AVHRR to develop prognostic models for interannual land surface phenology and, critically, to test whether or not the developed models are superior to use of climatological phenology values from the AVHRR. Using datasets for the conterminous United States, we first filtered data to select regions and plant functional types for which the best-possible remotely sensed signal could be obtained. We then used a generalized linear model approach to model the relationship between an integrative productivity index and estimates of the start of season (SOS) and end of season (EOS) derived from the AVHRR, yielding models capable of prognostically predicting SOS/EOS events independently of satellite data. Mean absolute errors between the model-predicted and AVHRR-observed SOS/EOS ranged from 5.1 to 20.3 days. SOS errors were uniformly lower than EOS errors. SOS models for the deciduous broadleaf forest and grassland plant functional types produced lower errors than use of the climatological SOS values while all other models produced errors higher than those obtained from the climatological dates. Based on this criterion for success, we suggest that the AVHRR may not be appropriate for further development of prognostic land surface phenology models. However, an intercomparison of phenological dates from an independent spring index model, our model predictions, and the AVHRR observations indicated that interannual predictions from our models may be superior to the satellite data upon which they are based, implying that a further comparison between models based on the AVHRR and newer, superior sensors, should be conducted.     

Exploring land use determinants in Italian municipalities: comparison of spatial econometric models

Environmental and Ecological Statistics - This study sets up a spatial econometric framework to explore the factors that best describe land consumption in Italy at the municipal level. By modelling...     

Simulating dynamic and mixed-severity fire regimes: A process-based fire extension for LANDIS-II

Fire regimes result from reciprocal interactions between vegetation and fire that may be further affected by other disturbances, including climate, landform, and terrain. In this paper, we describe fire and fuel extensions for the forest landscape simulation model, LANDIS-II, that allow dynamic interactions among fire, vegetation, climate, and landscape structure, and incorporate realistic fire characteristics (shapes, distributions, and effects) that can vary within and between fire events. We demonstrate the capabilities of the new extensions using two case study examples with very different ecosystem characteristics: a boreal forest system from central Labrador, Canada, and a mixed conifer system from the Sierra Nevada Mountains (California, USA). In Labrador, comparison between the more complex dynamic fire extension and a classic fire simulator based on a simple fire size distribution showed little difference in terms of mean fire rotation and potential severity, but cumulative burn patterns created by the dynamic fire extension were more heterogeneous due to feedback between fuel types and fire behavior. Simulations in the Sierra Nevada indicated that burn patterns were responsive to topographic features, fuel types, and an extreme weather scenario, although the magnitude of responses depended on elevation. In both study areas, simulated fire size and resulting fire rotation intervals were moderately sensitive to parameters controlling the curvilinear response between fire spread and weather, as well as to the assumptions underlying the correlation between weather conditions and fire duration. Potential fire severity was more variable within the Sierra Nevada landscape and also was more sensitive to the correlation between weather conditions and fire duration. The fire modeling approach described here should be applicable to questions related to climate change and disturbance interactions, particularly within locations characterized by steep topography, where temporally or spatially dynamic vegetation significantly influences spread rates, where fire severity is variable, and where multiple disturbance types of varying severities are common.     

火烧和植造桉林对南亚热带退化草坡土壤种子库的影响

对南亚热带典型火烧草坡迹地和人工造桉林后的退化草坡土壤种子库进行了取样调查,采用种子萌发试验研究了这些不同处理的土壤种子库特征,结果表明:(1)火烧处理对土壤种子库有显著影响,具体表现为种子库储量减小、物种种类减少、优势种组成单一、物种多样性指数和地上植被相似性系数低、土壤种子库物种组成与地上植被物种组成差异显著;(2)种植桉树对土壤种子库储量、物种生活型分布、土壤种子库物种组成与地上植被物种组成差异均没有影响,但能增加草本物种种类;(3)火烧并种植桉树对土壤种子库的影响与火烧处理结果一致,能使土壤种子库中优势物种种子密度增加,但不能增加草本种类在种子库中的密度.该研究表明采用火烧处理有利于南亚热带退化草坡植被恢复;采用种桉树的方法短期内植被恢复的效果不明显;退化草坡的植被恢复需要适度的人为干预与调控.     

Assessment of ecological risks induced by land use and land cover changes in Xiamen City,China

The landscape ecological risk (LER) in Xiamen City, China, from 1990 to 2030 was studied using an urban land use and land cover change (LUCC) model and LER analysis. The LUCC model was used to predict the LUCC of Xiamen from 2020 to 2030. We analyzed the characteristics of LUCC and landscape pattern changes and, finally, evaluated the effect of rapid LUCC on LER. Of the six landscape types investigated, built-up land and farmland demonstrated the most significant changes. The area of built-up land increased by 1.5 times in 2010 and is predicted to increase by 2.7 times in 2030 than that in 1990. The area of farmland increased from 34.5% in 1990 to 24.5% in 2010 and is predicted to decrease to 15.1% in 2030. The number of patches (NP) of built-up land decreased with increasing area, which promoted the dominance of built-up land over other landscape types. Five landscape types, those other than built-up land, increased in NP, landscape fragmentation, segmentation, and disturbance but decreased in dominance. The LER of Xiamen in 2010 was slightly lower than that in 1990. However, with the acceleration of urbanization, the LER in 2020 and 2030 will increase by 7.6% and 12.5% than that in 2010. The LER will significantly increase in areas such as the Huandong sea area, the second urban core of Xiamen, and northern Xiang'an. For the areas, some measures (e.g. optimum urban spatial growth patterns and control of coastal reclamation) must inevitably increase to reduce the LER posed by rapid urbanization.     

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.     

A method for mapping fire hazard and risk across multiple scales and its application in fire management

This paper presents modeling methods for mapping fire hazard and fire risk using a research model called FIREHARM (FIRE Hazard and Risk Model) that computes common measures of fire behavior, fire danger, and fire effects to spatially portray fire hazard over space. FIREHARM can compute a measure of risk associated with the distribution of these measures over time using 18 years of gridded DAYMET daily weather data used to simulate fuel moistures to compute fire variables. We detail the background, structure, and application of FIREHARM and then present validation results of six of the FIREHARM output variables that revealed accuracy rates ranging from 20 to 80% correct depending on the quality of input data and the behavior of the fire behavior simulation framework. Overall accuracies appeared acceptable for prioritization analysis and large scale assessments because precision was high. We discuss advantages and disadvantages of the fire hazard and risk approaches and a possible agenda for future development of comprehensive fire hazard and risk mapping is presented.     

Evidence, exaggeration, and error in historical accounts of chaparral wildfires in California.

For more than half a century, ecologists and historians have been integrating the contemporary study of ecosystems with data gathered from historical sources to evaluate change over broad temporal and spatial scales. This approach is especially useful where ecosystems were altered before formal study as a result of natural resources management, land development, environmental pollution, and climate change. Yet, in many places, historical documents do not provide precise information, and pre-historical evidence is unavailable or has ambiguous interpretation. There are similar challenges in evaluating how the fire regime of chaparral in California has changed as a result of fire suppression management initiated at the beginning of the 20th century. Although the firestorm of October 2003 was the largest officially recorded in California (approximately 300,000 ha), historical accounts of pre-suppression wildfires have been cited as evidence that such a scale of burning was not unprecedented, suggesting the fire regime and patch mosaic in chaparral have not substantially changed. We find that the data do not support pre-suppression megafires, and that the impression of large historical wildfires is a result of imprecision and inaccuracy in the original reports, as well as a parlance that is beset with hyperbole. We underscore themes of importance for critically analyzing historical documents to evaluate ecological change. A putative 100 mile long by 10 mile wide (160 x 16 km) wildfire reported in 1889 was reconstructed to an area of chaparral approximately 40 times smaller by linking local accounts to property tax records, voter registration rolls, claimed insurance, and place names mapped with a geographical information system (GIS) which includes data from historical vegetation surveys. We also show that historical sources cited as evidence of other large chaparral wildfires are either demonstrably inaccurate or provide anecdotal information that is immaterial in the appraisal of pre-suppression fire size. Since historical evidence is inadequate for reconstructing a statistical distribution of pre-suppression fire sizes to compare with post-suppression data, other more propitious methods of evaluating change are discussed.     

Hierarchical modeling for extreme values observed over space and time

We propose a hierarchical modeling approach for explaining a collection of spatially referenced time series of extreme values. We assume that the observations follow generalized extreme value (GEV) distributions whose locations and scales are jointly spatially dependent where the dependence is captured using multivariate Markov random field models specified through coregionalization. In addition, there is temporal dependence in the locations. There are various ways to provide appropriate specifications; we consider four choices. The models can be fitted using a Markov Chain Monte Carlo (MCMC) algorithm to enable inference for parameters and to provide spatio–temporal predictions. We fit the models to a set of gridded interpolated precipitation data collected over a 50-year period for the Cape Floristic Region in South Africa, summarizing results for what appears to be the best choice of model.

Hierarchical Bayesian Spatial Models for Multispecies Conservation Planning and Monitoring

Abstract: Biologists who develop and apply habitat models are often familiar with the statistical challenges posed by their data's spatial structure but are unsure of whether the use of complex spatial models will increase the utility of model results in planning. We compared the relative performance of nonspatial and hierarchical Bayesian spatial models for three vertebrate and invertebrate taxa of conservation concern (Church's sideband snails [Monadenia churchi], red tree voles [Arborimus longicaudus], and Pacific fishers [Martes pennanti pacifica]) that provide examples of a range of distributional extents and dispersal abilities. We used presence–absence data derived from regional monitoring programs to develop models with both landscape and site‐level environmental covariates. We used Markov chain Monte Carlo algorithms and a conditional autoregressive or intrinsic conditional autoregressive model framework to fit spatial models. The fit of Bayesian spatial models was between 35 and 55% better than the fit of nonspatial analogue models. Bayesian spatial models outperformed analogous models developed with maximum entropy (Maxent) methods. Although the best spatial and nonspatial models included similar environmental variables, spatial models provided estimates of residual spatial effects that suggested how ecological processes might structure distribution patterns. Spatial models built from presence–absence data improved fit most for localized endemic species with ranges constrained by poorly known biogeographic factors and for widely distributed species suspected to be strongly affected by unmeasured environmental variables or population processes. By treating spatial effects as a variable of interest rather than a nuisance, hierarchical Bayesian spatial models, especially when they are based on a common broad‐scale spatial lattice (here the national Forest Inventory and Analysis grid of 24 km² hexagons), can increase the relevance of habitat models to multispecies conservation planning.     

Correction to: Exploring land use determinants in Italian municipalities: comparison of spatial econometric models

Environmental and Ecological Statistics -     

退耕还林后圪针耳流域的土地承载力

山西省中阳县圪针耳流域土地承载力计算结果表明 ,流域内 1 5°以上坡地应全部退耕还林 ,人均基本口粮田不能低于 0 .0 66hm2 ;退耕还林后流域内极限土地承载力为 2 80 2人 ,大于预测人口数量。退耕还林后宜农地生产能保证粮食安全 ,并能使退耕地退得下、稳得住     

Task selection by workers of the fire ant Solenopsis invicta

The effects of worker size, age, and crop fullness on the flow of food into the colony were assessed using video recording and playback. Regardless of the level of colony satiation, small workers seldom had full crops and were more involved in larval grooming than in food traffic. Large workers played little role in larval care, but tended to be recruited easily to a food source and to store food in their crops. Medium workers had crops ranging from empty to full because they alternated between ingesting from and donating food to other colony members. Medium workers were the most versatile, engaging competently in food recruitment, larval grooming, and larval feeding. They displayed considerable variation in the frequency at which they fed larvae: some fed a few larvae before switching to other tasks, others fed over a hundred larvae before switching. The persistence, or lack thereof, of a worker's feeding response suggests a flexibility unaccounted for by the fixed-threshold-response hypothesis. Worker coverage of the brood pile was a dynamic equilibrium process unaffected by worker size, age, or crop fullness, or by differences in the nutritional or hygienic states of larvae. In summary, it appeared that worker size and age offered coarse regulation of task selection by workers, whereas crop fullness, flexible response, and task switching fine-tuned task selection. Received: 25 May 1998 / Accepted after revision: 20 August 1998