Haiganoush K. Preisler

Statistical Model for Forecasting Monthly Large Wildfire Events in Western United States

Abstract The ability to forecast the number and location of large wildfire events (with specified confidence bounds) is important to fire managers attempting to allocate and distribute suppression efforts during severe fire seasons. This paper describes the development of a statistical model for assessing the forecasting skills of fire-danger predictors and producing 1-month-ahead wildfire-danger probabilities in the western United States. The method is based on logistic regression techniques with spline functions to accommodate nonlinear relationships between fire-danger predictors and probability of large fire events. Estimates were based on 25 yr of historic fire occurrence data (1980–2004). The model using the predictors monthly average temperature, and lagged Palmer drought severity index demonstrated significant improvement in forecasting skill over historic frequencies (persistence forecasts) of large fire events. The statistical models were particularly amenable to model evaluation and production ...

Climate and weather influences on spatial temporal patterns of mountain pine beetle populations in Washington and Oregon

Widespread outbreaks of mountain pine beetle in North America have drawn the attention of scientists, forest managers, and the public. There is strong evidence that climate change has contributed to the extent and severity of recent outbreaks. Scientists are interested in quantifying relationships between bark beetle population dynamics and trends in climate. Process models that simulate climate suitability for mountain pine beetle outbreaks have advanced our understanding of beetle population dynamics; however, there are few studies that have assessed their accuracy across multiple outbreaks or at larger spatial scales. This study used the observed number of trees killed by mountain pine beetles per square kilometer in Oregon and Washington, USA, over the past three decades to quantify and assess the influence of climate and weather variables on beetle activity over longer time periods and larger scales than previously studied. Influences of temperature and precipitation in addition to process model output variables were assessed at annual and climatological time scales. The statistical analysis showed that new attacks are more likely to occur at locations with climatological mean August temperatures >15 degrees C. After controlling for beetle pressure, the variables with the largest effect on the odds of an outbreak exceeding a certain size were minimum winter temperature (positive relationship) and drought conditions in current and previous years. Precipitation levels in the year prior to the outbreak had a positive effect, possibly an indication of the influence of this driver on brood size. Two-year cumulative precipitation had a negative effect, a possible indication of the influence of drought on tree stress. Among the process model variables, cold tolerance was the strongest indicator of an outbreak increasing to epidemic size. A weather suitability index developed from the regression analysis indicated a 2.5x increase in the odds of outbreak at locations with highly suitable weather vs. locations with low suitability. The models were useful for estimating expected amounts of damage (total area with outbreaks) and for quantifying the contribution of climate to total damage. Overall, the results confirm the importance of climate and weather on the spatial expansion of bark beetle outbreaks over time.

Spatially explicit forecasts of large wildland fire probability and suppression costs for California

In the last decade, increases in fire activity and suppression expenditures have caused budgetary problems for federal land management agencies. Spatial forecasts of upcoming fire activity and costs have the potential to help reduce expenditures, and increase the efficiency of suppression efforts, by enabling them to focus resources where they have the greatest effect. In this paper, we present statistical models for estimating 1–6 months ahead spatially explicit forecasts of expected numbers, locations and costs of large fires on a 0.125° grid with vegetation, topography and hydroclimate data used as predictors. As an example, forecasts for California Federal and State protection responsibility are produced for historic dates and compared with recorded fire occurrence and cost data. The results seem promising in that the spatially explicit forecasts of large fire probabilities seem to match the actual occurrence of large fires, with the exception of years with widespread lightning events, which remain elusive. Forecasts of suppression expenditures did seem to differentiate between low- and high-cost fire years. Maps of forecast levels of expenditures provide managers with a spatial representation of where costly fires are most likely to occur. Additionally, the statistical models provide scientists with a tool for evaluating the skill of spatially explicit fire risk products.

PREDICTING THE OCCURRENCE OF RARE MOLLUSKS IN NORTHERN CALIFORNIA FORESTS

Terrestrial mollusks are important components of forest ecosystems, yet we know very little about the distribution and habitat of many of these species. We sampled for terrestrial mollusks in northern California with the goal of estimating the geographic ranges and developing predictive habitat models for five species that were assumed to be sensitive to land management activities. The species of interest were Ancotrema voyanum, Helminthoglypta talmadgei, Monadenia churchi, Monadenia fidelis klamathica, and M. f. ochromphalus. We randomly selected 308 plots for sampling from a grid of points across a 2.2 million-ha study area. We used Generalized Additive Models to estimate each mollusks geographic range and to develop predictive habitat models within their ranges. Models were developed at one microscale (1 ha) and six mesoscales (ranging from 12.5 to 1250 ha) using vegetation, physical, climatic, and spatial location covariates. Estimated geographic ranges varied from 4770 to 15 795 km2. Predictive habi...

Forecasting distributions of large federal-lands fires utilizing satellite and gridded weather information

The current study presents a statistical model for assessing the skill of fire danger indices and for forecasting the distribution of the expected numbers of large fires over a given region and for the upcoming week. The procedure permits development of daily maps that forecast, for the forthcoming week and within federal lands, percentiles of the distributions of (i) number of ignitions; (ii) number of fires above a given size; (iii) conditional probabilities of fires greater than a specified size, given ignition. As an illustration, we used the methods to study the skill of the Fire Potential Index - an index that incorporates satellite and surface observations to map fire potential at a national scale - in forecasting distributions of large fires.

Wildland fire probabilities estimated from weather model-deduced monthly mean fire danger indices

The National Fire Danger Rating System indices deduced from a regional simulation weather model were used to estimate probabilities and numbers of large fire events on monthly and 1-degree grid scales. The weather model simulations and forecasts are ongoing experimental products from the Experimental Climate Prediction Center at the Scripps Institution of Oceanography. The monthly average Fosberg Fire Weather Index, deduced from the weather simulation, along with the monthly average Keetch–Byram Drought Index and Energy Release Component, were found to be more strongly associated with large fire events on a monthly scale than any of the other stand-alone fire weather or danger indices. These selected indices were used in the spatially explicit probability model to estimate the number of large fire events. Historic probabilities were also estimated using spatially smoothed historic frequencies of large fire events. It was shown that the probability model using four fire danger indices outperformed the historic model, an indication that these indices have some skill. Geographical maps of the estimated monthly wildland fire probabilities, developed using a combination of four indices, were produced for each year and were found to give reasonable matches to actual fire events. This method paves a feasible way to assess the skill of climate forecast outputs, from a dynamical meteorological model, in forecasting the probability of wildland fire severity with known precision.

A statistical approach to estimate O3 uptake of ponderosa pine in a mediterranean climate

In highly polluted sites, stomatal behavior is sluggish with respect to light, vapor pressure deficit, and internal CO2 concentration (Ci) and poorly described by existing models. Statistical models were developed to estimate stomatal conductance (gs) of 40-year-old ponderosa pine at three sites differing in pollutant exposure for the purpose of calculating O3 uptake. Gs was estimated using julian day, hour of day, pre-dawn xylem potential and photosynthetic photon flux density (PPFD). The median difference between estimated and observed field gs did not exceed 10 mmol H2O m(-2) s(-1), and estimated gs within 95% confidence intervals. 03 uptake was calculated from hourly estimated gs, hourly O3 concentration, and a constant to correct for the difference in diffusivity between water vapor and 03. The simulation model TREGRO was also used to calculate the cumulative 03 uptake at all three sites. 03 uptake estimated by the statistical model was higher than that simulated by TREGRO because gas exchange rates were proportionally higher. O3 exposure and uptake were significantly correlated (r2>0.92), because O3 exposure and gs were highly correlated in both statistical and simulation models.

Modeling and risk assessment for soil temperatures beneath prescribed forest fires

Prescribed fire is a management tool used by wildland resource management organizations in many ecosystems to reduce hazardous fuels and to achieve a host of other objectives. To study the effects of fire in naturally accumulating fuel conditions, the ambient soil temperature is monitored beneath prescribed burns. In this study we developed a stochastic model for temperature profiles (values at 15 minute intervals) recorded at four depths beneath the soil during a large prescribed burn study. The model was used to assess the temporal fit of the data to particular solutions of the heat equation. We used a random effects model to assess the effects of observed site characteristics on maximum temperatures and to estimate risks of temperatures exceeding critical levels in future similar prescribed fires. Contour plots of estimated risks of temperatures exceeding 60°C for a range of fuel levels and soil depths indicated high risks of occurrence, especially when the moisture levels are low. However, the natural variability among sites seems to be large, even after controlling fuel and moisture levels, resulting in large standard errors of predicted risks.

Analyzing animal movement patterns using potential functions

The advent of GPS technology has made it possible to study human-wildlife interactions on large landscapes and quantify behavioral responses to recreation and other anthropogenic disturbances at increasingly fine scales. Of particular interest are the potential impacts on habitat use patterns, energetics, and cascading impacts on fecundity and other life history traits for key wildlife species that are exposed to human activities. Statistical models quantifying effects of human activity on animal movement on a heterogeneous landscape are essential for understanding these potential impacts. Here we present a statistical framework for analyzing movement data that is based on the concept of a potential surface. The potential surface is motivated by the assumption that animals are moving on a space-time surface with regions or points of attraction or of repulsion. We demonstrate the use of the technique by analyzing movement data from a long-term controlled experiment to evaluate the responses of free ranging Rocky Mountain elk (Cervus elaphus) to anthropogenic disturbances that vary in time and space. Our results demonstrated a strong avoidance of elk to all-terrain vehicles detected up to one km from the disturbance. Elk avoidance of mountain bikers was detected up to 500 m, and avoidance of hikers and horseback riders was detected to 200 m.

The Use Of Potential Functions In Modelling Animal Movement

Potential functions are a physical science concept often used in modelling the motion of particles and planets. In the work of this paper potential function based models are considered for the movement of free-ranging elk in a large, fenced ex- perimental forest. Equations of motion are set down and the parameters involved are estimated nonparametrically. The question of whether a potential function is plausible for describing the elk motion is considered. The conclusion is that it is not possible to reject this hypothesis for the data set and estimates considered.