Estimating canopy fuels and their impact on potential fire behavior for ponderosa pine in the Black Hills, South Dakota

We evaluate whether current procedures used in fire behavior prediction models such as FVS-FFE provide predictions of CBD and CBH suitable for evaluating fire behavior in response to fuel treatments. Currently, FFE-FVS uses a geographic non-specific set of tree allometries and assumes a uniform distribution of crown mass when estimating CBH and CBD. We develop allometric equations to predict crown mass specific to ponderosa pine in the Black Hills (Pinus ponderosa Dougl. ex Laws.) from a sample of 80 felled trees in 16 forest stands spanning a wide range in tree size and stand. We develop a non-uniform description of vertical crown mass within individual trees using the Weibul distribution. We relate the parameters of the distribution to stand structure, so that a vertical canopy biomass profile can be estimated for any stand from standard inventory information. We modify the existing FVS-FFE program to include our results. Estimates of CBD increased by an average of 78% when using our local biomass and non-uniform vertical distribution models compared to current procedures in FVS-FFE. On average, 47% of the underprediction of the current procedure compared to our new models resulted from site specific allometries and additional 31% of the under prediction resulted from a non-uniform distribution of crown mass. Our results suggest locally-derived crown mass equations in addition to non-uniform estimates of crown mass distribution should be used to calculate CBH and CBD as used in fire prediction models. Current management efforts to create stand structures more resistant to the initiation and spread of crown fire include increasing CBH and reducing CBD below the threshold where crown fire can be initiated and carried through the tree canopy. Of the 16 stands sampled in this study, only two had CBD estimates >0.100 kg m (i.e. the CBD where active crown fire would be expected) as currently implemented in FFE-FVS. When local crown mass equations and distribution models were applied to the data, 12 out of the 16 stands had CBD estimates >0.100 kg m threshold. Consequently, FFE-FVS, as presently formulated, would misdiagnose fire hazard in a substantial number of Black Hills ponderosa pine stands. Further, where FFE-FVS is used to design and evaluate fuels treatments, it is probable that either the amount of density reduction necessary to achieve a desired effect will be underestimated, or the longevity of effectiveness of a given treatment will be overestimated. BACKGROUND AND PURPOSE Long-term maintenance of fuels reduction treatments is a primary goal for forest management in the ponderosa pine (Pinus ponderosa Laws.) forests of the Black Hills, South Dakota. The Black Hills are dominated by young, dense, even-aged ponderosa pine stands and the prevalence of these stands can result in large, catastrophic wildfires, such as the 34,000 ha Jasper fire of 2000. Many of these stands are in the wildland – urban interface, where reducing the likelihood of crown fire behavior in the event of a wildfire is extremely important. Annually, ~2,300 ha are being treated to reduce the potential for crown fire, primarily by thinning to reduce canopy density. Fire resistant structures created by stand level fuels treatments are not static: canopy density increases with time as trees grow and regeneration is recruited into the overstory canopy. With this increased canopy density, active crown fire behavior again becomes likely. Accurate projections of how canopy density changes with initial treatment and how canopy density increases with time are crucial in determining how, when, and how often fuels treatments are preformed. We propose to develop improved methods for estimating the amount and vertical distribution of canopy fuels from forest inventory data and to integrate these estimators of canopy fuels with models of forest growth and fire behavior (i.e. the Forest Vegetation Simulator (FVS) and the Fire and Fuels Extension (FFE). Fuels treatments to reduce crown fire behavior in Black Hills ponderosa pine forests are most commonly commercial or precommercial thinning to reduce stand density below 12 m/ha (50 ft/acre) of basal area. These densities are thought to reduce the amount of canopy fuels below the level that will support crown fire. Stands thinned to these densities will have high basal area growth rates, with rapid development of canopy density. Also, ponderosa pine regeneration will be prolific at these stand densities, and will create ladder fuels and increased canopy density as regeneration grows in height and crown size. Initial treatment effectiveness and longevity of treatment effects are evaluated using estimates of surface fuels and canopy structure in surface and canopy fire behavior models. Two types of crown fire behavior are predicted based on stand structure and weather conditions – passive or active crown fire. Passive crown fire occurs when there is a sufficient density of canopy fuels to spread surface fire vertically from lower to upper canopy reaches at a given wind speed. The height in a canopy where there are sufficient fuels to spread flames vertically is called the canopy base height (CBH). Active crown fire occurs when there is sufficient density of fine fuels (e.g. foliage and small branches) at any height in the canopy to indicate sufficient continuity of canopy fuel to carry fire from tree to tree at a given wind speed. The density (kg/m) of needles and small branches used to determine continuity of canopy fuels is called canopy bulk density (CBD). Estimates of the amount and vertical distribution of canopy fuels are critical to accurate estimates of the threshold wind speed where passive or active crown fire will occur. The primary method by which federal land managers predict CBD is through the use of the growth and yield model, the Fire and Fuels Extension to the Forest Vegetation Simulator (FVS-FFE). While FVS-FFE provides a working prediction of CBD through time, the underlying assumptions and equations used to calculate and predict CBD may not accurately represent CBD or CBH. The equations used by FVS-FFE to predict crown mass for ponderosa pine created by Brown (1978) are based on trees from northern Montana and Idaho. These equations, therefore, may not capture variability in crown mass due to geographic, site, or stand variability. Currently, when calculating CBD, FVS-FFE uses a uniform distribution of foliage and branchwood within individual crowns (eg. crown mass divided by crown length). But, crown mass is not evenly distributed within individual tree crowns – crown mass is distributed as a skewed normal distribution, with less mass at the top and bottom a tree crown and most of the mass concentrated near the center of the crown . We hypothesize that both of these problems may result in the underestimation of CBH and CBD in FVS-FFE and therefore provide inaccurate estimates of potential fire behavior. There are many potential benefits from this proposed research. First, accurately predicting crown mass is the foremost important step in calculating CBD. If current crown mass equations for ponderosa pine do not adequately describe crown mass in different geographic regions, current predictions of CBD may be inaccurate. By comparing crown mass equations for ponderosa pine in the Black Hills to those calculated by Brown (1978) and used in FVS-FFE, this research will help determine if there is a need for region specific crown mass equations for individual species. Second, by incorporating a correction factor for foliage distribution into CBD calculations, effects of density and canopy position on canopy fuel distribution will be taken into consideration vastly improving the accuracy of CBD predictions. We suggest that the proposed modifications to the current methodology for predicting CBD will greatly improve land managers ability to evaluate initial treatment effectiveness and to plan for maintenance of fuels treatment projects. 1. We develop equations to predict the amount of forest canopy fuels for Black Hills ponderosa pine to be used with standard forest inventory data and that are compatible with fire behavior models such as FVS-FFE. Equations currently used to estimate canopy fuels are based on a small sample size from a specific and limited geographic range. Our results identify whether current techniques produce sufficiently accurate estimates of canopy fuels to provide a realistic analysis of changed fire behavior from fuel treatments. 2. We develop a technique to accurately predict the vertical distribution of fuels within treated canopies and untreated canopies of Black Hills ponderosa pine. The current approach to estimate CBD from tree inventories assumes a uniform distribution of crown mass. This approach may cause an underestimate of canopy bulk density and an inaccurate estimate of canopy base height. Our results identify whether a more accurate technique for vertical distribution of fuels within tree crowns will improve estimates of canopy bulk density. 3. We test the effect of our estimators of canopy fuel and canopy fuel distribution on the determination of canopy bulk density, canopy base height, and potential crown fire behavior in stands treated for fuels reduction in Black Hills ponderosa pine as compared to the current methods of prediction in FVS-FFE. We will provide our results to the Forest Management Service Center as a set of equations and documentation intended for incorporation into the FVS-FFE model. STUDY DESCRIPTION AND LOCATION