The LANDIS fuel module is a new addition to LANDIS. It tracks fine fuel, coarse fuel, and live fuels, simulates various fuel management practices, and estimates the impact of fire risk control (He et al. 2004). In the fuel module, fine fuels (FF) represent primarily foliage litter fall and small dead twigs less than ¼ inch in diameter. They are the primary prerequisite for fire ignitions. Coarse fuels (CF) include any dead tree materials that have a diameter >= 3 inches. These include snags, stems, boles, harvest stumps, and standing dead trees, which affect the intensity class of fire. Live fuels, also called canopy fuels, are live trees that may be ignited during high intensity fires (e.g., crown fires).
Calculating the amount of fine fuel involves deriving a fine fuel amount based on species age and modified by the fuel quality coefficient (FQC). Fine fuel from different species may have different flammability due to differences in physical and chemical attributes (Brown et al. 1982). We use a fuel quality coefficient (0 < FQC ≤ 1) to summarize such differences on a relative scale. Species with low FQC contribute less to the flammability of fine fuels than species with high FQC. FQCs are ranked and parameterized by users, and the same value (e.g., FQC = 1) can be used for all species if there are no discernible differences among species. The result is not the absolute quantity of fine fuels, but rather an “effective index” of the amount of fine fuel that accounts for its flammability. If there are n species in a cell, the total amount of fine fuel (FF) in this cell is calculated as:
where Agei is the age of the oldest cohort of the i th species, Longi. is the maximum longevity of the i th species (also used elsewhere in LANDIS), and FQC is as previously defined. Dividing by n averages the amount of fuel across all species present in the cell. Because LANDIS tracks only the presence and absence of species/age cohorts, the design for simulating fine fuels assumes that all species present in a cell have the same density. Such an assumption may not be realistic for individual cells, but at the landscape scale with millions of cells, relative species abundance can be realistically approximated (He et al. 1998). Fine fuel production changes through the lifespan of a given species. In this example (fig. 10), the amount of fine fuel created by each species (the thin line) is positively correlated with age until approximately 70% of the species lifespan is reached, and becomes negatively correlated with age as the species age approaches maximum species longevity. The actual value calculated is translated into five categorical fine fuel classes using the user-defined relationship (fig. 10). Such relationships can be defined for each species and specific ecosystems based upon empirical data.
Unlike fine fuels, coarse fuels are not derived using species-specific age cohorts. Instead, stand age (the oldest age cohorts) in combination with disturbance history (time since last disturbance) is used to determine the coarse fuel accumulation for a cell (Brown and See 1981, Harmon et al. 1986, Spetich et al. 1999, Spies et al. 1988, Sturtevant et al. 1997). Coarse fuel amount is the interplay between input and decomposition (Spies et al. 1988; Sturtevant et al. 1997). Such interplays may vary by land types (Harmon et al. 1986), which encapsulate environmental variables (e.g., climate, soil, slope, and aspect).
In the absence of disturbance, the accumulation process dominates until the amount of coarse fuel reaches a level where decomposition and accumulation are in balance (Bergeron and Flannigan 2000, Sturtevant et al. 1997), as depicted in figure 10. For example, on a mesic land type with high decomposition rates, the amount of coarse fuel can be low, whereas on a xeric land type with low decomposition rates, the amount of coarse fuel can be high. The decomposition process is modeled based on the decomposition curve based upon previous studies (Foster and Lang 1982, Hale and Pastor 1998, Lambert et al. 1980, MacMillan 1988) (fig. 10). Such a decomposition curve is also user-defined for each land type.The example suggests two decomposition trajectories of coarse fuels on two different land types (fig. 10).
The accumulation and decomposition curves together form the general “U-shaped” temporal pattern observed in many forest ecosystems (Spetich et al. 1999, Sturtevant et al. 1997). In many boreal and northern hardwood forest ecosystems, a land type can seldom accumulate enough coarse fuel to reach class five unless there are other disturbance events occurring such as wind, BDA, and/or harvest. Users can define these disturbance-related accumulations using coarse fuel accumulation and decomposition curves (fig. 10).
Due to the long temporal scales involved in estimating the amount of coarse fuels, uncertainty is high. Collapsing the estimates of the quantity of coarse fuel into five categorical classes (very low to very high) reduces the potential for false precision and the parameterization burden for the module.
Figure 10.- Fine fuel and coarse fuel accumulation and decomposition. The relationship can be further modified by land type, disturbance, harvest, and fuel treatment.
The elapsed time of fuel accumulation (TFC) is used to determine the current amount of coarse fuel as shown in figure 10. The various modifiers, once activated, will determine how much the coarse fuel class is increased or reduced. The relationships defined for each modifier (15.2) and the decomposition status defined in figure 10 is used to determine the final coarse fuel amount. The highest class (up to class 5) will be retained as the final coarse fuel class. For example, based upon the time of coarse fuel accumulation for a site, the coarse fuel class is determined to be class 2. However, an intense wind disturbance and an insect defoliation would each raise the coarse fuel class by 3. This leads to a final coarse fuel class for this site that is larger than 5 (2+3+3). In such a case, the model will set the coarse fuel class to 5.
Land type, fire, wind, harvest, and biological disturbance can modify the fine fuel and coarse classes derived from the relationships discussed above. Fine fuel decomposition rates may vary by land types (Agee and Huff 1987). Thus, a land type modifier may decrease or increase the fine fuel class derived from species age cohorts. For example, a fine fuel class 3 on a mesic land type might decrease to class 2 because the decomposition rate is relatively high, while the same fuel class (3) on a xeric land type might increase to class 4, since the decomposition rate is relatively low. The user defines the modifier in the fuel module input file, and the default is no modification. Disturbances may increase or decrease the fuel class in a similar way. For example, as a parameter for a Biological Disturbance Agent (BDA) (e.g., insect pest), the user defines how a BDA disturbance event will increase the fine fuel and coarse fuel class, depending on the type and intensity of the event. Similarly, fire events can reduce the fine and coarse fuel class (Armour et al. 1984) based on user-defined rules. The simplest and most common case is that fires remove all fine fuels. Alternatively, a rule could remove fine and coarse fuels in proportion to the intensity class of the fire. Wind disturbances primarily increase coarse fuel. However, it can also increase fine fuels by producing dead leaves and needles. Increases in fine fuel classes caused by wind can be determined by the intensity class of the wind. Harvest activities can also modify the fine and coarse fuel class. The user can also define how each harvest prescription defined in the Harvest module will modify the fine and coarse fuel class.
Potential fire intensity is determined by the combination of fine fuel and coarse fuel in each cell. A set of rules can be defined (see section 15.3.1) based upon the assumption that coarse fuel is the primary contributor to the fire intensity class, since in many forest ecosystems coarse fuel accounts for about 90% of forest floor mass (Grier and Logan 1977, Lambert et al. 1980, Lang and Forman 1978). Users can define other rules according to the ecosystems they study. In the example in section 15.3.1 high-intensity fires are not common compared to low-intensity fires. Seven fine and coarse fuel combinations result in fire intensity = 1 (very low), seven in fire intensity = 2, six in fire intensity = 3, three in fire intensity = 4, and two in fire intensity = 5 (very high),
Live fuels are live trees that may be ignited in high-intensity fire situations (such as crown fires). Thus, live fuels can be a fire intensity modifier. A mid-level intensity fire ( >= 3) may change into a crown fire (intensity class = 5) if there are suitable conifer species present (e.g., FQC = 1). However, changing from low-intensity fire to crown fire is not a deterministic event and a probability function is used to predict its occurrence. For example, the probability of low-intensity fire changing to a crown fire is 0.01 - 0.05 based on the empirical knowledge for Missouri central hardwoods (B. Cutter, Department of Forestry, University of Missouri-Columbia, personal communication). Such a probability (P) can be user defined. In the fuel module, when fire intensity reaches level 3 and there are species with FQC = 1 present, the fire intensity can reach 5 if the uniform random number > P.
Fire risk has a variety of definitions under different modeling frameworks. In the Fire Regime Condition Class fire risk is defined as the risk of loss of key ecosystem components (Hardy et al. 2001) . In the National Fire Plan, fire risk refers to the risk that communities and environment will experience a damaging fire (National Fire Plan Link). When the Forest Service forecasts fire risk, they define it as the probability or chance of fire starting determined by weather index. In LANDIS 6.0, the definition of fire risk reflects the two key aspects: 1) the probability of fire occurrence, and 2) the intensity and spread once a fire ignition occurs (He et al. 2004). Fire probability is derived from the fire cycle and the time since last fire for each cell and is intended to account for climate and other factors that can influence ignition probability, whereas fire intensity and spread are related to biotic factors affecting the level of various fuels (Yang et al. 2004). In the LANDIS 6.0 fuel module fire probability is converted into five classes, from very low to very high, based upon the “equal area” (the fire probability density function is divided into 5 areas of equal size) or equal interval approach. Potential fire intensity is determined by the combination of fine fuel and coarse fuel in each cell.
Fire risk is classified into five classes based upon fire probability and fire intensity, from very low (class 1) to very high (class 5). We assume that potential fire probability and fire intensity equally contribute to the fire risk. Thus, in the default fire risk table, five unique combinations of fire probability and potential intensity classes are identified for each fire risk class. Again, users can define this table based upon the characteristics of their study area. The module allows users to track the high fire risk areas on the landscape over time and to explore the effects of various fuel load reduction methods on the spatial and temporal dynamics of fire risk.
The LANDIS fuel module simulates fuel management practices that fall into two categories: prescribed burning and physical fuel load reduction (removal and mechanical thinning). LANDIS fuel management has s patial, temporal, and treatment components. The spatial component uses parameters on the desired treatment size (e.g., the percent area of a management unit to be treated) and determines where such a treatment can be spatially allocated (e.g., how stands are selected for treatments). The allocation criteria can be based upon rankings of potential fire risk, where stands with highest potential fire risk are treated first, or by using random selection. The temporal component of fuel management determines what year (decade) a given treatment is performed and how often it is repeated. Single, multiple, or periodical entry years can be specified. The treatment component specifies the treatment types (e.g., prescribed burning) and treatment intensity. Since the three components are independent, combinations of the three are capable of simulating most fuel treatment practices (He et al. 2004, Shang et al. 2004).
In LANDIS, prescribed burning mainly affects fine fuel but it can also reduce coarse fuel based upon the user specification. The choice of low- vs. high-intensity prescribed burning treatments depends on the field conditions, resources, and potential fire risk (Brose and Wade 2002). For example, a low-intensity prescribed burning might reduce the fine fuel load by a maximum of 2 classes. A high-intensity prescribed burn might remove most fine fuel loads (reduced to 1).
Mechanical thinning primarily targets coarse fuels, including reducing the fuel size and removing/reducing coarse fuel load. The low-intensity treatment reduces coarse fuel load by 1-2 classes when the coarse fuel class is >3. The high-intensity treatment removes most coarse fuels (reduced to class 1).
The following example illustrates how to specify chipping and thinning of coarse fuel. The chipping and thinning treatment is prescribed only for high ( = 4) and very high ( = 5) coarse fuel classes, and classes 4 and 5 are reduced to class 2. Note that coarse fuel treatments can result in an increase in fine fuels. In this example, all fine fuel classes are increased by 1.
Coarse fuel load class before treatment: 0 1 2 3 4 5
Coarse fuel load class after treatment: 0 1 2 3 2 2
Fine fuel load class before treatment: 0 1 2 3 4 5
Fine fuel load class after treatment: 1 2 3 4 5 5