Chapter 1 | Chapter 2 | Chapter 3 | Chapter 4 | Chapter 5 | home
DISCUSSION
Risk from wildfire to the Wildland Urban Interface will continue to increase across District 6 well into the next decade. There are three primary reasons for the increase in risk to rural populations: increasing rural population pressures, increasing forest fuel loads, and increasing ignition from human activity. The Census Bureau forecast continued growth in Arkansas' rural population over the next 10 years; 32% of this growth will occur in northwest Arkansas. By the year 2000 one in six persons will live in northwest Arkansas. The buildup of forest fuels across the nation will continue for several reasons: the publics resistance toward certain forestry management techniques, concerns about insurance liability, areas sensitive to smoke management, the tendency of the 'new' rural homeowner toward encouraging vegetative growth, especially bushes, inside of a fire free zone.
There is some evidence to suggest that the population estimates by the Census Bureau for several counties in northwest Arkansas (Benton, Carroll, Crawford, Madison, and Washington) may be too conservative. In 1994, Washington County had approximately 13,000 rural residences. The Census Bureau estimated rural housing to increase to over 20,000 by the year 2000. In 1998, the Washington County 911 system had over 24,000 residences logged into its emergency response system (Gibson, 911 Coordinator, 1998, oral communication).
Two noteworthy developments in northwest Arkansas will ensure long-term economic development in the region. First, a major transportation interchange will be completed by the year 2,000, bringing about the development of additional farm land and
timberlands. Secondly, a regional airport will open this fall providing employment and economic opportunity to the region. Along with this economic bonanza will come pressure on rural infrastructure, more rural residences, and increased risk from wildfire.
Although not part of this study weather is a key element in understanding the behavior of wildfires. Future improvements of the Risk Potential Model would be the integration of weather elements, especially dew point, drought conditions, and wind vectors. Since ancient times lightning and anthropogenic fires have shaped our environments, landscapes, and even human culture. Weather's effects on wildland fire cannot be overstated.
Weather elements: temperature, wind speed, wind direction, and relative humidity have direct influences on the extent and severity of wildland fires. Weather not only influences a fire but in some cases may be the source ignition, for example ground lightning from a cumulous thunderstorm. Large wildfires can modify local weather conditions; wildfires are not simply two dimension events but are also meteorological phenomenon's providing thermal energy and condensation nuclei for development of thunderstorms. Examples of weather influences on wildfires in both the temporal and spatial domains can be found, from a millisecond bolt of lightning to an extended drought, from a single thunderstorm to global changes such as El Nino. The tools used in this study, remote sensing and geographic information systems are useful tools for analysis of weathers effect on wildfire.
Three environmental elements affect wildfires: fuels, topography, and weather. From a fire fighters perspective weather is the single most important element influencing the behavior of a wildland fire. Fire fighting efforts to control wildfires relies to a large extent on the cooperation of the weather. In past wildfire incidents, a disregard for changing weather conditions has had catastrophic results in the loss of life and increased property damage. Mild winters and wet summers contribute toward fire danger through increased fuel loads. Changing global weather patterns and climatic elements such as long term drought can predispose an area to a higher degree of risk to wildfire.
Drought records along with the Palmer Drought Severity Index may give some insight into patterns of wildfire occurrence. Phillips (1987), using tree-ring indices noted drought cycles of every 14.8 years in Arkansas. Other studies have shown drought periodicity ranged from 16 to 21 years (Jacoby et al. 1985, Currie 1979, Mock and Hibler 1976, Wagner 1971). In this study the Palmer Drought Severity Index (PDSI) was found to be useful as an indices in wildfire occurrence. The Pearson Correlation for the 1992 to 1996 wildfire incidence reports showed a strong negative relationship between PDSI values and wildfire occurrence, -0.90 statewide and -0.93 within District 6. A similar inverse relationship also exist in the Kluender et.al (1988) study which showed a -0.84 relationship between PDSI and wildfire occurrence between 1983 and 1987. PDSI for 1995 moved from a 'wet' value of +1.0 in the spring, then to normal 0.0 in midsummer, and then into 'incipient' drought -0.5 by fall. That year fire occurrence increased 53% statewide and 46% in District 6.
With the frequency and extent of wildfire damage held in check fire becomes excluded from plant communities it help shape or even dominate. Wildfire has two effects on a plant community. First it reduces small fuels in the understory, fuels which act as a ladder into tree canopies. Secondly, it removes litter buildup (twigs, limbs, leaves/needles) on the forest floor reducing the intensity of the next wildfire. This action also reduces certain forest pest improving the overall health of the forest.
State and Local Rural Fire Organizations
Presently protection against wildfire at the Wildland Urban Interface comes from two sources. The Arkansas Forestry Commission maintains a wildfire protection presence in every county in Arkansas and uses planes to constantly patrol for wildfires. However, when wildfire occurs in the Interface the local volunteer fire department is usually first respond, especially in counties with 911 emergency responder systems. No one single fire organization has yet to evolve in the Interface, currently two separate agencies must coordinate their efforts to be effective against wildland fires in the Wildland Urban Interface. State foresters bring wildland fire fighting techniques to the Interface using: air tanker drops, fire plows, hand equipment, and backfires to modify fuels ahead of the wildfire. Local volunteer fire departments bring urban fire fighting techniques to the Interface using: pressurized water systems and hand equipment to combat wildland fires.
On the front lines with the Arkansas Forestry Commission are the rural volunteer fire departments (RFD) in Arkansas. Of the 1000 fire departments in the state 80% are volunteer fire departments. They often lack the equipment and training to combat wildfires should the fire escape into the forest. Water sources may be distant and the RFD's may be limited to the water that they carry to the scene. The Arkansas Forestry Commission supplies federal surplus equipment, grants, and loans to rural volunteer fire departments as part of a cooperative fire protection program. To be eligible the rural volunteer fire department must agree to respond to any fire within a five mile radius of their station (Larry Nance - Arkansas Forestry Commission, oral communication).
The fire regime in the northwest Arkansas is very similar to that of eastern Oklahoma. Oklahoma is plagued with large wind driven grass fires that can have a rate-of-spread that is faster than a person can walk. The fact that these areas are lightly populated has prevented a major loss to rural residences. Northwest Arkansas on the other hand is much more densely populated, however the pastures and hay fields tend to be interspersed with more woodlots than eastern Oklahoma.
Fuel models if left undisturbed will move through successional stages along with the changes in the vegetative land cover. For example, when farmland (primarily grazing lands and hay fields) is taken out of production it will quickly move from a short grass fuel to a tall grass fuel within one growing season. As pioneer tree species, especially eastern red cedar (Juniperus virginiana) and white cedar (Juniperus ashei), invades the fuel types become more volatile. These fuels are more susceptible to weather changes, dew point and wind patterns. The successional progression if left unchecked will move from grassland to woodlot to closed canopy.
The Wells and Mckinsey (1991) study indicated that lack of variability in the number of variables limited the ability to accurately map their region wildfire potential and model wildfire behavior. However, personnel with the Arkansas Forestry Commission feel the need for simpler model which field personnel can be trained to recognize and accurately report. In some cases usage of the national fire behavior models, which are based upon West Coast fuel types, led to errors when reporting fuel types. Additionally, it is also unrealistic to expect untrained volunteer fire department personnel to be able to recognize the behavior of wildfire in 13 or 20 different fuel types. Part of this study looked at way to generalize through the use of matrices fuel types into usable volatility classes. Anderson's (1982) 13 fuel behavior models were reduced into three volatility classes (high, moderate, low) based upon the rate-of-spread as described by Anderson.
The Risk matrix is comprised of four separate components: ignition potential component, fuel (volatility) potential component, response potential component, and access potential component. The Risk Potential matrix values are: high risk potential, moderate risk potential and low risk potential, and expresses the overall risk to a rural residence. The four components, when combined together, could produce over 100 separate categories. This large number of categories in impractical in an operational environment and requires a high degree accuracy when developing the individual components. A decision matrix was used to simplify the models. If a model showed a high risk level a query into the individual data sets, which make up the model, would give more site specific information. Spatial operators, such as coincidence tables, can store information about individual layers within the risk potential data set.
The Risk Potential Matrix (Figure 12) map is comprised of two separate matrices, the threat matrix and the suppression matrix, its values are high risk potential, moderate risk potential, and low risk potential. The risk potential of wildfire for District 6 (Table 31) indicates: (1) 13% (124) rural residences are at a high risk, this risk category comprises 1.45% (33,145 Ac) of the total area, (2) 17.35% (7,138) rural residences are at a moderate risk from wildfire, this risk category comprises 76.14% (1,740,525 Ac) of the total area, and (3) 81.82% (34,118) rural residences are at a low risk from wildfires, this risk category comprises 20.44% (44,352 Ac) of the total area.
The Threat Matrix (Figure 10) has three values: high risk potential, moderate risk potential and low risk potential. It represents threats to rural residence based upon the likelihood of a wildfire ignition and the proximity of different fuel volatility classes.
The Threat Matrix (Table 27) of wildfire to rural residences in District 6 indicates: (1) 1.67% (698) rural residences are at a high risk from the threat of wildfire, this risk category comprises 1.82% (41,100 Ac) of the total area, (2) 3.75% (1,565) rural residences are at a moderate risk from the threat of wildfire, this risk category comprises 4.96% (112,006 Ac) of the total area, and (3) 94.27% (39,311) rural residences are at a low risk from wildfire, this risk category comprises 91.28% (2,061,265 Ac) of the total area.
The Suppression Matrix (Figure 11) has three values: high risk potential, moderate risk potential and low risk potential. It represents the risk that once a wildfire has begun or threatens that adequate personnel and equipment will arrive in sufficient time to protect a rural residence from a wildfire. This risk potential is comprised of: (1) the response potential component, can adequate equipment and personnel reach a rural residence in a timely manor, and (2) the suppression potential component, once on the scene of a wildfire can a rural fire department gain access to property and the fire scene. Road conditions play a key role in the level of risk assigned to this component. The suppression matrix (Table 29) of wildfire by rural fire departments in District indicates: (1) 13.34% (5,563) rural residences are at a high risk from inadequate wildfire suppression, this risk potential component comprises 77.97% (1,781,767 Ac) of the total area, (2) 4.31% (1,797) rural residences are at a moderate risk from inadequate wildfire suppression, this risk potential component comprises 1.8% (41,104 Ac) of the total area, and (3) 82.32% (34,327) rural residences are at a low risk from inadequate wildfire suppression, this risk potential component comprises 20,32% (462,185) of the total area. In summary the further a rural residence is from a rural fire department or the type of road used to access the property indicates the level of risk the rural homeowner faces in regard to wildfire suppression.
A. Ignition Potential Component
Ignition Potential Component (Figure 6) has two categories based upon the occurrence of a wildfire from 1992 to 1996. Each wildfire point source was buffered 20 acre to represent the historical average size of a wildfire. The Ignition Potential Component comprises of: (1) 1.69% (43,535 Ac) with a history of wildfire and contains 705 rural residences in District 6.
B. Volatility Classes (Table 22 & 23)
The Volatility Classes (Figure 7) were made up of three categories: high risk from forest fuels, moderate risk from forest fuels, and low risk from forest fuels. The high risk category is comprised of 1.74% (39,732 Ac) of the total area, 3.79% (1,579) of the rural residences are located in this volatility class. The moderate risk category is comprised of 92.46% (2,108,620 Ac) of the total area, 95.72% (39,916) of the rural residences are located in this volatility class. The low risk category is comprised of 0.68% of the total area, 0.19% (79) of the rural residences can be found in this category.
Access Potential (Figure 8) is made up of four categories: all weather roads (paved and gravel), seasonal roads (graded), limited access roads (no-maintenance), and no access (residences > 300 feet from roads). Each road category was buffered to identify those areas that are accessible by rural fire department equipment. All weather road category contained 74.48% (31,059) of the rural residences. Seasonal road category contains 21.22% (8,850) of the rural residences. Limited road category contains 0.17% (67) of the rural residences. No access road category contains 4.13% (1,724) of the rural residences.
D. Response Potential Component
Response Potential (Figure 9) Component has two categories: those areas within 5 miles and those areas that are more than 5 miles away from a rural fire station. Each rural fire station was buffered 5 miles. The within 5 miles category 93.22% (38,874) of the rural residences are located. The greater than 5 miles category 6.78% (2,826) of the rural residences are located.
The aggressive nature of wildfire has not dissipated since its beginning, what has changed are societys attitude toward wildfire. These attitudes have led directly or indirectly, knowingly and unknowingly, to changes in the source of wildfire igniton, wildfire frequency, and a buildup of fuels. A wildfire is a creature of fuels, weather, and topography and most suppression efforts focus on removal or modification of fuels directly adjacent to the wildfire. Because of social, economic, and political reasons the wildfire policy is one of extermination. Today, firefighters wage a constant battle to hold wildfires in check. Thrown into these wildfire combat zones are new, rapidly growing, rural residential suburbs; scattered, isolated, remote homes. Two events could upset the balance between wildfires and state and local fire organizations. First, the continuing expansion of the wildland urban interface into remote areas increases pressures on the notion of a 'adaquate' fire protection whether the fire's orgin is structure or wildfire. The second event which has a potential of overwhelming suppression capabilities is 'periodic' drought.
This project demonstrates that those factors which influence wildfires ignition and volatility can be measured and mapped using spatial tools such as GIS, GPS, and remote sensing. These tools have proven effective in collecting, cateloging, analysising, and producing risk potential models. These analytical tools allows development of digital data sets which can be recodded, reclasified, buffered, and analysised in a geospatial environment. Existing data sets provide adaquate spatial and temporally accurate information leading to useful products, the AHTD data set which provided residential and road type informantion is scheduled to be updates once every seven years. Once the model is completed future changes in data can be quickly assimulated into the model thus updating the model products. Some of the Wildland/Urban Risk Model products identify: the location of residences in high risk fuels, high risk roads, high risk from inadaquate fire protection (based on response time or distance from a RFD), etc. The products could provide information for both short-term fire suppression decisions and long term-strategic planning decisions within the Interface. These products could be used by firefighters to predict fire risk based upon information compiled about human activity, fire behavior, fuel types, fuel loads, environmental conditions, and topology. With this information in hand the fire supervisor can then determine priority of multiple fires based upon other considerations such as public safety, economic losses, and political implications.
Chapter 1 | Chapter 2 | Chapter 3 | Chapter 4 | Chapter 5 | home