|
<br />.'
<br />
<br />wildfire areas. Extensive coordination was made with the various government agencies involved
<br />with hazard mitigation, particularly the Nationai Weather Service (NWS). An integral part of our
<br />research focuses on helping to determine threshold-rainfall conditions, which will change with time
<br />that could cause hazardous flooding in burned areas. Additional research objectives are to: 1.) ,
<br />determine the length of time for the basin to return to pre-fire conditions; 2.) assess the
<br />effectiveness of watershed-management practices used to mitigate water and sediment runoff; 3.)
<br />compile literature on wildfire-hazard mitigation; and 4.) develop a paleoflood-based monitoring
<br />program that provides data needed for subsequent ecological and forest rehabilitation activities.
<br />Limited resources preclude extensive, instrumented data-collection efforts. There also is a
<br />likelihood floods may not occur in instrumented basins. Monitoring needs to begin immediately
<br />after a wildfire because important data are lost shortly after each storm and hazard-mitigation
<br />efforts. Monitoring in burned and unbumed areas consists of determining rainfall amounts, peak
<br />flows, and channel agradation and degradation.
<br />
<br />The study approach was applied to the community of Buffalo Creek, located about 50 km
<br />southwest of Denver, Colorado (fig. 1). On May 18, 1996, an intense wildfire, now known as the
<br />Buffalo Creek wildfire, burned about 50 km2 of forest. Following the fire, rainstorms produced nine,
<br />1 OO-year or larger flash floods (pre-fire conditions). Two people were killed and several million
<br />dollars in public and private property damage were caused by the largest flood on July 12, 1997
<br />(Colorado Water Conservation Board, 1997). Additionally, about $ five million dollars were spent
<br />fighting the fire and on watershed-rehabilitation efforts (USFS, written commun., 1998). Colorado
<br />is the first state in the Nation to create a statewide wildfire fire assessment map, which outlines
<br />areas most vulnerable to catastrophic wildfires (USFS, written commun., 1997). Buffalo Creek is
<br />located in a moderate fire-hazard area, thus, illustrating the potential hazards in other wildland-
<br />urban interface areas.
<br />
<br />Background
<br />
<br />Wildfires, which change flood and sediment production, are a natural process critical to maintaining
<br />healthy ecosystems and have occurred numerous times in the western United States during
<br />about the last 8,000 years (Weise and Martin, 1995; Meyer et aI., 1995). In 1997, wildfires
<br />burned more than 25,000 km2 in the western United States, which was the most area since 1952
<br />(National Interagency Fire Center, written commun., 1997). Substantial resources are directed
<br />towards mitigation of water and sediment runoff in bumed areas; however, little is known about
<br />the effectiveness of watershed-rehabilitation practices. Most studies done to assess the effects
<br />of wildfires on watershed hydrology have been for chaparral vegetated areas (dense, low
<br />evergreen oaks) in southern Califomia (e.g., Florsheim et aI., 1991; Weise and Martin, 1995).
<br />Fewer studies have been done for other forest ecosystems, and most of these focus on sediment
<br />runoff (e.g., White and Wells, 1981; Parrett, 1987; Meyer et aI., 1995; Cannon et aI., 1995; Ewing,
<br />1996). Because southern California has different vegetation and soil types than other forested
<br />areas in the United States, study results may not be transferable or may need to be modified for
<br />different forest ecosystems. Hydrophobic (water-repellent soil) conditions often develop after a
<br />wildfire from the combustion of vegetation and decomposing organic matter, which produces
<br />aliphatic hydrocarbons that move as a vapor through the soil and substantially reduce infiltration
<br />(USFS, 1979a). Hydrophobic soils, decreased vegetation cover, and reduced surface storage
<br />following wildfires dramatically increase the potential for extreme flooding, and sediment transport
<br />and deposition (USFS, 1979b). Fire-suppression activities make it difficult to estimate the natural
<br />frequency of fires from historical fire records.
<br />
<br />One goal of state and federal (e.g., USFS, BLM, and NPS) forest managers is to baiance
<br />maintaining healthy forest ecosystems and minimizing hazards to the public and property. In the
<br />late 1950s, after more than half a century of active fire suppression, which increased the threat of
<br />catastrophic wildfires, greater emphasis was placed on prescribed burning in forest lands to
<br />reduce the buildup of fuelwood (Weise and Martin, 1995). Since 1984, prescribed burning has
<br />been used on an average of 1,200 km2 per year in US national forests, which have a total area of
<br />773,000 km2 (USFS, written commun., 1997). The effects of fire intensity on water and sediment
<br />runoff and the ecosystem are poorly understood.
<br />
<br />Draft 3/30/98
<br />
<br />2
<br />
|