5.4 FLOODS AND FIRES: FAIRBANKS RESPONSES TO DISASTERS

[Table of Contents] [Previous Chapter] [Next Chapter] [Main Page]

[Next Section] [Previous Section] [Chapter Head]

Bowers, Peter M., and Owen K. Mason (1998) 5.4 Floods and Fires: Fairbanks Responses to Disasters. In Historical Development of the Chena River Waterfront, Fairbanks, Alaska: An Archaeological Perspective, edited and compiled by Peter M. Bowers and Brian L. Gannon, CD-ROM. Alaska Department of Transportation and Public Facilities, Fairbanks.

Throughout the nearly one century Fairbanks has been in existence, the community has been beset with disasters, both natural and man-made. This section looks at some of the floods and fires which have affected Fairbanks over the years, and identifies a few of the ways the community has coped with these disasters. We do not present a comprehensive list of all such problems; instead we focus our discussion only on the evidence uncovered at the Barnette excavations and relevant historical records which corroborate those data. Because the Chena River has played such an important role in the development of downtown Fairbanks, we focus much of this discussion on the hydrology and flood history of the river.

The city of Fairbanks began in 1901-02 along the banks of the Chena Slough as a consequence of an historical accident. Due to the placement of Barnette’s Cache on the bank of a seasonally flood-prone slough, the waterfront has suffered from repeated ice jams, over-bank floods, and bank erosion. Until the Alaska Railroad was completed in the 1920s, the river was the main route for transportation and shipping, and the main focus of the city’s landscape. However, the depredations of flooding were at variance with the commercial value of riverfront property; because of the value of river navigation to the town, residents attempted to control the river by a variety of shoring techniques in the downtown area and by constructing levees upstream (e.g. Moose Creek Flood Control Project).

Much larger engineering measures were undertaken in the 1930s, after riverboat traffic had ceased. This involved re-channeling the Chena River, resulting in a decreasing flow from the Tanana River main channel.³⁰ These flood control efforts led to a decrease in stream flow that restricted navigability and contributed to the re-definition of downtown Fairbanks away from its former riverfront. Today, the shallow and narrow Chena River more closely resembles a drainage ditch than the vibrant and vital transportation artery it once was.

Hydrology of Interior Alaska Rivers

The hydrologic pulse of northern rivers is driven by two types of water input: spring melt of snowfall (exacerbated by ice damming) and summer rain storms.³¹ The Chena River flow is at a minimum during winter as ice and low temperatures restrict its movement. Typically, river flow increases steadily after spring as snow melts, the ground thaws, and water is added to the river by overland flow. Peak flows often occur during May or June along the presently-controlled Chena River through Fairbanks. Ice breakup on the modern Chena River proceeds quickly and effortlessly due to the addition of warm waste water from the utility plants. However, as many old-time residents recall, breakup on the Chena Slough once resembled other interior rivers by its sudden and often catastrophic nature.

The ice debris produced during breakup can be re-assembled downstream into a temporary ice dam that impounds water and causes a local rise in river level. Ice jam flooding was a perennial likelihood each spring in Fairbanks and represented the most probable type of flood to reach the town. The greatest physical danger was the incorporation of drift debris and the transportation of huge ice blocks onto the waterfront, making these floods more damaging than simple water level rises. Ice blocks weighing several hundred pounds or more, moving erratically or suddenly, presented a formidable threat to the integrity to the town’s frame buildings. The efficacy of early erosion control methods was limited, as was demonstrated when the 1911 flood lifted pilings and destroyed revetments placed by the NC Co. along the river bank (Figure 5.6).³²

Figure 5.6. Breakup of Chena Slough, May 1911, taking out the pilings of the Turner Street bridge. Vertical Files [VF 94.01], Alaska and Polar Regions Archives, University of Alaska, Fairbanks.

Water levels can also rise suddenly if rainfall is locally heavy and exceeds the ability of the ground surface to absorb the water. Extreme rainfall events are comparatively rare in Fairbanks before July, and the largest rains in the interior typically occur during August. However, exceptional rainfall-generated floods can occur earlier in the year. The May 1948 flood, for example, the second largest on record, was a result of high rainfall combined with breakup flooding. The most substantial rainfall (and flood) on record occurred during several days in mid-August 1967 over the headwaters of the Chena River, and culminated in a widespread flood that engulfed the entire town.

Today, the Chena and Tanana rivers respond to different hydrological regimes. The comparatively small drainage of the Chena River receives most of its flow from snow melt and occasional precipitation events. The Tanana River drains a considerable portion of interior Alaska north of the Alaska Range, and its discharge reflects mid-summer glacial melt as well as the cumulative effects of breakup and meteorological conditions in six sizable rivers. The Chena commonly reaches a discharge peak early in the year while the discharge on the Tanana peaks later, usually in summer.³³ Hydrographic data collected since 1949 at a downtown Fairbanks gauge suggests that maximum discharge in the Chena can vary dramatically, but most often (19 of 43 years of record or 44%) occurs during May, and in fewer than 14% of years after August 31. Generally, discharge is between 6,500 and 9,000 cubic feet per second (cfs), with the extreme values of 24,200 cfs in May 1948 and 74,400 cfs in August 1967. High-water events on the Tanana River are more restricted, with over 80% of high discharge events occur during July or August. The effects of flood control engineering have negated the dominance of the Tanana River pattern and rendered the Chena River more sensitive to drainage from just the Chena instead of the entire Tanana drainage.³⁴

Flood History of Fairbanks

At present, the only means of reconstructing prehistoric flooding frequencies is through the use of geologic data, determined from the radiocarbon dating of organic surfaces capped by flood sediments.³⁵ Such data are available only for the Tanana River, which may be out of phase with the Chena River because of differences in the size of catchment areas of each river.

Summer floods are more likely to be visible due to higher sediment yield during summer events.³⁶ The Luke’s Slough locality on the Tanana River, about 10 km southwest of downtown Fairbanks, contains sediments deposited during more than 20 floods over the last 3,000 years. Mason and Begét³⁷ estimated that large floods have increased in frequency and magnitude over the last 400 years. Whereas floods had occurred on average every 110 years between AD 100-600 (a warm interval associated with frequent forest fires), flooding intervals became shorter, averaging every 60 years, during the Little Ice Age, circa AD 1550-1800. The Luke’s Slough data, derived from a study locality downstream from the mouth of the Chena River, indicate that flooding in interior Alaska accelerated during glacial episodes and declined during periods of warmer climatic conditions.

Major floods have hit Fairbanks every 10 to 20 years since 1905.³⁸ The most frequent flood type is due to the breakup or ice jam event that is generated in May or June, but floods caused by high rainfall or a combination of both factors have also occurred. Major floods occurred in the Fairbanks Area (Table 5.22) in 1905, 1911, 1930, 1937, 1948, 1960, 1963, and 1967.³⁹ In terms of economic loss the 1948 and 1967 floods were the most catastrophic: the 1948 event impacted over one third of Fairbanks⁴⁰ and the 1967 event covered nearly all of the city.⁴¹ The 1967 flood was the most costly because of damage to capital equipment and infrastructure (i.e. sewer lines, water and electricity facilities, store inventories and automotive equipment). However, unlike the ice jams of the pre-war period, few buildings were structurally damaged in 1967.

Table 5.22. Selected Data on Floods in Fairbanks, 1905-96.⁴²

Floods And River Bank Stabilization: Historical Data

In an effort to protect the town from flood events, flood control efforts began almost as soon as the town was settled; by August 1905, erosion control had already cost the nascent city some $10,000 (Figure 5.7).⁴³ Following major floods in 1905, 1911, 1930, 1937 and 1948, numerous attempts were made at shoring the river bank in the downtown area. Following the May 1937 flood, residents and engineers alike believed that the slough was in the process of becoming the main channel of the Tanana (Figure 5.8).⁴⁴ As a result, an upstream levee was built between 1938 and 1941 to mitigate ice jam flooding effects from the Tanana River; this represented the first flood control efforts in Alaska by the U.S. Army Corps of Engineers. However, these flood control designs failed to prevent the 1967 event, the worst flood to hit Fairbanks in recorded history.⁴⁵ A major re-engineering of the flood control system in the 1980s incorporated a dam system in North Pole designed to check Chena River flooding.

Figure 5.7. The Fairbanks Waterfront in 1908, showing the City Dock and erosion control measures that were taken following the 1905 flood. First Avenue has been rebuilt; pilings and brush beneath the street can be seen to the right of the dock on the far side of the river. Lulu Fairbanks Collection [32.05], Alaska and Polar Regions Archives, University of Alaska, Fairbanks.

Figure 5.8. Breakup in 1937, view to the southwest along First Avenue. City of Fairbanks Collection.

Within the small spatial focus of our research, a relatively large number of photographs have been located which reveal a variety of historic construction techniques for stabilizing the river bank. Table 5.23 lists photographs found in Rivertown⁴⁶ to illustrate the range of techniques and materials used to help stabilize the river bank. The reader is referred to that report for a fairly comprehensive analysis and chronology of river bank stabilization techniques.

Table 5.23. Observations of First Avenue river bank erosion and improvements based on analysis of historical photos in Rivertown.⁴⁷

Floods And River Bank Stabilization: Archaeological Evidence

Substantial evidence of past flood events, represented mostly by prominent sand layers and erosional unconformities, were found throughout the Barnette Site excavations. These are documented in the discussion of site stratigraphy in Appendix 1. In addition, numerous bank stabilization and bank extension features were documented during the two seasons of fieldwork. These include:

• drainage features oriented perpendicular to the river;
• log "deadman" timbers set perpendicular to the river;
• brush and sand used as fill to stabilize river bank;
• log cribbing below the NC Co. steamboat dock;
• NC boiler ash used as fill and bank protection;
• trash disposal used as fill and bank protection;
• street fill used to raise street grade; and
• rock riprap used to armor the river bank.

"Deadman" Timbers. Deadman timbers, used to help stabilize the river bank and anchor dock foundations, were found in Areas A (Features 7 and 8) and B1 (Feature 5). These large timbers are set in the bottom of trenches at least 1 m deep and 50 cm wide, oriented perpendicular to the Chena River. The bottoms of the trenches are in sterile sediments. A stratigraphic profile (Figure A1.11) shows that the trenches are capped by early historic deposits (Levels 1-4). Based on stratigraphic position and on historic photos showing construction of similar features along Front Street, they probably date between 1904 and 1906.

Log Cribbing Below the NC Dock. The log structure below the site of the NC Co. Dock and warehouse was found at the northern end of Area B2 (Figures 4.51 and 4.52). It is located less than 2 meters south of the modern Chena River. The feature was buried by a mixed deposit of NC boiler ash (Level 2) and alluvial silt (Level 3). The deeply-buried logs extend into the natural (pre-1901) alluvial silt found on the south bank of the Chena River. This feature consists of log cribbing, constructed from large logs set into the river bank. Logs (greater than 2 m long) running east-west and parallel to the river bank were corner-notched and joined at a right angle with shorter logs (about 1.5 m long), parallel to the river bank. Probably built in 1907, this feature primarily provided erosion control for the river bank below the structure. Our excavation exposed a 2 meter-long section of the structure, roughly 2 meters (or 6 logs) high. Although two logs had rolled off this section, the feature was extremely well preserved. Dendrochronology (Appendix 7) of logs from the base of Feature 2 suggests two separate construction events, dating to between 1901 and 1910, but the feature also could have been built by using both salvaged and newly cut timber as a single event. Located in the spaces between the log "cells" was a thick deposit of brush, sand, and small branches (Figure 5.9). These are believed to relate to the rebuilding of First Avenue following the 1905 flood.

Figure 5.9. Log cribbing placed along First Avenue in the former location of the NC Dock, circa 1930. Robert King Collection.

Box Drain. A wooden box drain was found above the remains of the NC Co. Dock (B2, Feature 1). Located in association with Level 4, this feature is capped by mixed deposits of NC boiler ash (Level 2) and alluvial silt (Level 3; Figure 4.46). This feature is a wooden box constructed from dimensional lumber that functioned as a drain, probably conducting water off First Avenue and into the Chena River. The drain extends over what was once the edge of the sloping river bank. It is more than 4 m long but as it continues south beyond the wall of our excavation, it was impossible to determine its total length. It appears that this feature was laid on the surface of the river bank and soil placed around it. Based on the stratigraphic position of this feature and historic photographs which illustrate the construction of similar features along First Avenue, the drain probably dates to about 1904. The photo shown in Figure 2.2, a south facing view of Barnette’s Cache, illustrates a nearly-identical box drain feature.

NC Boiler Ash and Trash. Coal ash and clinkers were deposited over the river bank by the NC Co. between the mid 1920s and 1950. These deposits were found in Areas A, B1, and B2. The ash is highly compacted and hard, making an excellent medium to protect against floods. Trash was also expediently used as erosion control from the earliest days of the gold camp through the 1930s. Approximately 45,000 artifacts, almost one half of the total recovered during the Barnette Project, were recovered during our excavations of the river bank areas; we speculate that most of the material was deposited in an attempt to not only dispose of it, but also to help stabilize the river bank. This distinctive stratigraphic marker horizon, referred to as NC boiler ash in Level 2 of Areas A, B1, and B2, is shown in Figures 4.51 and 4.55. A photograph (Figure 4.59) from the late 1940s depicts this activity quite dramatically.

Modern Fill. Modern fill was encountered throughout the riverside excavations, in Areas A, B1, B2, and C. This includes street fill, which has effectively raised the level of the streets some 3 feet over the years, and boulder rip-rap, which is a recent attempt at securing the riverside sediments against flooding.

Fires

Over the years, fires have plagued the city. The effect of fires has been exacerbated by such factors as the extreme cold, which makes pumping water difficult, and construction techniques such as sawdust insulation and buildings placed adjacent to one another with no fire breaks. Our excavations found direct evidence of two of the major, widely-publicized fires (Figures 5.10 and 5.11), and other evidence of lesser fires.

Figure 5.10. The beginning of the May 22, 1906 fire in downtown Fairbanks. View to the east. Charles Bunnell Collection [1.11], Alaska and Polar Regions Archives, University of Alaska, Fairbanks.

The May 22, 1906 downtown fire destroyed the area between Turner and Lacey Streets and First and Third Avenues.⁴⁸ Archaeological evidence for the 1906 fire was located at several places in our excavations. It was demonstrated most prominently in Area A, where it was located near the base of our excavations in Levels 4b, 4c, and 4d (Figures A1.8, A1.9, and A1.10). At that location, we identified rubble from the fire which had been thrown over the river bank during cleanup of the damage. We documented burned canvas (tents?), burlap, wood, charcoal, charred architectural hardware and burned artifacts from the lowest cultural deposits in Areas A and B1.

Figure 5.11. Historic Photo of the May 22, 1906 fire engulfing the east end of Fairbanks. View to the east along the Chena River waterfront. Charles Bunnell Collection [1.12], Alaska and Polar Regions Archives, University of Alaska, Fairbanks.

We also found extensive archaeological evidence of the April 25, 1946 fire on Lots 1 and 2, Block 16, which destroyed the Chena Bar, the Northern Hotel and partially damaged the Pioneer Hotel (Figure 4.35). Extensive evidence of this was found throughout our excavations in Area C, within several stratigraphic contexts. The fire at that location began on April 25 at 10:15 PM and continued until April 26 at 3:00 AM. The Northern Hotel and coffee shop were a total loss. The Pioneer suffered $4,000 losses in fire damage and looting. The Chena Bar and Grill was heavily damaged and the Chena Liquor Store suffered water damage during attempts to keep that structure from burning.⁴⁹

Other fires in the downtown area, but not directly represented in our excavations, include the April 20, 1935 fire that burned the Alaska Hotel at the corner of First and Wickersham. This fire is reported to have been contained inside the building because of the metal covering on the outside walls.⁵⁰ On Christmas Eve, 1946, the Northern Commercial Telephone Co. was destroyed by a major fire, regarded by many firefighters as the worst fire in modern Fairbanks history.⁵¹ As a result of this disaster, telephone service was severely disrupted.⁵² On July 13, 1952, the Pioneer Hotel was again affected by fire, but this time the building was a total loss.

The ease with which potentially disastrous fires could begin is attested to in the cellar of the Miners’ Home Saloon (Figure 4.64), where we found evidence of fire from a wood stove accident. An area of charred floorboards and melted glass bottles and fragments was uncovered in an area beneath where a stovepipe can be seen in historical photographs rising from the roof (Figure 4.71).

Another fleeting glimpse of fires as seen in the archaeological record of Fairbanks is evidence of the 1920 fire which destroyed the Alaska Daily Citizen building on Garden Island.⁵³ Within our backhoe test trench T1 (Figure 4.4), we found charred wood, charcoal, and one piece of linotype.

Discussion: Floods and Fires

From the perspective of historical archaeology, Fairbanks’ attempts to fight the periodic battles against the forces of the river and fires are remarkable. Until major flood control projects were engineered in the 1980s, flooding was always a constant threat to the community. Similarly, fires were a continual threat, and continue to present dangers today, especially during the times of extreme cold winter. Specific to our research area, erosion of the river bank along First Avenue was regarded with concern by the townspeople, and was dealt with through repeated attempts at bank stabilization. Based on a combination of archaeological data and historic accounts and photographs, the following developments relative to our research can be chronicled.

As early as the summer of 1904, work was underway to provide drainage across Front Street to the Chena River. This was accomplished by digging a trench and employing a wooden box drain (Figure 2.2).

At the time of construction of the NC Co. Dock in 1905, log revetments were built along the riverfront in the vicinity of the dock and upriver along First Avenue between Wendell Street and Barnette Street.⁵⁴ These revetments were built in a contiguous series of "cells," eight foot-long by 6 to 8 logs high, pinned firmly together and placed parallel to the river bank. In our study area the revetments served the purpose of helping to protect the bank from erosion, and perhaps contribute to the support of the NC Co. Dock.

After the severe 1905 flood, which destroyed some 400 feet of First Avenue, the City of Fairbanks let a contract to rebuild the road and provide erosion control.⁵⁵ This effort resulted in a thick matrix of brush, tree limbs, and sand being placed in the void spaces created by the flood. This firmly compacted matrix was then anchored in place with vertical pilings at the water’s edge. Evidence of this technique was revealed in our excavations at the base of the former NC Co. Dock. According to historic sources, this technique was an engineering process widely used at that time.⁵⁶ Reportedly "the brushy lattice laid down in the sandy-gravel matrix would set up as hard as concrete if installed by a competent contractor."⁵⁷ This reconstruction effort was delayed by the disastrous May 22, 1906 fire, although the fire rubble provided a ready source of fill for the gash along First Avenue.⁵⁸ Since part of the pre-fire fill had been swept away by the flood, more log cribbing, in the form of a retaining wall between Wendell St. to the Turner St. Bridge reportedly, was built right after the flood. This wall was apparently extended after 1907 breakup, but we have no material evidence of this. Additional brush and pilings were placed, creating a continuous retaining wall extending from the Turner St. Bridge to the lower end of the Pioneer Dock. The rest of this project was apparently not completed until the summer of 1910, and much of this rebuilt matrix was again swept away in a 1911 flood event.⁵⁹

The erosion control features along First Avenue apparently underwent little change until after the NC Co. and Pioneer docks were torn down. Thereafter, townspeople continued to dump refuse over the river bank, ostensibly for erosion control and convenience of disposal. The by-product of this activity is well represented archaeologically by the thousands of commercial and domestic artifacts and NC boiler ash.

According to several sources⁶⁰ an additional log revetment was added between Cushman Street and Bonnifield Street (located four blocks west of Barnette Street) in about 1926.⁶¹ The remains of this logwork was apparently still visible up until the early 1950s. This structure was formed by a series of non-linear segments of log walls. However, no unequivocal material evidence of this feature was found in our excavations.

Finally, the main erosion control technique documented by our research is the coal ash and clinker deposited by the NC Co. as they cleaned out their boilers from time to time. This activity and resulting refuse material in the ground is well documented both historically and archaeologically (Figure 4.59).

While the Chena River presented a constant threat from erosion and flooding, it also offered a continual downstream flow that provided a ready place of disposal for a wide variety of waste materials. The threat of erosion also led residents to develop several "folk" methods of bank stabilization by dumping unrecyclable materials, debris from fires, and other refuse in the hope of staving off erosion.⁶² This method may have had minimal effects on bank erodibility, but it did preserve a wealth of archaeological data.

Development of Fairbanks had another, more indirect impact on the Chena River. Rivers integrate the hydrological balance of entire drainage basins and the effects of large scale logging contribute to siltation and shallowing of rivers.⁶³ De-forestation around Fairbanks proceeded rapidly due to heavy demand for construction lumber, cordwood for steamboats and NC Co. boilers, and for mine-related activities.⁶⁴ By 1903, local loggers had ventured up the "Big" Chena River, proceeding farther upriver each year. By 1913, loggers had reached 100 river miles from Fairbanks, at the phenomenal rate of 9 miles per year.⁶⁵ One local newspaper indicates that "timber [was] cut from both sides of river for a distance of a mile or so from the stream. Farther away than that the timber becomes scrubby and useless for sawmill purposes."⁶⁶ One historical summary describes the second decade of Fairbanks existence as a time of, "acute fuel shortages, as most of the accessible timber had been cut."⁶⁷

The short term (decadal) effects of historic logging on the Chena River are unstudied but probably included increased siltation due to the fewer number of trees capable of binding sediments. The contribution of logged timber to the debris load of the river is uncertain. These and other historic land-use patterns are illuminated by the detailed studies of the Barnette Site stratigraphy, which has preserved a record of the battle between town and river. Several decades (1910s to circa 1950) of dumping on the river bank reflect a convenient disposal location, erosion control efforts, and a mind set that eschewed aesthetic considerations.

[Next Section] [Previous Section] [Chapter Head]

[Table of Contents] [Previous Chapter] [Next Chapter] [Main Page] [Top of Page]