EFFECTS OF WICKERSHAM DOME FIRE
ON
WATER QUALITY OF WASHINGTON CREEK
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENVIRONMENTAL
RESEARCH CENTER
ALASKA WATER LABORATORY
College, Alaska
-------
EFFECTS OF WICKERSHAM DOME FIRE
ON
WATER QUALITY OF WASHINGTON CREEK
Frederick B. Lotspeich
for the
ENVIRONMENTAL PROTECTION AGENCY
ALASKA WATER LABORATORY
COLLEGE, ALASKA
Working Paper No. 14
-------
A Working Paper presents results of investigations
which are to some extent limited or incomplete.
Therefore, conclusions or recommendations—expressed
or implied—are tentative.
-------
Conclusions and Recommendations
Based on field observations during three visits to the Wickersham
Dome fire area, and analyses of water samples collected on each trip,
the following conclusions and recommendations are offered:
1. Retardants used on this fire did not appear to cause high phosphate
or nitrogen concentrations in Washington Creek, the major stream draining
the burned over area.
2. Suspended sediment load increased a few days after the fire as a
result of thermal erosion of fire trails. Most of this erosion was
brought under control by constructing water diversion dams across fire
lines to divert melt water from streams to vegetated areas.
3. Where possible, retardants should be confined to areas some distance
from running water to prevent adding loads of these concentrated chemicals
to streams. There appeared, however, to be little damage under the con-
ditions of this fire, even after sufficient rain had fallen to cause
Washington Creek to rise.
4. Bulldozed firelines should terminate before reaching stream banks
and diversion dams should be installed at frequent intervals across fire-
lines as part of the overall fire plan.
5. Followup work on diversion dams should be done at critical points to
insure that melting permafrost does not cause failure as was observed in
the field.
6. Where possible, firelines should be located in areas of rocky ridges,
rather than at midslope with deep soils, to prevent erosion.
-------
Introduction
On June 24, 1971, lightning started a forest fire on a ridge about
five miles north of the Washington Creek bridge crossing of the Elliot
Highway. Later, smoke from this fire was plainly visible from Fairbanks,
about 25 miles south. Because the fire was so near Fairbanks, the Bureau
of Land Management (BLM) mounted a strenuous effort to control the fire;
a total of 710 firefighters and 40 pieces of equipment were mobilized.
About 70 miles of firelines were built in a week's time. Despite these
efforts, the fire quickly spread southwest in a narrow band until, by
June 27, it had run about 13 miles (Figure 1). This fire did not spread
further and, by July 1, it was considered to be under control although
"mopping up" continued for several more days. A total area of about
20,000 acres was burned over with a range of severity from complete kill
to unburned islands. In addition to the ground effort at fire control —
bulldozed firelines and hand-clearing—planes dropped about 60,000 gallons
of chemical fire retardant in large concentrations at critical areas
(Figure 1).
Nearly the entire burned area drains into Washington Creek, a medium-
size stream with measured discharge of about 28 cfs below the fire. A
sizeable proportion of the retardants was dropped near Washington Creek
at the southwest portion of the burned area. Since the chemical retardant
contained both nitrogen and phosphorous compounds, streams draining the
burned area could have been enriched with these elements if they were
flushed off the drop zones by rain or melt water. No runoff was evident
-------
WICKERSHAM DOME FIRE
JUNE 24-30,1971
ELLIOTT HIGHWAY
DIRECTION OF FIRE RUN
SMALL STREAMS
R R R RETARDANT DROPS
1000 GALLONS OF RETARDANT
X2-4 STATIONS ON W. CREEK
LIMITS OF FIRE
Figure 1. Map of fire area showing major streams, dates and direction of fire runs, points where retardants
were dropped, and sampling points on Washington Creek.
-------
because rainfall was very light; any water entering the drainages was from
permafrost meltwater on bulldozed firelines.
Another stream pollutant associated with fires and fire fighting is
silt introduced into streams. A previous study by AWL concluded that
melting permafrost may introduce more sediment from improperly sited
bulldozed firelines than that arising from erosion of burned over forest.
-------
Field Work and Results
A trip was made to the fire on July 1 to collect samples of Washington
Creek and record observations. Emphasis was placed on checking the effects
of fire retardants on water quality and any increase in sedimentation.
BLM provided a helicopter and two stations were established below
the burn. Station 1 (Figure 1) was located at the lower-most fireline
and Station 2 at a good riffle about 3 miles downstream from the fireline.
Station 3 was established about 1/4 mile above the Elliot Highway bridge,
across Washington Creek. Later (July 8) Station 4 was established on
Washington Creek, about 1/2 mile upstream from Lost Horse Creek. This
site was selected because Lost Horse Creek drained the upper fire area
where retardant had been used in the first day of fire fighting. A total
of three sampling trips was made—July 1, 8, and 13.
Field work consisted of collecting the following samples and field
measurements: biological samples at all stations on July 1 and 8, chemical
samples for nitrogen and phosphorus at all stations on all dates, sediment
samples at all stations on July 8 and 13, discharge measurements at all
stations on July 1 and 8, and field chemistry measurements of water at
stations 1 and 3 on July 1 and at all stations on July 13. Data from
these analyses, except for the biological samples, are presented in Tables
1 and 2.
Waters of Washington Creek are similar in properties to those of many
streams of Interior Alaska; they are brown colored but not turbid except
at high water stage. Table 1 shows some chemical properties (measured
-------
TABLE 1
PROPERTIES OF WASHINGTON CREEK WATER
Field Measurements
Station
1
2
3
1
2
3
4
1
2
3
4
Date
7/1/71
7/8/71
7/13/71
n
Water
Temp. °C
11.3
8.5
12.8
13.8
12.8
11.4
5.8
6.5
4.1
4.7
Cond.
umho/cm
220
220
82
90
68
77
8.2
8.2
Alk. mg/1 P.O. mg/1
87
91
7.3
7.3
28
32
29
28
11.0
11.3
11.5
11.2
-------
in the field) at high and low discharges. The lower values for conduc-
tivity, alkalinity and pH at high water represent the diluting effects
of runoff after moderate rainfall on July 11 and 12. Water temperatures
were also reduced because of cold water entering the stream and lowered
insolation during prolonged cloud cover. Table 2 summarizes the laboratory
data derived from all samples collected.
Sediment samples were not collected on July 1, but suspended sediments
in samples collected on July 8 were low at all stations. Such low values
probably resulted from little or no runoff since only light rainfall
occured during that time interval. On July 1 some sediment was observed
entering Washington Creek from small tributaries that were carrying
meltwater from bulldozed firelines. Collection and analysis of sediment
was not anticipated at that time, therefore no data is available. BLM
personnel informed us that considerable sediment was added to the streams
on July 2, following termination of fire fighting. This was due to the
thawing of unprotected permafrost in several firelines, causing silt-laden
meltwater to enter the streams. BLM personnel, who observed this increased
sediment in Washington Creek, estimated the total suspended load to be
several times greater than any we measured. This sediment load had ter-
minated about 10 miles downstream at the time of observation on July 2.
BLM halted most of this sediment movement by building intercepting
water dams across firelines throughout the burned area; they retained
seven bulldozers for this work and treated all firelines between July 3-5.
This measure was evidently effective because the samples gathered on
July 8 show low sediment values from all stations. It was noted that,
following the temporary heavy sediment load reported by BLM, the riffles
-------
TABLE 2
PROPERTIES OF WASHINGTON CREEK WATER
Laboratory Measurements
Suspended
Station
1
2
3
1
2
3
4
1
2
3
4
Date
7/1/71
ii
it
7/8/71
n
ii
n
7/13/71
n
n
n
Time
1600
1715
1845
1145
1300
1500
1610
1300
1340
1520
1430
Sediment
mg/1
M>
-
-
2
3
9
5
68
97
23
34
NH
mg/1
0.03
0.03
0.03
0.03
0.03
0.04
0.03
0.07
0.18
0.06
0.23
NO
mg/1
0.02
<0.01
0.01
0.01
0.03
0.02
0.02
0.09
0.05
0.15
0.03
NO
mg/1
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.004
0.004
0.004
0.005
T-PO
mg/1
0.048
0.055
0.021
0.034
0.034
0.021
0.021
0.07
0.06
0.04
0.05
0-PO
mg/1
0.014
0.013
0.013
0.011
0.012
0.013
0.012
0.010
0.012
0.007
0.012
Q
CFS*
28.2
28.7
13.8
33.7
36.4
12.7
9.2
_
-
-
-
*To convert CFS to m3/s, multiply by 0.028.
8
-------
at stations 1 and 2 were running clear and had not collected visible silt
when visited on July 8. Figures 2-5 illustrate some of the phenomena
that were observed during the sampling trips.
Sediment samples collected on July 13 contained a heavier load of
silt than previous samples at all stations. Both stations below the fire
had nearly three times the sediment load of the two upper stations,
probably reflecting the influence of silt contributed by bulldozed fire-
lines. On July 13 Washington Creek was estimated to be discharging two
to three times greater volume than at the previous sampling dates, which
may explain the higher sediment concentration for all stations. However,
the greatest sediment load below the fire (97 rng/1) was only slightly
greater than the 80 mg/1 level which permits the continuance of good
freshwater fishery (as established by the European Inland Fisheries Advisory
Commission, FAO UN, 1965).
Based on field observations and sediment samples, it must be concluded
that the measures employed by BLM were fairly effective in preventing
extensive sedimentation of Washington Creek. However, this preventative
plan was a "one shot deal" and future protective measures should include
followup work, especially at critical areas where firelines cross streams.
Improved construction procedures should include supervising individual
bulldozing operations, compacting each dam as it is built, and followup
work to detect and repair leaks. Figures 6-7 illustrate the type of water
bars used.
A major objective of the sampling program was to detect and measure
the amounts of nitrogen and phosphorous compounds carried in runoff waters
to the streams from about 60,000 gallons of chemical retardant used.
-------
Figure 2. Washington Creek at Station No. 1. This site is at the downstream bulldozed fireline. When this
photo was taken the water was clear. Note the opposing slopes at the crossing and diversion dams on each
side. The dam on the south bank (upper right) is plainly visible; the one on the north bank is partially
obscured by spruce trees at the mid-left of the photo. July 8, 1971.
-------
Figure 3. Bulldozed fireline on the south bank of Washington Creek at Station No. 1. The tracks shown
here are 1 to 1V4 feet deep and the bottom is ice; the soil is silt, some of which is forming a small delta
shown in this scene. A small water diversion dam had been hand built at the brow of this slope to prevent
surface water from running down these tracks; however, a leak on the bottom was contributing water which
ran down the track and formed the delta; followup work at this dam would have detected and stopped this
leak which later enlarged by thermal erosion. July 1, 1971.
-------
Figure 4. North bank of fireline at Station No. 1. This bank is gravel and is very stable with no evidence of
erosion or melting. The material is probably unfrozen since it is a small slipoff slope formed by the creek at
this bend; hence is very stable compared to the undercut bank of frozen silt on the south bank. Perhaps a
better way to cross streams is to go up or downstream to make use of slipoff slopes on both banks. July 1,
1971.
-------
^at* '::*£
^WJfer
Figure 5. Thermal erosion on a slope exposed to insolation when vegetal cover was removed by fire fighting
activities. Such areas may be unstable for several years and illustrates one of the environmental problems
posed by permafrost. None of the water shown in this scene is from rainfall; all is meltwater from thawed
permafrost. In the fire area covered by this burn, such examples as shown here were few but might have
been avoided by proper siting of fire lines to take advantage of rocky ridges and shallow soils. July 6, 1971.
-------
Figure 6. The upper side of the detention dam on the north bank of fireline. Note the ponded meltwater
retained by the structure. Standing water, such as shown here, retards melting which is accelerated when
released water is free to run off. July 8, 1971.
-------
Figure 7. A bulldozed fireline on fairly level terrain after diversion dams were installed. These dams are
angled across the fireline in an effort to divert the meltwater into undisturbed forest. Bureau of Land
Management has plans to seed and fertilize these firelines in 1972, to establish a plant cover as a temporary
measure before native plants take over. July 6, 1971.
-------
Data presented in Table 2 do not indicate significant increase in these
materials over that found in waters at the control station. Although
there does appear to be a slight increase in total phosphate in samples
below the fire, compared to samples above the fire, none contained as much
as the upper limit of 0.100 mg/1 recommended by the Committee on Water
Quality Criteria (1968). Nitrogen concentrations do not appear to have
been affected. Samples collected on July 13 should reflect any effects
of runoff because sufficient rain had fallen to cause Washington Creek to
rise; hence, the water should have contained chemicals from retardants
if they were carried to drainage waters by surface runoff.
Discharge measured below the fire (Station 1) on July 1 was about 28
cfs, and about 14 cfs at the bridge above most firelines (Station 3).
Discharge at the bridge had decreased to about 13 cfs on July 8, whereas,
below the fire, discharge had increased to about 34 cfs. Evidently,
meltwater from thawing firelines was entering streams influenced by the
fire, yet no increase in sediment load was measured on that day. These
discharge data confirm that considerable water was running off the fire
area before rains caused Washington Creek to rise. If retardant chemicals
were to have seriously affected the drainage waters, they should have been
present in samples taken during the period of increased discharge.
16
-------
Summary
Water samples collected at the Wickersham Dome fire in July 1971 were
analyzed at the AWL to obtain a measure of sediments caused by the fire
and activities such as bulldozed firelines directly crossing streams,
thawing permafrost on firelines and traffic from ground supply vehicles,
and to assess the effect of chemical retardants on waters draining the
burned over area. Initially, a considerable sediment load, estimated
to be about 300-500 mg/1, was observed by BLM personnel. Corrective
measures taken by BLM immediately following the fire consisted of
building diversion dams across firelines. These appeared to be
effective and samples later collected were low in suspended sediment;
the highest sediment in samples collected on July 8 was 19 mg/1. Some
sediment from bulldozed firelines was apparently entering the drainage
system after sufficient rain had fallen to cause Washington Creek to rise,
for the highest quantity measured was 97 mg/1.
Samples collected below the burned area did not show an increase in
nitrogen attributable to chemical retardant use compared to those from
the control site. Total phosphate in samples collected below the fire
was somewhat higher than in the control area but never exceeded the minimum
content recommended for aquatic life in Water Quality Criteria (Report of
the National Technical Advisory Committee, FWPCA, 1968, pp. 27-110).
Interception dams constructed by BLM to divert meltwater from exposed
permafrost to vegetated areas were largely effective. However, on future
fires, improvements in supervision and techniques should be made and
followup work done at critical areas where firelines cross flowing streams.
17
-------
|