-------
21
98!
JFMAMJJASO
1982
DlJFMAMJJASOND
0
-s_
.
— ,
1
SO
O
§L
z*
i
X
"Z.
— .
o»
^2_
2!
CO
H
200
150
100
50
0
40
32
24
IS
3
0
450
360
270
ISO
90
0
j J J i ! ! 1 ! 1 1 1 | 1 1 1 1 ! ' ' ' ' ' '
-
•
-••• *.» . -
• •
O
* •• • •
u * - -
. • • • • .
* •» • • •
•
. * • • • •
*o . . • • •
~ t 1 1 1 1 1 1 1 1 ! 1 1 1 1 1 ! 1 1 1 . L 1 _. 1. ! ~
-
«
.
»* • • *
« * « * • •
— * * • • —
• * * • * • *
9 • * • • •
• • * • *
1 1 1 1 t J ! 1 1 1 1 I ! ! t 1 1 1 ! 1 1 ' ! !
* •
A A
855 - 825
• *
* •
— —4
* *
*
*
I
~l
•
9 •
* •
• * *
• « *
* •
• • • ** **
— *» » » * • • —
• * *
4
1 1 1 i ! t | | | | i ] i | ] | | , , , , i
JFMAMJJASONDJFMAMJJASOND
1981 1982
Figure A4a. Values of selected variables for the Snake River as it
enters Lake Dillon.
-------
1981 1982
JFMAMJJASONDlJFMAMJJASOND
—
0
flfc
Q.
25
"o
CO
"o
o»
w
Q.
O
Q_
X.
o>
E
tf>
.£
"o
3
O
"•£
o
CL
10
8
6
4
2
0
10
5
0
8
6
4
2
— i — i — 1 — i — i — I — i — i — i — i 1 — 1 — i — i ! — i — m — nr-T — ! — IT-: 1
A A A
13 25 18
-
•
• '. "
. " -
• * • * * *."*
- • •• • . -
• •• • • • «
i
1 1 1 1 1 ! 1 1 1 1 1 1 1 1 ! I 1 1 1 1 • i i
A
20'
'
•
— • -H
• • • • *
* •
• • • •
• • # • • • •
• ••**• •
i
1 1 1 1 1 1 1 1 1 1 I I ! 1 1 ! i 1 ! ! ' ! !
A A A
19 . ' 13 15
™ """
- • —
• • *
• * •• • •
* •
• • • • |
* 1 •
I 1 ! 1 I 1 1 I I I 1 I !' I '!'!!! 1 .
JFMAMJJASONDJFMAMJJASOND
198! 1982
b. Values of selected variables for rhe Snake River as it
enters Lake Dillon.
-------
216
broader in 1982 and the highest values did not occur until the end of
June. In both years there was a small early peak in May, probably due
to early melt in the valley bottoms, and a shoulder or low peak in
August or September. The amount of variation around the seasonal
pattern was small, which provides an ideal situation for the estimation
of weighted mean concentrations and transport.
Stream temperatures varied between 0 and 16°C. The highest
temperatures occurred during the last half of July. Maximum
temperatures were lower in 1982 than in 1981 by 2-4°C. The seasonal
change in temperatures was relatively smooth, although some
irregularities were caused by variations in time of day when
temperature was taken. Variation of temperature with season and
between years affected the depth of penetration of water entering the
lake, as already explained.
Conductance showed the dilution effects that one would expect to
be associated with changes in discharge for materials that are soluble
and little affected by biological activity. For the Snake River, the
conductance faithfully reflected the early small peak in discharge
(May), the subsequent brief decline in discharge, and the major peak of
discharge caused by spring runoff- This sequence of events was visible
both years for the Snake River and was also characteristic of the other
two rivers. However, the winter conductances of Tenmile Creek were
especially high because of the presence of especially potent sources of
soluble materials in the upper part of this watershed, at Climax
Molybdenum tailings ponds.
-------
Nitrate concentrations of the Snake River were high at the time of
runoff and then decreased steadily until August, after which they
increased steadily until December or January. Midwinter concentrations
were typically high. The low nitrate concentrations in the late sunnier
months were almost certainly due to high summer biological demand for
inorganic nitrogen, which became all the more effective in reducing
nitrate levels as the amount of water movement declined at the end of
summer. With the onset of cool weather, and reduced biological demand,
nitrate concentrations again rose to higher winter levels. The winter
nitrate levels of the Snake were somewhat higher than would be expected
for unaltered watersheds in this region (cf. Lewis and Grant 1979,
1980b). They probably reflect the influence of dispersed nutrient
sources, especially from residential areas, which have an especially
noticeable influence on winter concentrations because of the very lov
discharges in winter and the low biological demand for nitrogen at this
time. The seasonal pattern of nitrate concentrations observed in the
Snake River is, however, essentially what would be observed in an
unaltered watershed (Lewis and Grant 1980b). The same pattern is
visible in the data for Blue River and Tenmile Creek, but with
considerable scatter due to the influence of point sources. The late
summer minimum and winter maximum are evident, however.
Tenmile Creek was unusual in that nitrate concentrations were
higher at all times of the year than in other rivers. In 1981 nitrate
nitrogen approached 1,000 ug/1 in winter (Figure 45). The high nitrate
concentrations are explained by the large amounts of nitrate exported
from the site of Climax Molybdenum at the head of Tenaile Creek. This
-------
218
1000
800
600
400
200
0
Z
I
fO
o
—9
1 ! 1 1 \ ! ! I 1 ' 1 1 ! 1 1 \ i 3 '- 1 1
JFMAMJJASONDJFMAMJJASOND
1981 1982
Figure 45. Nitrate values for Tenmile Creek at its point of entry to
Lake Dillon, illustrating the very high nitrate levels at
certain times of the year.
-------
O 1 Q
special nitrogen source will be considered further in the
segment-fay-segment analysis of Tenmile Creek.
For ammonium, there vas also a seasonal pattern of concentrations
in 1981 and 1982, but it was not so clear as the pattern for nitrate.
In part this was due to the much lower concentrations of ammonium than
of nitrate, especially in the Snake River. The Snake River data do
show indications of an early spring maximum in concentration before
discharge had begun to increase drastically, a late summer minimum
corresponding to the postulated peak of biological demand for inorganic
nitrogen, and a fall increase after the stream water had begun to cool.
Two high values in early August of 1981 for the Snake River departed
from the seasonal pattern and were not repeated in 1982. These high
values were confirmed, however, by other samples upstream and were
almost certainly caused by some temporary or sporadic man-iaade source.
Patterns in the ammonium data for the Blue River and for Tenmile
Creek are scarcely discernible because of the large amount of
variability introduced by point sources. Ammonium concentrations were
affected simultaneously not only by seasonal factors but by variations
in release from point sources and variations in the conversion rate
from ammonium to nitrate.
In all three rivers, total soluble nitrogan was accounted for
approximately half by organic nitrogen and half by ammonium plus
nitrate. Winter values of total soluble N tended to be higher than
summer values, in parallel with nitrate, a major contributor to to~a~
soluble X. Seasonal pattern was weak, however, especially for the Blue
River and Tenmile Creek.
-------
220
Total soluble phosphorus concentrations showed no strong seasonal
pattern in the Snake River, nor did orthophosphate concentrations. The
average concentrations were often high at times of high discharge,
however. The implication is that dilution effects associated with
major seasonal increases in discharge were effectively cancelled by
increased transport at such times (cf. Lewis and Grant 1979). There
were no clear seasonal patterns in concentration for the Blue River or
Tenmile Creek, where the point sources introduced considerable
additional random variation. For the Blue River, the increase of
orthophosphate and total soluble phosphorus in September, October, and
November of 1981 caused by treatment plant malfunctions has already
been mentioned. A similar but less exaggerated phenomenon was observed
in Tenmile Creek in the first half of 1981, but was not repeated in
1982 and was of obscure origin.
Total particulates, particulate phosphorus, particulate nitrogen,
and particulate carbon are united by a tendency to show patterns
similar to each other. In the Snake River there was a trend toward
high concentrations around the time of maximum discharge. Superimposed
on this was an occasional exceptionally high value associated with some
kind of short-lived earth disturbance in the watershed, as shown for
the Snake River in late August of 1981 and in the summer of 1982. Such
short-term spikes obviously can be extremely important in total annual
transport of particulate materials.
-------
79"
Chemistry of Miner's Creek and Soda Creek
Because of the relatively small proportion of total land area
drained by Miner's Creek and Soda Creek, the potential contribution of
these two watersheds to total loading of the lake with nutrients is
small. Their chemistry is considered here mainly for the sake of
completeness, since these two watersheds are separate from the three
major river drainages. Table 41 gives the discharge-weighted averages
for the chemistry of the two streams and the standard error for each
variable. As in the case of the three rivers, the time-weighted means
were calculated but are not shown in the table. For Miner's Creek and
Soda Creek, the time-weighted means are even closer to the
discharge-weighted means than for the three major rivers.
The small size of contributions of these two creeks to total
loading of the lake is evident from the discharge figures in Table 41.
The chemical data are unexceptional. Soluble nitrogen concentrations
were in the range of those observed in the Snake River, and
considerably below those of the Blue River and Tenmile Creek as
reported in Table 40. This is not unexpected, since Miner's Creek and
Soda Creek at the sampling points were not influenced by point sources
of nutrients, but were influenced by some relatively dispersed human
influences analogous to those characteristic of the Snake River
drainage.
For Miner's Creek, the phosphorus concentrations, including both
soluble and particulate fractions, were relatively low, and were
comparable to the values observed in the Snake River. For Soda Creek,
however, both particulate and total soluble phosphorus concentrations
-------
N02-N, up,/!
fJO^-M. uc/1
v O > « '
Nil * — N uc/j-
Total Soluble N, ug/1
Participate N, ug/1
PO.-P. up/1
4
Total Soluble P, ug/1
Particulate P, ug/1
Total Pavticulotes , mg/1
Particulate Carbon, mg/1
Alkalinity, mg/1
pll
Conductance, uinho/cm
Discharge » I /sec
Discharge, cfs
Mi n e r ' s
Mean
0.8
32.2
14.9
141.
34.
1.9
3.2
6.4
6.9
320.
28.9
7.4
74.
88.
3.1
1981
Creek
S.E.
.11
4.3
3.0
34.
13.
0.23
0.34
1.2
1.9
70.0
1.7
0.08
3.4
10.
0.37
Soda Creek
Mean S.E.
1.2
12.7
12.2
90.0
56.
8.7
18.3
11.0
6.1
351.
59.0
8.0
168.
12.
0.42
.32
4.5
2.8
21.
17.
1.2
3.3
1.5
1.2
63.2
3.2
0.15
9.8
3.1
0.11
Miner's
Mean
1.4
20.2
10.0
102.
83.
2.9
5.1
5.8
7.7
531.
26.0
7.5
58.
246.
8.7
1982
Creek
S.E.
.20
4.1
1.9
26.
16.
0.47
0.60
0.76
1.7
116
1.2
0.05
2.6
47.
1.7
Soda Creek
Mean S.E.
2.3
63.
11.5
202.
100.
13.4
20.5
21.6
16.3
501.
62.3
8.0
142
54.
1.9
.23
14.
1.4
41.
29.
2.7
3.0
4.0
3.1
103.
2.4
0.08
8.5
15.
0.52
Table 41. Discharge-weighted means for the two amall streams entering the ]ake (sampling stations above WWTP
cf II iicnts) .
-------
123
were higher than for any of the three major rivers, and the discharge
was especially low in relation to drainage area. Conductance was also
notably higher for Soda Creek than for Miner's Creek. Water is
evidently used within the drainage, and this results in reduced
discharge. Water use and possibly spri-gs may account for high
conductance.
Frisco and Snake River Treatment Plant Effluents
Table 42 summarizes the chemistry of the effluents from the Frisco
and Snake River wastewater treatment plants. Total soluble nitrogen
concentrations were 50 to 100 times higher than the stream
concentrations and were relatively similar between plants and between
years. The phosphorus values present quite a different picture.
First, the effluents of the two plants were rather different in the
partitioning of total phosphorus. In the Snake River Wastewater
Treatment Plant effluent, the largest amount of phosphorus was
connected with the particulate fraction, while in the Frisco effluent
the largest fraction was connected with the soluble components. For
the Snake River Plant, the phosphorus concentrations were about 10
times above those of the receiving stream. A few exceptionally high
concentrations were checked against plant records. High values
typically coincided with construction or equipment problems. The
average total P concentrations for the Frisco Plant were considerably
higher than for the Snake River Plant, and there was a rajor difference
between 1981 and 1982 for the Frisco Plant. In 1982 the
discharge-weighted averages were exceptionally high at the Frisco
Plant. Examination of the raw data shows that the high averages were
-------
1981
Frisco WWTP
Effluent
NG^-N, ug/1
N03-N, ug/1
Nll^-M, ug/1
Total Soluble N, ug/1
Particuiate N, ug/1
P04-P, ug/1
Total Soluble P, ug/1
Particuiate P, ug/1
Total Particulates , mg/1
Particuiate Carbon, mg/1
Alkalinity, rag/1
pll
Conductance, uniho/cm
Discharge, 1/s
Discharge, cfs
Mean
135.
1579.
2787.
8911.
167.
53.6
76.7
17.6
1.3
741.
42.2
6.8
417
12.8
0.5
S.E.
15.8
407.
670.
1168.
23.
18.
19.
2.2
0.28
163.
2.9
0.08
29.
0.89
0.02
Snake WWTP
Effluent
Mean
116
3527.
1825.
6798.
496.
4.5
22.
45.9
6.0
4710.
50.5
6.5
450.
13.4
0.5
S.E.
17.
786.
620.
774.
160.
2.4
3.4
11.2
2.4
2384.
7.9
0.16
18.
0.92
0.04
]982
Frisco WWTP
Effluent
Mean
113.
1228.
2868.
7096.
-
496.
556.
57.0
5.1
-
65.8
7.2
449.
14.5
0.5
S.E.
25.
281.
543.
867.
-
169.
177.
16.
1.2
-
7.7
0.10
21.
0.50
0.02
Snake WWTP
Effluent
Mean
268.
2251.
1957.
8398.
-
24.4
38.6
120.
9.3
-
68.5
7.0
429.
14.6
0.5
S.E.
50.
910.
453.
1627.
-
12.
13.
32.
2.
-
6.
0.
20.
1.
0.
2
9
09
2
04
K>
Table 41'.. Dlscharge-wc I ghl ed inemir. for the Lwo effluents discharging near the mouths of Miner's and
r>od.i Crt-oks .
-------
225
caused by discontinuous but repeated occurrence of very high
concentrations (1000 to 3500 ugP/1). Checks of the dates of high
values with plant operators showed that, for the most part, the dates
of high P values coincided with known equipment problems.
*
Precipitation Chemistry at the Main Station
The precipitation chemistry data consist of the continuous record
from the collector situated near the Snake River Wastewater Treatment
Plant and shorter records from two other locations. We will consider
the more extensive record first and then compare it with the other two
records.
All of the precipitation chemistry data to be dealt with here
refer to bulk precipitation, which is the total of all wet and dry
deposition on a collecting surface. A collecting surface accumulates
dry deposition when there is no rain or snow. The occurrence of wet
precipitation augments the delivery of materials onto the collector
surface, especially for certain substances that are scrubbed out of the
atmosphere effectively by water droplets (e.g., nitrate). There ara
typically changes in the concentration of dissolved materials in wet
precipitation during the course of a rainstorm, principally as a result
of the cleansing of the air by rain. Because of the continuous
deposition of materials on a collector surface during dry weather and
the continuous change of concentrations during a wet precipitation
event, an average concentration is not very meaningful for
precipitation chemistry. For this reason, we express all the results
-------
for precipitation chemistry in terms of loading rates, i.e., the amount
of material delivered to a unit surface over a specified time.
Table 43 summarizes the means and standard errors for the loading
of chemical constituents analyzed in the samples from the main
precipitation collector at the Snake River Wastewater Treatment Plant.
The means are derived from 61 separate collections spread over the
two-year study period. Since the collection intervals were not all of
identical length, the means are time-weighted averages.
The hydrogen-ion loading was lower (4x) than loading reported for
the Como Creek watershed at similar elevation near Ward, Colorado
(Grant and Lewis 1982), but precipitation was frequently below pH 5.65,
indicating the presence of strong mineral acids. Because of the
dominant influence of stream chemistry on the chemistry of a lake such
as Dillon, however, direct hydrogen ion loading from precipitation is
of minimal significance to the biology of the lake. As expected,
nitrite was a minor contributor among the inorganic soluble nitrogen
species, but nitrate nitrogen was present in substantial amounts. As
at Como Creek, which has been studied thoroughly in this regard, the
precipitation at Dillon appears to be substantially augmented beyond
background in nitric acid as a result of the presence of substantial
amounts of NGv in the atmosphere from fossil fuel combustion. The
ammonium loading was relatively high but secondary to the nitrate
loading as a nitrogen source.
The loading rates for both PC^-p and total soluble phosphorus
*ere considerably higher than at Como Creek (3-5x). Sinca the
mechanisms of phosphorus transport through the air are poorly known, no
-------
1981. - iag/m2/wk
N02-N
*
A
Total Soluble N
POA-P
Total Soluble P
Total Particulates
Particulate P
HCO~
H+
so4
Mean
0.079
2.616
3.479
18.380
0.732
1.012
187.
0.219
19.8
0.055
12.8
S.E.
0.008
0.284
0.353
1.776
0.087
0.143
8.01
0.015
8.60
0.006
0.86
1982 - ng/m2/wk
Mean
0.125
3.724
1.627
6.002
0.630
0.695
152.
0.178
23.4
0.040
46.7
S.E.
0.011
0.351
0.098
0.364
0.123
0.134
7.54
O.OC9
1.56
0.006
5. 7
Table 43. Loading rates for bulk precipitation at the main collecting
station, 1981 and 1982.
-------
228
explanation can he offered. However, the observed phosphorus transport
values are within the range of values reported for a variety of sites
in the literature (Likens et al. 1977). Particulate phosphorus
transport was significantly lower than at Como Creek, and this
partially offset the higher loading for soluble fractions. Sulfate
was a major ion, as would be expected by analogy with the Como Creek
studies.
Figure 46 shows tine plots of loading rates for some of the
nutrient fractions that have a significant amount of seasonal pattern.
The transport of both nitrate and ammonium is facilitated by moisture,
as is evident from a comparison of the amount of rain or snow and the
loading rates for these ions. Total soluble phosphorus showed a
pronounced maximum in June of both years. This early summer maximum,
although quite high, appeared both years and at more than one station
and was thus most definitely not an error.
Precipitation Chemistry at Other Stations
Phosphorus loading rates by bulk precipitation were determined at
two additional stations in 1982 as a comparison with the main station
located near the Snake River Wastewater Treatment Plant. The first of
the comparison stations was located near Frisco. A different type of
collector (tube collector, diameter 25 cm) was used at this station;
the dates of comparison were June 1 to November 17, 1982. The means
r.r.d standard errors for the main station and the Frisco station over
this interval are summarized in Table 44. There was no significant
difference in the average loading rates for the two stations (p >
-------
Jtf
-v
E
E
E
„
z
'•0
O
Z
^
$
V.
w
E
o>
E
„
Z
1
X
JJ.
^
M
E
0
CL
^J
^^^
3
"o
"o
K
•t u
32
24
16
8
0
8
6
4
2
0
10
8
6
4
2
0
3
2
1
.
*
* •
4
— • __
*
• • 4
•*••*.. * * **
* • • • • •• ••• •
- -1 *i "i *i . ". i 1 . <• i P 1 •, , , • ,« , , , , . , -
A35 24ii.
30
*
— — '
• • 1
- * • u
* * • • • •
• *
* • • —
* • ••• ••• •
"" 1 *1 1 1 1 t 1 1 1 1* 1 1 1 1 1 1 1 1 l 1 • • ;
• • *
14AAA26
14
-
_ * —
_ • _
• •
* ... . .* * ~l
• ••» • •*•
" *| * 1 1 I 1 1 I 1*1 1 ! 1 1 !*! ! 1 !*! ' * ! ! ' ~*
1
i
•
-
• • •
I «| * ! *l ! ! ! » <••!•*! • ! • 1 ' j • 1 I ! ! «*• . :
JFMAMJJASONDJFMAMJJASONO
1981 1982
Figure -6. Selected variables for bulk precipitation rear the Snake
River WWTP.
-------
130
0.05). The period of collection incorporated the high phosphorus
loading pulse in June, which results in relatively high standard errors
for both stations. However, the pulse appeared simultaneously at the
two stations, verifying our conclusion that it was not an artifact.
1
1
Dates of Comparison
June - 17 November
July - 1 November
Total P
Main Station
mg/m /wk-
Mean S.E.
1.354 .969
.427 .065
Comparison
Station
Frisco
Raft
Total P
Comparison Station
9
mg/m*"/wk
Mean S.E.
1.561 .688
.533 0.099
Table 44. Summary of comparison for phosphorus loading by bulk
precipitation at the main station and two other stations.
The second comparison station was situated on the Denver Water
Department's raft, which was anchored on the water surface. This
comparison station is of special interest because we wanted to rule out
the possibility that a significant amount of phosphorus transport
observed at the Snake River station could be accounted for by localized
dust in that area. The dates of comparison were July 1 through
November 1. From July 1 through mid-September the raft was anchored in
midlake, and from mid-September through October it was anchored at the
shoreline. The means for the main station and the station on the raft
ever these dates are summarized in Table 44 along with their standard
errors. The standard errors are much lower in this case because the
interval of sampling did not include the period of high phosphorus
-------
"1
loading in June. There was no significant difference in the loading
rates between the two stations.
From the data at the two comparison stations, we have every reason
tc believe that the main precipitation collection station at the Snake
River Wastewater Treatment Plant gave an accurate impression of the
bulk precipitation loading to the surface of Lake Dillon.
-------
232
Total Nutrient Loading of the Lake
Water Budget
Reliable estimations of total nutrient loading of the lake over
the period of study presume the availability of sound information on
the amounts of water reaching the lake by various pathways. If the
flow measurements summarized in the previous chapter are accurate,
their summation plus the unmeasured flows should be close to the
calculated input determined by the Denver Water Department, which is
obtained by an indirect method based on measured outflow and change in
lake volume. Table 45 summarizes the flows from various sources in
1981 and 1982. The sources of the estimates for the rivers, streams
and effluents are as given in the previous chapter. Two sources of
water were estimated without direct measurement: miscellaneous surface
runoff and groundwater.
Miscellaneous surface runoff was the contribution from a number of
small areas near the lake that did not drain through watersheds where
there was gauging or discharge measurement. These areas total 4300 ha,
or 5.8% of the total land area of the Lake Dillon watershed. Their
unmeasured contribution to runoff was estimated in two parts. The
first of these is the Meadow Creek drainage, which includes some
high-elevation areas. This drainage was considered to be
hydrologically similar to Miner's Creek. The runoff was thus set at an
r.~cu"t .-r.qual to that of Miner's Creek, with appropriate adjustment
tor area. The second portion consists of smaller drainages that are
low-lying and without steep contours. Because of low accumulation of
-------
233
Rivers
Snake
Blue
Tenmile
Subtotal
Strearas not on Rivers
Miner's
Soda
Subtotal
Effluents
Frisco
Snake
Subtotal
Diffuse Surface Drainage
Precipitation
Groundwater
Grand Total
1981 -
m /yr
40,100
49,900
58,000
148,000
2770
330
3150
403
423
826
3366
4336
8^41
168,119
Thousands
Acre-ft/yr
32.54
40.45
47.02
120.0
2.25
0.31
2.55
0.33
0.33
0.66
2.73
3.52
6.90
136.4
1982 - Thousands
m /yr Acre-ft/yr
81,800 66.35
98,200 79.61
104,000 84.31
284,000 230.3
7760 6.29
1700 1.38
9460 7.67
457 0.57
460 0.37
917 0.7i
10,385 8. -2
4057 3.29
84-1 6.90
317,260 257.-
Table 45. Summary of water flow into the lake in 1981 and 19?2
-------
234
snowpack in such areas, the water yield was expected to be small. This
second portion was assumed to have the water yield of Soda Creek
drainage, with appropriate adjustment for area. Any errors in the
approximation procedure will obviously not have a very great effect on
the water budget of the lake because of the relatively small proportion
of the watershed in the miscellaneous surface drainage category.
Groundwater also potentially supplies water, and thus nutrients,
to the lake. A complete water budget was constructed on a monthly
basis for 1976-1979 from the outflow figures of the Denver Water
Department, the inflow data of U.S.G.S gauges, measures of
precipitation, estimates of evaporation, and an allowance for
miscellaneous surface drainage. This water budget should sum to zero
if all estimates were made with good accuracy and if there was no
groundwater entry into the lake. Some variance around the zero sum is
expected due to inaccuracies of the various estimates, but a trend
toward greater outflow than inflow would suggest the presence of an
unmeasured term, which would have to be groundwater. The residuals
over the three years had an average of 575 acre-feet per month. There
was no strong seasonal or annual pattern in the size of the residual,
and the standard error was about 300 acre-feet per month. This
residual may have been affected by the amount of drawdown as well as
other factors. Since the irregularities in the residual cannot be
pinned to a specific pattern, we treat the groundwater inflow to the
lake as constant at 575 acre-feet per month for both years of the
s:ady. Since this amount of flow is very small by comparison with the
total flow to the lake, any errors in the estimate within the range of
-------
235
one or two standard errors would not be significant to the overall
conclusions of the study.
When all sources of water for the lake were sunned (Table 45), the
estimated inflow was 136,000 acre-feet for 1981 and 257,000 acre-feet
in 1982. This figure can be compared with the independently derived
estimate of the Denver Water Department for total input over the sane
interval. First, however, evaporation must be added to the Denver
Water Department estimate, which does not include evaporation. From
the averages of several years of study by the Denver Water Department,
we estimate evaporation as 5900 acre-feet per year, which gives a
corrected Denver Water Department input of 111,000 acre-feet per year
for 1981 and 223,000 acre-feet per year for 1982. The deviation of
this from the totals based on U.S.G.S. data as described above is
10-20%, which is more than anticipated. Reexamination of the data did
not reveal any potential errors of this magnitude. For nutrient
loading computations, we assume that the U.S.G.S. data for input are
correct, although this assumption is to some degree arbitrary.
Phosphorus Loading
Table 46 summarizes the phosphorus loading of Lake Dillon
according to the various pathways of nutrient flow that were recognized
in the previous chapter and in the section on water balance. For
rivers, streams, and effluents, the numbers in Table 46 vere octainec
by multiplving the discharge-weighted average concentrations as
reported previously (Tables 40, 41) tines the time-weighted average
discharge and dividing the product by lake area. The estimate of
-------
Phosphorus -
Rivers
Snake
Blue
Term J 1 e
Subtotal
Streams not on Rivers
Miner's
,°3da
Subtotal
Effluents
Frisco
Snake
Subtotal
Miscell aneoua Surface
Drainage
Free ipi tat i on
Ground water
OR A Ml) TOTAL
PO -P
4
.036
.524
.048
.608
.004
.002
.006
.016
.001
.018
.011
.381
.006
1.030
Total
Soluble
P
.060
.524
.156
.740
.007
.005
.012
.023
.007
.030
.019
-520
.010
1.331
- kg/ha/yr
Part. P
.127
.206
.308
.641
.013
.003
.016
.005
.015
.020
.035
.114
.000
.826
- 1981
Total P
.187
.729
.464
1.380
.020
.008
.028
.029
.022
.051
.054
.634
.010
2.157
Phosphorus •
%
8.7
33.8
21.5
64.0
.9
.4
1.3
1.3
1.0
2.4
2.5
29.4
.5
100.1%
PO -P
4
.166
.316
.164
.646
.017
.017
.034
.170
.008
.178
.045
.328
.028
1.259
Total
Soluble
P
.386
.492
.428
1.306
.030
.026
.056
.190
.013
.204
.080
.361
.066
2.073
- kg/ha/yr
Part. P
.423
.411
.459
1.293
.034
.028
.061
.020
.041
.061
.091
.093
.000
1.599
- 1982
Total P
.808
.903
.887
2.598
.064
.054
.117
.210
.054
.265
.171
.454
.066
3.671
%
22.0
24.5
24.1
70.6
1.7
1.5
3.2
5.7
1.5
7.2
4.7
12.4
1.8
99.9%
Table 46. Summary of phosphorus loading of the lake in 1981 and 1982.
-------
237
loading for precipitation was taken directly from Table 43 of the
section on precipitation chemistry. For miscellaneous surface
drainage, the water contribution has already been obtained by
approximation as reported in Table 45. The concentration of phosphorus
fractions in this water must also be approximated before an estimate of
loading can be made. Phosphorus concentrations in the miscellaneous
surface drainage category were assumed equal to those observed in
Miner's Creek because of parallel land usage, and these concentrations
were applied to the discharge approximations made in connection with
Table 45 to produce the estimate of loading shown in Table 46. A
similar approach was taken for groundwater. The groundwater phosphorus
concentrations were assumed to be equal to those of the Snake River
drainage, a large watershed with mixed land use but lacking point
sources. It was also assumed, however, that particulate phosphorus
would not be transported through groundwater and this fraction was
therefore set to zero. The concentrations were then multiplied by
the amount of groundwater as reported in Table 45 to obtain the loading
of phosphorus via groundwater. Both the miscellaneous surface drainage
and groundwater categories were minor contributors, so the details of
these approximations are of no great concern.
More than half of the total phosphorus from all sources entered
the lake in soluble form. Table 46 shows that the bulk of total
soluble phosphorus was in the orthophosphate fraction rather than the
organic phosphorus fraction. The particulate fraction was significant
both years and was slightly higher proportionally in 1982 than in 1°?1,
probably because of higher discharge in 1932. The total phosphorus
-------
233
transport in 1981 was about 60% of the total phosphorus transport in
1982.
Trie rivers contributed about two-thirds of the total phosphorus
reaching the lake. The individual rivers made very different relative
contributions in the two years, however. All were affected by higher
discharge in 1932, but not identically. Loading from the Snake River
was drastically affected by construction in the river bottom in the
second year. Loading from Tenmile Creek was approximately proportional
to the difference in flows between the two years. The Blue River
loading was exaggerated in 1981 by problems at the Breckenridge WWTP.
Other differences in sensitivity cf the three rivers to changes in the
hydrologic conditions between years can be traced to their differing
land use patterns; these will be considered further when land use is
analyzed.
Precipitation accounted for a large proportion of the residual
loading beyond the contributions of the three major rivers. The
proportion of total phosphorus loading attributable to precipitation in
1981 was considerably larger than in 1982. This is principally
explained by the much greater contributions from surface runoff in
1932, which reduced the overall influence of precipitation loading.
The combined contribution of the Frisco and Snake River wastewater
treatment plants was only 2.4% in 1981 and 7.2% in 1982. The
contribution in 1982 should actually have been lower in view of the
much higher phosphorus loading from non-point sources that year, but
tr..'accent plant malfunctions produced an exceptionally high effluent
contribution or, certain dates in 1982. There are two other major
-------
239
wastewater treatment plants in the watershed: Breckenridge and Copper
Mountain. The contributions of all four plants will be specified more
exactly when the nutrient loadings are broken down more completely
according to sources and land uses. It is obvious from Table 46 that
the bulk of phosphorus entering Lake Dillon does not originate from
point sources, however.
Nitrogen Loading
Table 47 summarizes the Dillon nitrogen loading. The methods of
computation were similar to those used for phosphorus. Particulate
nitrogen was not measured in the effluents in 1982 because it was such
a small contributor to effluent N. The particulate N column in the
table was filled in for 1982 by use of the 1981 data.
The inorganic fraction of nitrogen was slightly larger than the
organic fraction for all sources combined. Unlike phosphorus, nitrogen
loading was mainly associated with the soluble fractions; particulate
nitrogen made a relatively small contribution. The difference between
years in total nitrogen loading was very similar on a proportional
basis to that observed for phosphorus: the loading for 1982 was al^osz
double that of 1981. For nitrogen, the dominating effect of the three
rivers on total loading was even more pronounced than it was for
phosphorus. The influence of precipitation on nitrogen loading was
relatively small by comparison with phosphorus. The contributions of
the two effluents on a percentage basis were slightly larger than for
phosphorus, but still relatively minor by comparison with total
loading (Figure 47).
-------
Rivers
Snake
Blue
Tenmlle
Subtotal
Streams
Miner's
Soda
Subtotal
Effluents
Fr laco
Snake
Subtotal
Miscellaneous Surface
Drainage*
Tec i pi tat Ion
Iround water*
CKAND TOTAL
NO -N
.02
.37
.15
.55
.002
.00
.002
.04
.04
.08
.005
.004
.002
.64
„,-.
2.86
8.49
9.20
20.55
.07
.004
.07
.48
1.12
1.59
.18
1.36
.23
23.98
Nl trogen
»v»
.33
6.99
1.1 3
8.45
.03
.003
.03
.84
.56
1.40
.08
1 .81
.03
11 .79
- kg/ha /y
Total
Soluble
N
5.49
20.61
15.49
41.59
.29
.03
.32
2.69
2.15
4.84
.79
9.56
.44
57 . 54
r - 1981
Part.
N
.99
1.05
2.86
4.90
.07
.02
.09
.05
.10
.21
.19
1.17*
.000
6.56
Totiil N
6.49
21.65
18.36
46.50
.36
.04
.41
2.74
2.31
5.05
.99
10.72
.44
64.11
'
10.1
33.8
28.6
72.5
0.6
0.1
0.7
4.3
3.6
7.9
1.5
16.7
.7
100.0%
"V
.09
.24
.86
1.17
.008
.003
.01
.04
.09
.13
.02
.07
.01
1.41
t>
MVN
7.16
7.98
13.81
28.95
.12
.08
.20
.42
.78
1.20
.32
1.140
.58
32.39
li trogen -
Nil, -N
4
.73
7.86
7.32
15.91
.06
.02
.07
.98
.67
1.66
.16
.85
.06
18.71
- kg/ha/yr
Total
Soluble
N
14.33
28.06
37.05
79.44
.59
.26
.85
2.43
2.89
5.32
1.59
3.12
1.16
91.48
- 1982
Part.
N
6.55
5.36
6.93
18.84
.48
.13
.61
.05*
.16*
.21*
1.30
.95*
.000
21.91
Total N
20.88
33.42
43.98
98.28
1.07
.38
1.46
2.48
3.05
5.53
2.&9
4.07
1.16
113.39
*
18.4
29.5
38.8
06.7
.9
0.3
1.2
2.2
2.7
4.9
2.5
3.6
1.0
99.9%
estimated
}ble 47. Summary of nitrogen lending of the lalce In 1981 and 1982.
-------
All other sources
Frisco & Snake
effluent
Snake R
14%
Precipitation
21%
Blue R.
30%
Tennmle
23%
241
P
loa
All other sources
Frisco S Snake
effluent
Precipitation
Snake R
13%
Blue R.
32%
Tenmile
34%
load
Figure £7. Itenization of sources of ? and X loading for lake Zil-.cn
at the points of entry to the lake (average 1931. 19821).
-------
Overview of Phosphorus and Nitrogen Loading
Table 48 expresses the phosphorus and nitrogen loading in several
different ways. Because of phosphorus removal at the point sources and
lew background phosphorus in the watersheds, the ratio of nitrogen to
phosphorus in the loading was very high. The tnolar ratios of nitrogen
to phosphorus were 69 in 1981 and 66 in 1982.
1981
N
P
1982
N
P
Per unit
kg/ha/yr
85.3
2.9
134.0
4.2
lake area*
g/m /yr
8.5
0.29
13.4
0.42
Entire
kg/yr
85,600
2,900
151,400
4,800
lake
Ib/yr
189,000
6,300
333,800
10,700
*Based on mean lake areas of 1004 ha in 1981 and 1130 ha in 1982.
Table 48. Summary of N and P loading.
Figure 48 puts the phosphorus loadings into perspective by use of
the widely-used Vollenweider diagram (Vollenweider 1968), which
predicts the trophic status of a lake on the basis of its mean depth
and its phosphorus loading. The position of Lake Dillon on the diagras
is shown with relation to the position of a small selection of other
lakes. The diagram is presently considered to provide only a crude
approximation of trophic status because it does not allow the
phosphorus loading to be discounted for varying degrees of phosphorus
_-edir.entation and loss through outflow. However, the diagram is in
escieral agreement with the lake trophic indicators. A more detailed
treatment of this matter will be given in connection with modelling.
-------
243
1.0
I 0.5
O
_J
O.I
EUTROPHIC
L.Erie, 1968.
L.Washington, I960
Dillon, 1982
MESOTROPHIC
, 198!
OLIGOTROPHIC
L. Tahoe
I
10 100
Mean Depth, m
1000
Figure 48. Position of Lake Dillon on the original Vollenweider
diagram (Vollenweider 1968).
-------
244
Nutrient Export in Relation to Land Use
Nutrient export must be considered in relation to land use. The
central information base for this purpose is the representative
watershed program, in which nutrient loss was measured for individual
small watersheds dominated by particular land uses. The names and
locations of the representative watersheds are shown in Figure 49. The
watersheds represented eight different land uses: undisturbed
(background), roads, interstate highways, residential development on
sewer (nonpoint component), urban development on sewer (nonpoint
component), residential development on septic, ski areas, and Climax
{•folybdenum operations. The nutrient yields from these representative
watersheds are considered in sequence below. For each of the watershed
types except the undisturbed type, the background nutrient yield is
subtracted from the total yield and the residual is related
quantitatively to the intensity of land use by some index such as the
number of persons per unit area or the percentage of affected area.
After the nonpoint sources have been treated in this way, consideration
is given to the contributions per capita from small and large point
sources.
Wacer yield enters into many of the nutrient yield relationships
developed below for the representative watersheds. The water yields,
given as mm/year of runoff, are in all instances based on discharge
measurements. Water yields vary a great deal; the causes of this
variation are numerous. There is a pronounced increase in water yield
with elevation. Long-term records of snow accumulation also show that,
for a given elevation, the western portion of the catchment tends to
-------
•JO
c
50
(D
XI
i-l
ro
in
rt
H-
rt
n>
n
en
3"
n>
Q.
en
LASKEY GULCH
(W7B)
DILLON VALLEY
STRAIGHT CREEK
PORCUPINE GULCH
WILDERNEST
(W2A %8
/ ~"—---_
WEST TENMILE (W5B)
KEYSTONE GULCH
(W6B)
COPPER
MTN (W4B)
GULCH (W2A 81)
milts
i j
4.~
Ui
-------
246
receive more winter moisture than the rest. Slope, vegetation, and
land use also have complex effects. A certain amount of water is also
pumped or diverted in various places (e.g., local irrigation). For
1981, the catchment-wide average runoff at lakeside was 180 mm and in
1982 it was 340 mm.
Undisturbed Watersheds
Porcupine Gulch and Laskey Gulch provided information about
nutrient yield from undisturbed areas. Table 49 summarizes the amount
of runoff from these watersheds and the phosphorus and nitrogen yield
in each of the two years. In these and a number of other
representative watersheds the particulate nitrogen contribution was
computed from the percent carbon and total particulates of the stream
by use of a common C:N ratio equal to that of the Snake River, since
the amount of particulate N was often below detection. The standard
error for phosphorus is reported on the basis of weighted averaging.
For nitrogen, the estimation of particulate nitrogen from carbon
precluded estimation of the standard error, but the ratio of S.E. to
mean would be roughly the same as for phosphorus.
The phosphorus yields of Porcupine Gulch and Laskey Gulch were
close to those reported in the literature for cold temperate areas of
similar soil characteristics. For example, Wright (1974) reported 1.5
ng/m~/yr P export for the Boundary Waters Canoe area of Minnesota,
and ::'chindler et al. (1976) reported 5.0 mg/m2/yr for the
~'--:"&-.-imental Lakes Area, Ontario. The data also show great similaritv
-------
247
Water Yield
nan/yr
P yield - mg/m /yr
Total S.E. % Part.
N yield - mg/m /yi
Total % Part.
Porcupine Gulch (W7A)
1981 315
1982 541
1.55 0.34 17
3.52 0.43 32
75
126
7
35
Laskey Gulch (W7B)
1981
1982
Keystone Gulch (W6B)
1981
1982
Upper Snake (SR3)
1981
1982
129
306
108
328
240
500
0.50
2.77
0.57
3.19
1.14
4.37
0.10
0.55
0.08
0.44
0.15
0.55
33
33
33
38
61
54
29
62
14
46
64
103
12
58
19
41
15
51
Table 49. Yield of water, total P, and total N from two watersheds
representing background (undeveloped) conditions (W7A, W7B), and free.
two watersheds with roads but otherwise essentially undeveloped (W6B,
SR3).
-------
248
to data for the Coino Creek watershed at a similar elevation in Eoulder
County (Lewis and Grant 1979).
Both the total phosphorus and nitrogen export showed considerable
variation between years and between watersheds. Table 49 suggests,
however, that this variation was not random; it was associated with the
amount of water yield. As shown by the study of Lewis and Grant
(1979) for the Como Creek watershed, the increase in soluble phosphorus
yield between a dry year and a wet one is expected to be slightly
higher than the increase in water yield. A quantitative relationship
was sought between the water yield and the nutrient yields reported in
Table 49. Since the Como Creek studies suggest that the relationships
are likely to be slightly curvilinear for undisturbed watersheds, the
following equation was used:
v = -y b
"n a *w
where YW equals the annual water yield, Yn equals the annual
nutrient yield (P or N), and a and b are the critical parameters of the
relationship between nutrient yield and water yield. The equation was
log transformed and then tested by linear regression for fit to the
data. The fit was excellent (P < 0.01). A similar procedure for
nitrogen also showed a good fit (P < 0.01).
Casual examination of data from watersheds with only gravel roads
or dirt roads suggested that the effect of these roads might be
sufficiently small that they would not raise the nutrient export much
above background. Since this would be an advantage in that it would
offer a larger number of watersheds from which the terms relating water
and nutrient yield could be determined for background conditions, two
-------
249
watersheds differing from background only by the presence of dirt
or gravel roads were examined. These included Keystone Gulch, which
was a representative watershed (W6B), and a segment of the upper Snake
(SR3). The data for these watersheds are shown in Table 49. Keystone
Gulch actually has near the watershed crest a small septic field that
serves the upper lodge at Keystone ski area, but it is of sufficiently
small size that it cannot account for more than 2% of N or P export
from Keystone Gulch, so the septic effect was ignored.
The data analysis showed that the two background watersheds can be
lumped with the two watersheds containing roads. When the equations
for the two undisturbed watersheds only were used to project the yield
from the two watersheds with roads, and the projections were subtracted
from the observed yields, two of the nitrogen residuals and one of the
phosphorus residuals were slightly negative, suggesting that the yields
of watersheds with roads were very near background. The exponent (b)
derived for phosphorus yield on the basis of the two undisturbed
watersheds was 1.39, and with all four watersheds it was 1.37. The
coefficient (a) was higher by 20% for all four watersheds than for the
two undisturbed watersheds, but this is a small amount in view of the
certainty of variation in background caused by differences in
vegetation, exposure, elevation, and slope. Ve therefore conclude that
all four watersheds should be treated together, and that the effect of
small roads in undeveloped areas is minor enough to merge with
background for present purposes. Thus eight watershed years are
available for determination of background yield equations for N and ?.
-------
250
Table 50 summarizes the results of the log-transform regressions
used to determine the coefficient and exponent of the background yield
equations for the eight watershed years. The goodness of fit is
excellent both for phosphorus and for nitrogen. For phosphorus the
exponent is slightly greater than 1, as expected from the Como Creek
work. For nitrogen the exponent is lower than for phosphorus. Once
again, this is expected for undisturbed watersheds of this type from
the Como Creek studies (Lewis and Grant 1979). The Corno Creek work
shows that nitrate yield typically increases more slowly than water
yield, but that particulate N yield increases faster than water yield.
The two effects partially cancel, producing an exponent in this
instance only slightly greater than 1.0.
__
Exponent Error Coefficient Equation
(b) (M (a)
Phosphorus* 1.372 0.20 0.000782 Y = 0.000782 Y1'3/2
n w
Nitrogen* 1.154 0.17 0.0842 Y = 0.0842 Y1'154
n w
r\
* mg/m~/yr
Table 50. Summary of statistical information on empirical
establishment of the relation Y = ai for background vield
n w 3
(Yv = mm/yr water yield; Yn = mg/m2/yr nutrient
yield). Both equations are highly significant (P<0.001).
Residential Area on Sewer
A. number of developments and subdivisions in the Lake Dillon
w--:nershed consist of clusters of homes or condominiums served by sewer.
-------
25
These make a point-source contribution, which will be considered at the
end of this chapter, and they also potentially make a nonpoint
contribution to nutrient loading. The nonpoint contribution is
associated with the presence of pavement, with changes in vegetation
and soil cover within the development, with fertilizer application,
with more or less continuous minor earth disturbance, and with other
miscellaneous factors arising from human habitation. We consider this
classification to apply to settlements or developments that have enough
vegetative cover to break up runoff and thus slow the transport of
materials to streams. This classification does not apply to areas that
are very densely settled (more than 80% land coverage by concrete or
dwellings). Such densely settled areas are treated below as urban area
on sewer.
Several representative watersheds were initially chosen for
quantification of the effect of residential development served by
sewer. All of these except one proved impossible to ase. Some were
settled too sparsely and others lacked sufficiently coherent drainage
to give good quantitative information. The information is thus based
on a single watershed containing the developments of Wildernest and
Mesa Cortina, the composite of which we shall call Wildernest. Data
are available only for 1982. The data are summarized in Table 51.
This watershed has an area of 246 ha, of which 126 ha is taken up by
the residential development on sewer. The watershed is occupied by
an annual average of 662 persons per day. It is ideal for present
purposes in that it is settled densely enough to show any effects
rather clearly, but is not so densely settled as to constitute an urban
-------
Water yield, mm/yr
Use Intensity, persons/ha
Phosphorus 0
Total yield mg/a~/yr
Std. Error
% Particulate»
2
Background, mg/ra /yr
2
Net yield, ng/m /yr
Net yield per use unit: g/person/yr
Nitrogen
Total yield mg/m /yr
% Particulate,,
Background, mg/m"/yr
2
Net yield, mg/a /yr
Net yield per use unit: g/persan/yr
Residential on Sewer
Wildernest (W2A)
246
2.69
3.29
0.40
29
1.49
1.30
5.63
131
39
43
83
308
Urban on Sewer
Dillon Valley
(W2B)
326
6.83
14.7
4.2
100
2.19
2.89*
4.22
456
55
67
218*
318
* Net yield after subtraction of interstate highway effect as well as
background.
Table 51. Yield of water and nutrients from a watershed supporting residential
area on sewer (1982 data, nonpoint contribution only), and a watershed
supporting urban developtaent on sewer (1982 data, nonpoint only).
-------
253
area. Most of the dwellings or groups of dwellings are separated by
trees and most areas are unpaved.
Table 51 gives the total water yield and total phosphorus and
nitrogen yields for the watershed. The background, as computed from
the background equations, has been subtracted from these to give the
net yield attributable to the land use under consideration here. Table
51 expresses this net yield in terms of the use intensity index, which
in this case is the number of persons. The net yields per use unit in
1982 were 5.63 g of phosphorus per person per year and 308 g of
nitrogen per person per year. These are relatively small in relation
to the sewage export per capita, which would be in the vicinity of
400-600 g per year for phosphorus and about five times this amount for
nitrogen.
The question arises whether or not the yields shown in Table 51
are dependent on water yield. It seems certain that there is a
positive relationship between nutrient yield and water yield, but this
relationship cannot be quantified without more data. It will be
assumed hereafter that total phosphorus and total nitrogen yield have a
dependence on water yield of the same degree as documented for
background yields. With the exponent fixed by this means, the
coefficient can be determined algebraically from Table 51. The
coefficient for P is 0.00295 and for N is 0.536 to give g/person/yr.
Urban Area on Sewer
Although we were not able to sample in such a way as to segregate
cleanly any of the cities within the watershed, we did find one snail
-------
254
watershed segment almost exclusively on sewer that could be treated as
an urban area. This watershed segment, which is referred to here as
Dillon Valley, is shown in Figure 49. Development is much more dense
in Dillon Valley than in Wildernest, which accounts for the difference
in classification. The watershed segment has an area of 115 ha. The
settled area is 65 ha, of which 62 ha is occupied by dwellings or
roads. The area is occupied by a time-weighted average of 786 persons.
The segment is sufficiently small and complex hydrologically that
discharge could not be estimated easily. We therefore assumed runoff
per unit area equal to Straight Creek. Computation of the urban effect
also required subtraction of interstate highway effects. The basis of
this correction is given in the section on yield from interstate
highways. The results are summarized in Table 51.
The total yield of nitrogen and phosphorus for Dillon Valley was
much higher per unit area than for residential area on sewer, but this
was expected because of the higher density of development. On a per
capita basis, the yield above background was similar to the yield for
the residential area on sewer. As with the sewered residential area,
we assume that the yield of both nitrogen and phosphorus is related to
the water yield by an equation incorporating a common exponent (1.372
for P, 1.154 for N). Once the exponents are thus set, the coefficients
of the equations can be determined algebraically from the data in Table
51, are 0.00150 for P and 0.410 for N to give the yield in units of
-------
255
Residential Areas on Septic Systems
Representative watersheds for residential areas on septic systems
were South Barton Gulch and Illinois Gulch. Data are available both
years for South Barton Gulch and in 1981 only for Illinois Gulch. The
data are summarized in Table 52.
South Barton Gulch has a moderate density of development; 194
persons are distributed over 817 ha of watershed. Illinois Gulch is
much more sparsely settled; it has a time-weighted average of 25
persons distributed over a watershed of 496 ha. The three watershed
years provided a good range of water yields. Table 52 shows the net
nutrient yield from each of the watersheds after subtraction of the
background. The net yield is then expressed as per-capita phosphorus
and nitrogen contribution. The data for nitrogen and for phosphorus
suggest that water yield greatly affected nutrient export from these
watersheds. This is to be expected, since nutrients stored by septic
systems in the soil are carried out in proportion to the amount of
water percolating through the soil.
The per capita phosphorus yields were plotted against water yields
for each of the three years, and showed evidence of a strongly
curvilinear relationship. By regression following log transformation
it was established that the equation of best fit is Yn =
3.44YW0'759 where Yn is yield above background expressed as
g/person/yr and Yw is water yield (mra/yr). The fit is good (P =
0.04, S.E. of exponent is 0.10). The exponent is less than l.C, unlike
the exponent for background P yield. The lower exponent implies that
higher water yields were accompanied by P yields that were higher by a
-------
256
South Barton Gulch Illinois Gulch
(W1A) (W2A/81)
Water yield, cm/yr
Use intensity, persons/ha
Phosphorus „
Total yield Eg/m /yr
Std. error
% Particulate,,
Background, mg/nVyr
2
Net yield, mg/m /yr
Net yield per use unit: g/person/yr
Nitrogen ^
Total yield ng/ra"/yr
% Particulate
Background, mg/ia /yr
2
Net yield, mg/m /yr
Net yield per use unit: g/person/yr
1981
190
0.147
4.14
1.61
74
1.04
3.06*
208
40
56
36
1.9*
129
1982
522
0.147
9.83
1.13
61
4.19
5.50*
374
397
72
115
275*
18722
1981
70
/ .1
0.05C
0.70
0.16
38
0.28
0.42
83
13
50
11.7
1.3
257
* Contribution of 74 persons on residential sewer has been subtracted as well
as background.
Table 52. Yield of water, nitrogen, and phosphorus for watersheds containing
residential septic systems.
-------
257
less than equal proportion. Individual septic fields thus appear to
respond in a manner intermediate between that of background sources,
whose yield increases faster than water flow, and a true point source,
whose yield would not be affected by water flow.
The regression procedure was repeated for nitrogen. The fit with
a curvilinear model was poor; fit with a linear model was much better.
The resulting equation is Yn = 44.3YW-5205, where Yn is yield of
N as g/person/yr and Yw is mm/yr water yield. The standard error for
the slope is 11.8 and the significance of the relation is P = 0.08.
The cause of the negative intercept is clear from Table 52: N yield
above background becomes vanishingly small at water yields well above
0. This is explained by biological uptake and retention up to a
certain flushing threshold, by denitrification, or by a combination of
these.
Ski Slopes
The water and nutrient yields from Keystone and Copper Mountain
ski slopes (including lodges, trails, and lifts) were quantified in
1981 and 1982. Contributions from the associated residential areas and
point sources are not considered here. Table 53 summarizes the data.
The yield of phosphorus was well above background. The nitrogen yield
was also above background, but less markedly so than phosphorus. The
percentage of each watershed accounted for by cleared areas was used as
a neasure of use intensity; for both resorts about 40% of the area is
cleared. The number of skiers would probably serve equally well as ar.
-------
258
Water yield, mm/yr
Use intensity, % open
Phosphorus „
Total yield, mg/m /yr
Std. Error
% Particulates
Background, mg/m /yr
2
Net yield, mg/m /yr
2
Net yield per use unit - mg/m used/yr
Nitrogen „
Total yield, mg/m /yr
% Particulate
2
Background, mg/ra /yr
Net yield, mg/m~/yr
2
Met yield per use unit, ag/n used/yr
Keys tone
(W4A)
1981
98
40
1.94
0.41
49
0.42
1.52
3.84
27
59
17
10
25
Copper Mountain
(W4B)
1982
195
40
5.51
1.30
67
1.08
4.43
11.20
75
53
37
33
96
1981
348
42
3.56
0.99
44
2.40
1.16
2.40
74
27
72
2
4
1982
329
42
6.29
1.49
47
2.2'
4,07
8.43
88
37
67
21
44
Table 53. Yield of water and nutrients from watersheds supporting ski slopes
(nonpoint contributions only).
-------
259
intensity measure. In 1981 and 1982 Copper Mountain and Keystone had
between 500,000 and 700,000 skier days per year.
The water yields from Copper Mountain were higher than for
Keystone; this is explained partly by elevation and partly by location
along the east-west moisture gradient. There is no indication in Table
«
53 of the close relation between water and nutrient yield documented
for most other land uses. The explanation is not obvious, since the
mechanisms of nutrient release are complex. Presence of the trails and
slopes undoubtedly facilitates transport, even without human activity,
especially since ski trails and slopes must be cleared parallel to the
gradient. Possibly more important, however, is mechanical damage to
the ground surface when snowpack is minimal. There is also a certain
amount of spring fertilizer application on the slopes. Yields at
either Keystone or Copper Mountain could have been augmented by
construction on or near the slopes.
Since the yield from ski slopes does not show any indication of
sensitivity to water yield, we use a mean. The mean for the four
watershed years in Table 53 is 6.5 mg/m^ of cleared area/year above
background for P and 42 for N.
Interstate Highways
The effects of interstate highways were isolated in the upper
segments of Straight Creek and West Tenmile Creek. Table 54 su^narizes
the data and expresses yield above background in relation to area of
roadway plus right of way. For phosphorus, the yield was unexpectedly
high in both watersheds. The relation of P yield to water yield
-------
260
Straight Creek
(W2B2)
Water yield - mm/yr
Use intensity, % roads
Phosphorus 9
Total yield - mg/n*~/yr
Std. Error
% Particulate
2
Background - mg/m /yr
2
Net yield - mg/m /yr
2
Net yield per use unit - mg/m used/yr
Nitrogen „
Total yield - mg/m /yr
% Particulate
2
Background - mg/m /yr
2
Net yield - mg/m"/yr 9
Net yield per use, intensity - nig /m" used/yr
1982
326
2.97
4.09
0.84
60
2.19
1.90
64
102
61
68
34
1146
West Tenmile
(W5B)
1981
522
0.78
5.38
1.76
47
3.80
1.57*
200
181
18
115
66*
8426
1982
662
0.78
7.19
0.67
50
5.51
1.67*
213
123
36
151
-28*
-3574
* a small contribution from Vail pass public toilet septic field has been subtracted.
Table 54. Yield of water, N, and P from watersheds with interstate highways but
little else.
-------
261
was tested for fit to a function of the form already used for other
nutrient sources. The fit was reasonably good, and resulted in the
equation Yn = O.OOZOgy^'799 where Yn is nutrient yield per
m2 of highway plus right of way and YW is water yield (ran/yr). The
slope has a standard error of 0.53 and P = 0.09 (ideally P should be
below 0.05, but the power of the test is low with N=3). High yield is
probably due in part to major disturbance of vegetative cover. Also
the materials added to roads in the winter may contribute, as does the
fertilization associated with revegetation programs now in progress.
The nitrogen yield associated with roads was low by comparison
with other sources and less consistent than P, as shown by Table 54.
The N yield above background for West Tenmile was negative one year.
This is quite possible due to excess of uptake or denitrification over
yield, but presents some difficulties in prediction. For purposes
of curve fitting, the two yields from West Tenmile were averaged,
giving a low positive yield. The fit to an exponential equation was
then reasonably good: Yn = 1.71Y,,1-13* S'E' sl°Pe = °-40> p =
0.11, where the units are the same as for the phosphorus equation.
Mining
The Climax Molybdenum mining operation is a unique mixture of
nutrient sources and must therefore be treated separately fror. other
land uses. Water from Climax Molybdenum has a high suspended and
dissolved solids content. The water is retained in tailings por.cs,
where sedimentation occurs, and the ponds are purged during high water-
The ponds receive not only industrial wastewater but also domestic
-------
262
effluent derived from the toilets and kitchens of the plant. The
domestic effluent is not treated, except by sedimentation in the
pcnds.
Table 55 summarizes the water and nutrient yields from upper
Tenmile Creek where the active mines are located. There is little else
in this watershed except old mines and State Highway 91. Roads,
tailings, and bare areas around the plant account for 753 ha of the
total 6447 ha. Virtually all of the nutrient yield above background
can thus probably be attributed to Climax Molybdenum.
The P yields were above background and the N yields were even more
so in 1981 and 1982. Neither seems related to water yield. Unlike
most watersheds, the total nutrient yields for upper Tenmile were
higher in 1981 than 1982. In 1981 the time-averaged work force at
Climax Molybdenum was 2650 persons and in 1982 it was 1100 persons.
Tne yield of both N and P thus seems related to size of work force.
Table 55 expresses the yields both years on a per capita basis; the
yield per capita was almost identical for the two years. We will thus
treat N and P yield at Climax Molybdenum as a function of work force
size.
Tne ? yield per capita at Climax Molybdenum suggests that the ?
output above background is almost entirely a sewage contribution and
not an industrial effect, since the yield per capita is in the range of
what would be expecced from the domestic output of the workforce. The
ponds appear to be effective in reducing particulate P losses that
would result from earth disturbance. N export is very high, suggesting
major augmentation by industrial activity. Ammonium nitrate is used in
-------
263
Upper Tensile - W5A
1981
1982
Water yield - nnn/yr
Use intensity - persons/ha
Phosphorus „
435
0.41
513
0.17
Total yield - rag/m /yr
Std. Error
% Parciculate
2
Background - mg/m /yr
2
Net yield - mg/m /yr
Net yield per use unit -
g/person/yr
Nitrogen -
Total yield - mg/m /yr
% Particulate
2
Background - mg/m*"/yr
2
Net yield - rag/m /yr
Net yield per use unit -
g/person/yr
7.07
1.55
20
3.26
3.81
93
821
5
93
728
17710
5.84
0.64
58
4.09
1.75
103
491
15
113
378
22150
Table 55. Yield of water, N, and P for the raining area on upper
Tenmile Creek. Use intensity is based on nunber of persons
working at Climax Molybdenum.
-------
blasting and may be the source of this augmentation. Revegetation may
also contribute.
Point Sources
There are two types of point sources in the Lake Dillon catchment:
large sources with tertiary treatment and small sources with secondary
treatment. We will establish here a contribution per capita for each
category. Large point sources with tertiary treatment include the
Snake River WWTP and Frisco WWTP, whose effluents were sampled
directly. The results for these two have already been reported in the
chapter on total loading of trie lake, but have not yet been converted
to a per capita basis. In addition, data can be obtained for the
Breckenridge WWTP as the difference between stations ER1 and BR2 on the
Blue River (just downstream and just upstream of the Breckenridge
WWTP outfall). These data are summarized in Table 56. The Copper
Mountain WWT?; designed as a tertiary plant, was in a transition from
secondary to tertiary during the course of the study and will therefore
not be considered here as a representative of tertiary treatment.
PersonsP - g/person/yrN - g/person/yr
Served* 1981 1982 1981 ' 1982
Frisco WWT? 3300 11.7 84.8 1108 955
Snake River WWT? 3230 9.1 22.3 1003 1260
Breckenridge WWT? 6700 113. 44.0 1378 1479
* Annual average persons per day
Table 56. Yield of P and N from three WWTP plants on a per capita
basis. All three plants practice tertiary P removal.
Table 56 shows the effectiveness of the tertiary treatment, since
all of the per capita ? yields are well below the raw per capita
-------
contributions coming to point sources (400-600 g/person/yr). At the
same time, there was considerable variation in P yield per capita
between years, even at the same plant. This is explained entirely by
breakdown or shutdown of tertiary treatment, which occurred at all
plants on one or more occasions over the course of the study. Although
it could be argued that tertiary sources should be represented by their
yields when treatment facilities are operating properly, which would
give yields about half as great as the observed average (Table 57), it
is probably realistic to assume that the plants will be impaired for a
certain amount of time each year.
g/person/yr
Phosphorus
Secondary treatment 270
Tertiary treatment 47
Nitrogen
All treatment 1197
Table 57-Average point source yields on a per capita basis.
Nitrogen yields from the point sources were much more uniform,
since they were not affected as much by variations in plant operation.
The Breckenridge plant practices ammonium removal, but this does not
affect total N yield per capita (Tables 58, 57).
The P contribution from small point sources (package plants)
subject to secondary treatment was estimated from data taken upstream
and downstream of such sources along the Blue River above Goose Pasture
Tarn. The results are reported in Table 57.
-------
266
Overview of Nutrient Yields
Export of N and P from various sources is expressed in Figure 50
on a per capita basis wherever this is meaningful. In Figure 51 the
export rates are recast on an area basis, assuming runoff is near that
of an average year. Among the point sources, the potency of sources
without tertiary treatment is striking. For nonpoint sources, septic
systems stand out, as expected, as do interstate highways, which might
not have been expected.
-------
P, g/person/yr 25:
0 40 80 120 160 200 240 280
_ Plant with Secondary P Treatment
P
N
Plant with Tertiary P Treatment
P
N
Climax molybdenum, per worker year
P
N
\ \ V\]
1
Septic
N
Urban on sewer
N
Residential on sewer
pg
Nf
Ski Slopes
\ \ I Point source
I I Non-point source !
6 9 12 15 18 2!
N, kg/person/yr
Figure 50. Nutrient yield from various sources expressed on a pe.
carita basis, assuming 300 :m runoff. Sources that car.r.cr
be ixoressed on a per capita basis are r.et shown.
-------
P, mg/mVyr
0 20 40 60 80 100 120
140
Urban with Tertiary P Treatment, 12 persons/ha
p|\\ \\\\\\\\1i
Residential with Tertiary P Treatment, 2.7 persons/ha
Residential with Secondary P Treatment, 2.7 persons/ha
P i\ \ \V\ \ \\X\\\\\ \T1
?
N
P
P
PQ
P
N
Residential on Septic, 0.5 persons/ha
CMmax , 753 ha in use
interstate, pavement plus right cf way
Ski areas, 40 % cleared
Undisturbed
Point source
Non-point source
0 2000 4000 6000 3000
N, mg/m2/yr
Figure 51. Nutrient yield from various sources expressed on an areal
basis, assuming runoff of 300 mm.
-------
269
Separation of Nutrient Sources within the Watershed
Chemistry and discharge measurements were taken at a number of
points along each one of the three major rivers (Figure 3). Each of
the river drainages was thus divided into a number of watershed
segments. If the land use within any given segment is known
quantitatively, the equations that have been developed in the previous
chapter for nutrient yield associated with various land uses should
provide a prediction of the nutrient yield from that segment. Three
important factors may cause deviations between the observed and
predicted nutrient yields from any given watershed segment.
(1) Estimation errors. The observed yields for any given segment
are based on chemistry and discharge measurements, both of which are
subject to analytical error. We may also include under the heading of
error variance the confidence limits around any one of the curves used
to predict nutrient yield from land use. These confidence limits
reflect not only error variance in the original data from which the
curves were developed but also a certain amount of individuality ir.
watersheds with respect to soils, vegetation, exposure, slope, and
elevation. If these sources of error were the only cause of deviations
between predicted and observed, we would expect to see a random
assortment of positive and negative deviations, between the observed and
predicted values.
(2) Unexpected sources. The second cause of deviation between
•'bserved and expected yields for individual segments is the preser.c. of
nutrient sources that are either unquantifiable or unknown. In the cr.e
-------
170
instance where such a source proved to be important, it was possible to
determine the source of unusual and unexpected nutrient yield.
(3) Storage. The third factor affecting the agreement between
observed and predicted yields is a storage effect associated with areas
of low relief in the river bottoms. This has not been incorporated
into the equations used for prediction. The equations were developed
on the basis of small watersheds off the main channel where individual
land uses could be isolated. When nutrients flow to lower elevation,
and particularly into a much broader stream channel incorporating
low-lying areas and wetlands, significant amounts of nutrient can be
stored in years of low flow and purged in years of high flow.
Mechanisms of phosphorus storage include sedimentation of particulate
phosphorus, biological uptake of soluble phosphorus, and ligation or
inorganic precipitation of soluble phosphorus. For nitrogen, the
possibilities are more complex, including all of the possibilities for
phosphorus plus cenitrification, by which oxidized nitrogen species are
converted to nitrogen gas, which is subsequently released into the
atmosphere. All of these phenomena are most likely in the river
bottoms and associated flood plains and wetlands.
Nutrients lost by sedimentation and ligation or uptake in years of
low water may be returned to the open channel and thus enter the lake
during a year of high water when the physical forces at work in the
river bottoms are sufficient to move large amounts of accumulated
materials. Thus the river bottoms can act as a storage site that
retains a portion of the nutrient yield in dry years and gives back
much or all of this retained portion in wet years. These storage and
-------
release phenomena are most correctly regarded as temporary changes in
the watershed nutrient inventory from one year to the next, and will be
referred to below as the "inventory change." If inventory change is
important in determining the yield of nutrients to Lake Dillon, we
should see in dry years a consistently lower ratio of observed to
predicted yield in river segments incorporating areas where the river
water moves through wetlands, gravel beds, or ponds. If much or all of
the storage (increase in inventory) is given up in a wet year (decrease
in inventory), we would expect consistent increase in the ratio of
observed to predicted yields from these segments in a wet year. If
nutrient inventory change is shown by the data from individual streaa
segments, then additional equations must be developed to account for
year-to-year inventory change of nutrients due to their accumulation in
low-lying areas in dry years and their release from these areas in wet
years.
Snake River Drainage
The Snake River drainage was divided into four segments as shown
in Figure 52. Histograms of the observed and predicted yields are
superimposed on each one of the stream segments. Phosphorus yields are
presented on the left side of the river channels and nitrogen yields or.
the right side of the river channels. For each year and for each
segment, one histogram bar shews the observed yield fron the segner.t,
ard the adjacent bar shews the predicted yield. The scales for t':.e
yields are all the sanie for a given nutrient, but differ between the
two nutrients because of the larger yields of nitrogen.
-------
OQ
C
i-f
fD
fu O pd O
3 3" H- cr"
'<: fu < C/i
3 it)
3
OQ
H-
<
(D
fD
ft)
M s;
- |D
rr
fu (D
en 3 n
fD
«8
fD
0) O.
3 fD -a
H- a. t-t
o H-
OQ 'rJ n
fD ^ rt
3 O ft)
cn fi,
H- 'a
cn 3^ '•^
o H-
o ^i m
3 £ M
cn fx
rr w
D" H-
(D W Hi
i-i O O
OQ
3* rt CO
rt 3" ft)
(t> 00
O S
HI h-- a,
ft 3
rt t-ti rt
3J rt Ln
fl)
O O
D Hi Hi
3'
fu (t rt
3 (6 fD
(D
I-1 cn cn
rt 3
H* M [11
3 fD JT"
P ft)
PHOSPHORUS
3
\
E
0
NITROGEN
- 150
SNAKE
RIVER
OBSERVED
PREDICTED
ro
~-j
10
-------
For phosphorus the agreement between predicted and observed yields
from individual stream segments is generally very good. One notable
exception is the phosphorus yield in 1982 from segment SR2, which
incorporates Keystone Gulch, Keystone Ski Area, and Keystone
Development. The observed phosphorus yield from this segment was
extraordinarily high in 1982; it was far above any expected random
deviations from the predicted. This exceptionally high observed
phosphorus yield is explained by construction activities. Keystone Ski
Area installed six 300-hp pumps in the river bed during 1982 and
constructed an associated pipeline to carry the water from thesa pucps
to ski slopes, where it was to be used for snowmaking. Construction
was carried out at several other locations in various watersheds during
the course of the study but did not produce such extraordinary yields
of phosphorus. The especially high yield in this instance is
undoubtedly accounted for by work in or near the river bed, where water
movement greatly facilitated the transport of phosphorus-bearing
materials loosened by the construction activities. Since the
prediction equations do not take into account any such special
activities, the gap between observed and predicted P of 11 mg/m'Vyr
can be accounted for specifically by the construction. Total 1982
transport due to the construction in this segment was 295 kg of
phosphorus. This was approximately 7% of the runoff yield of
phosphorus for the entire Lake Dillon watershed during 1982. Expressed
in another way-, it was equivalent to the annual yield of phosphorus
from a tertiary point source serving 6300 persons, assuming efficiency
of point source operation as shown in Table 57.
-------
274
There is little evidence for significant valley bottom inventory
changes affecting phosphorus yield in the four Snake River segments.
The inventory change effect is unlikely in SR3 and SR4 because of the
steep relief in these segments. The effect is most likely in segment
SR2, where it would be obscured by the large construction effect in
1982.
Predictions for nitrogen are also in good agreement with the
observed values with the exception of the 1982 value for segment SR2.
Once again, we associate the extraordinarily high yield in 1982 for SR2
with the Keystone construction activities.
Blue River Drainage
The Blue River drainage was originally divided into seven
segments (BR1-BR7). Figure 53 shows the segment yields based on a
total of four segments. Segment BR1 has been omitted because this
small segment isolates only the Ereckenridge Wastewater Treatment
Plant, whose yield has already been discussed in the previous chapter.
Some of the smaller segments between BR2 and BR7 have been combined for
present purposes because a considerable land area is required to show
a quantifiable effect on the chemistry and discharge of a river as
large as the Blue River (BR2 in Figure 53 includes the original 3R2 and
ER3; BR4 in Figure 53 includes original BR4 and BR5). In addition,
diffuse drainage in the vicinity of Breckenridge made impossible any
reliable determinations of discharge there.
There were no extraordinary departures of observed from predicted
phosphorus yield in the Blue River drainage. The Blue River drainage
-------
275
PHOSPHORUS
3r
0>
E
0
NITROGEN
150
BLUE RIVER
Observed
Predicted
53. Observed and predicted yields from segments of the 31ue
River watershed. Phosphorus is on the left of the stra?.:
channels, nitrogen on the right.
-------
276
is a good place to look for valley bottom inventory change because the
main channel, even as high as BR7, incorporates extensive wetlands,
small ponds, and accumulated old tailings in the river bed. In every
segment of the Blue River drainage, the ratio of observed to predicted
phosphorus was lower in 1981 than in 1982. This is explained by tha
inventory change effect, which leads to the entrapment of a certain
amount of the yield in dry years, and a release of a portion of this
material during wet years.
In BR7, the nutrient sources drain through extensive areas of low
relief. The predictions were consistently above the observed yields in
this portion of the watershed for phosphorus, and the ratio of observed
to predicted was higher in 1982 than in 1981. The data thus suggest
that significant retention occurred even in the wetter year of 1932,
although it was especially pronounced in 1981. The same was true of
BR6, which incorporates Goose Pasture Tarn as well as an extensive
diffusely drained area above the Tarn.
Segment BR4 has a much different character from that of BR7 and
BR6. The river bottom has been extensively modified, both by mining
and by settlement around the town of Breckenridge. The phosphorus data
show phosphorus accumulation in low-lying areas in 1981 and a strong
purging of the stored material in 1982. The behavior of Segment BR2
was more similar to that of BR6 than BRA.
There was considerable difference between the yield patterns of
nitrogen and phosphorus in the Blue River drainage. For several
reasons, phosphorus and nitrogen cannot be assumed to behave similarly
'-" any given watershed segment. First, a great deal of phosphorus is
-------
277
transported in particulate form, whereas the portion of nitrogen
transported in particulate form is generally much smaller. Thus the
factors that lead to entrapment of significant amounts of particulate
phosphorus in a given watershed segment will not necessarily have a
similarly strong effect on total nitrogen yield from that segment.
Furthermore, nitrogen yields can be affected by denitrification losses.
Denitrification is especially likely in low-lying areas with diffuse
drainage where anoxia can occur. Finally, the biological uptake of
phosphorus and nitrogen is likely to differ-
The ratio of observed to predicted nitrogen yield for individual
segments was higher in 1982 than in 1981 for all segments except BR7.
The data thus indicate operation of the inventory change effect that
was observed for phosphorus. However, it is clear that other factors
complicated the yield of nitrogen from most of the segments. Negative
yields (i.e., N sources plus incoming N greater than outgoing I,") were
observed for segments BR2 and BR6 in 1981. Yields in segment BR6 were
exceptionally low even in 1982 by comparison with the prediction. The
major deviations of segment BR6 from the predictions are almost
certainly connected with losses of nitrogen associated with Goose
Pasture Tarn. Concentrations of nitrogen above and below the Tarn
indicate that it is a major nitrogen sink. Nitrogen loss in the larr.
is probably due to a combination of uptake by weedbeds, sedimer.taticr.,
and denitrification in the mud.
-------
273
Tenmile Creek Drainage
The data for Tenmile Creek are shown in Figure 54. Three segments
are shown. Smaller segments have been combined for reasons similar to
those mentioned for the Slue River. The upper segment is the site of
Climax Molybdenum, for which predictions are good. The middle segment
(TM4) has numerous dispersed sources of nutrients, and is also the
discharge point for the Copper Mountain Wastewater Treatment Plant.
The presence of the wastewater treatment plant complicates the
interpretation of the data. Tne Copper Mountain plant was doing
secondary treatment during almost all of 1981, and the effluent was for
the most part pumped into a pond instead of into the river directly.
Tertiary treatment was in force, but was considered not fully effective
according to plant operators, during the first ten months of 1982,
after which secondary treatment was resumed for the months of November
and December. The presence of the wastewater treatment plant, and the
problems associated with its new treatment practices, make the
estimates of other factors difficult because the plant introduced nuch
more irregular variation than would be expected in other watershed
segments. There was major overestimation of phosphorus for TM4 in 1981
but not in 1982. The 1981 overestimate is probably due to storage in
the pond and in the river bottom in the lower half of the segment.
Storage would be very likely in the poorly drained areas known as
Curtain Ponds and Wheeler Flats. However, segregation of this storage
and purging effect is more difficult in T>!4 than in any other location
because of the changing operations of the point source.
-------
279
PHOSPHORUS
3r
NITROGEN
-
-
-
150
100
50
0
w
>•%
CVJ^
£
\
E
81 p 82
ENMILE CREEK
Observed
i Predicted
54. Observed and predicted yields froci segments of the Ter.r.ile
Creek watershed. Phosphorus is on the left of the strea~
channels, nitrogen on the right.
-------
ISO
Segment TM1 incorporates a large portion of the town of Frisco.
Since the town is located just adjacent to the lake, the drainage of
Tenmile Creek near its mouth is modified in a major way by the presence
of the city. There is a large area of flat relief in the vicinity of
Frisco. Probably because of this flat area, the phosphorus and
nitrogen yield data for TML show evidence of inventory change between
the dry and wet years. Both phosphorus and nitrogen yields for 1981
were negative, indicating that the areas of low relief within this
drainage stored not only some of the yield from segment IM1 but also
some of the yield of upstream segments. However, the observed storage
in 1981 was almost exactly compensated by excess of observed over
predicted yield in 1982, indicating a purging effect. The purging
effect was less pronounced for nitrogen, possibly because of permanent
nitrogen losses to denitrificaticn or biological uptake.
Relationship between Runoff and Nutrient Storage
Since the data show that nutrients originating at various points
in the watershed are stored in the river valleys during years of low
runoff and purged from these areas during years of high runoff, the
yield equations cannot predict total output from the three rivers until
they have been coupled with an inventory change function. An estimate
can be made of the function from the data at hand. Inventory change
will be positive in those years when part of the yield from areas of
steeper relief accumulates in areas of low relief and will be negative
wner. part of this stored inventory is released during wetter years.
-------
281
For nitrogen, the function will be considered to include any effects of
denitrification.
The first step in estimating the inventory change function is to
sum up the observed and predicted total yields from runoff for each of
the two years of the study. The lakeside point sources (Frisco, Snake
River, and Breckenridge WWTP) are excluded from the summation, since
they are so close to the lake that the nutrients they give off are not
subject to storage. The summations are corrected for major
discrepancies from the predicted values that are known not to be
connected with storage. In the two years of record, the only example
of this is the boost of yield in the Snake River watershed during 1982
caused by construction in the vicinity of Keystone. After this
correction, the remaining discrepancy between observed and predicted
yields from all sources combined is assumed to be due to inventory
change. Since inventory change is dependent on amount of runoff, we
postulate a decline from positive to negative inventory change as
runoff goes from the lowest to the highest annual values. A coroll.-sr/
is that the average change in inventory over a number of years is not
very great. Even if the watershed is aggrading, as it may be with net
accumulation of vegetation and filling in of excavated areas, the
resulting departure from equilibrium over a number of years would
amount to a relatively small percentage of nitrogen and phosphorus flux
(cf. Bormann and Likens 1979).
Figure 55 shows the apparent change in nutrient inventory for tn=
two years of record for both nitrogen and phosphorus. The nutrient
inventory change is expressed as a percent of watershed overland
-------
282
o>
c
o
JZ
o
>»
V.
o
•«••
>
o
13
C
Average Runoff
t
P: y»- 0.40x + 87.6
: y*-0.57 x + 112.4
100 200 300
GAUGED RUNOFF, thousands of acre feet/yr
Figure 55. Linear plot of the inventory change functions for nitrogen
and phosphorus. Positive inventory change indicates
storage. Inventory change is expressed as a percentage of
the summed runoff nutrient sources.
-------
283
nutrient yield, excluding lakeside WWTP sources. The percentages were
calculated as follows:
(Y - Y )
wn = _J^ £•£.. x 100
n,o
where Wn is nutrient inventory change (%), Yn o is observed
overland yield, and Yn e is the expected overland yield from
summation of all sources with the assumption that inventory change is
zero. Figure 55 shows the relationship of Wn to discharge. The two
points (1981, 1982) for each of the two nutrients are joined by
straight lines. Although the form of the relationship is not
necessarily linear, numerous additional points would be required to
show curvature. Unless there is very marked curvature, the line will
provide reasonable approximations.
The watershed appears to be aggrading with respect to phosphorus,
since an average water year is accompanied by a 10-15% retention of
yield in valley bottoms. Nitrogen also shows some net retention in an
average year, but less than phosphorus. The equations for the lines,
as indicated in Figure 55, will be used to represent the expected
change in inventory in any particular year.
Comparison of Observed and Predicted Total P and N Loading
The land-use equations, point source equations, and
inventory-change function lead to predictions of total runoff yield.
All identifiable sources are dealt with by these equations except rnajor
construction in or near stream beds, such as that observed near
-------
284
Keystone in 1982. Although the frequency of this activity is largely
unpredictable, it will surely be above zero. As an approximation of
the effect, we use the average increase in yield over the two years of
record expressed as percent of other runoff. This boosts the P
prediction by 3.5% and the N prediction by 2.2%.
When the equations for yield from all watershed sources are used,
and a mean figure is used for contributions from precipitation and
groundwater, the result is an overall predicted loading of N and P for
the lake in 1981 and 1982. These predicted loadings are shown in Table
58, which also shows the observed loading for comparison. The observed
and predicted values agree very well (<10% difference). Deviations of
predicted from observed are caused mostly by the irregular nature of
failures in tertiary treatment plants, and to some extent by irregular
variation in other sources.
P load, kg/yr N load, kg/yr
Year Observed Predicted Observed Predicted
1981 2900 - 2700 85,600 83,500
1982 4800 4600 151,400 154,300
Table 58. Comparison of predicted and observed N and P loads for the
two years of study.
Itamization of Lake Nutrient Sources for 1981 and 1932
In the chapter dealing with total nutrient loading of the
lake, nutrient sources were itemized according to the major pathways by
which they entered the lake (Tables 46, 47). On the basis of the
r-;trient yield equations, nutrient sources can be further broken down
according to land use. In making this breakdown, we use the values
-------
285
reported in Tables 46 and 47 for precipitation, groundwater, and
effluent contributions from the Frisco and Snake River plants. The
contributions of the Breckenridge Wastewater Treatment Plant are taken
from the analysis of the watershed segment BR1 as reported in Table 56.
The contributions of all other sources are obtained by application of
the yield equations and the inventory change equations to the land use
data and water yield data for 1981 and 1982. Since the ability of
these equations to predict observed yield accurately for 1981 and 1982
has already been demonstrated, it is expected that the sum of
individual sources obtained by this means will be very close to, but
not exactly equal to, the observed loading.
Table 59 shows the complete breakdown of sources. The table gives
the kilograms of nitrogen and phosphorus from each source reaching the
lake in 1981 and 1982, and also expresses these as a fraction of total
annual loading. Although data for both years are given, it should be
noted that the distribution of contributions for 1982 is much more
typical of the median or average situation under present land use
conditions than is the distribution for 1981. There are several
reasons for this. First, the runoff for 1981 was exceptionally low.
The sum of flows for the 22 years of record at the four U.S.G.S. gauges
averages 185,600 acre feet per year. The gauged flow for 1981 was
116,500, or 63% of the average. Although the runoff for 1982 was above
the average (210,900), it was only slightly so (14%) and thus
represents more accurately the median distribution of various
contributions to the total nutrient loading of the lake. In addition,
the wastewater treatment plant at Copper Mountain was operating under
-------
Phosphorus
Nitrogen
1981
1982
1981
1982
Mean %
Tertiary Outfall.-,
Frisco
Snake River
Breckenridfie
Copper Mountain
Subtotal
Secondary Outfalls
Package plants
Copper Mountain
Subtotal
Climax Molybdenum
Keystone Construction
49 1.6
322 10.8
371 12.4
152
5.1
49
47
96
104
295
1.1
1.0
2.1
2.2
6.4
1.3
5.9
7.2
3.4
3.2
198
1305
1503
0.2
1.5
1.7
28583 33.0
Mean
39
29
757
-
825
1.3
1.0
25.3
-
27.6
280
72
295
41
688
6.1
1.6
6.4
0.9
14.9
3.7
1.3
15.9
0.5
21.3
3656
3240
9233
-
16129
4.2
3.8
10.7
-
18.7
3152
4070
9909
1893
19024
2.1
2.7
6.5
1.2
12.5
3.2
3.2
8.6
0.6
15.6
198
378
576
23736
3883
0.1
0.2
0.4
15.6
2.6
0.2
0.8
1.0
24.3
1.3
Dispersed Nonpoint Sources
Residential
Urban areas
Septic
Interstate
Ski slopes
Background
Subtotal
Preci pi tation
Croundwater
areas (sewered)
(sewered)
highway
runoff
GRAND TOTAL
9
9
161
21
36
545
781
846
13
2988
0.3
0.3
5.4
0.7
1.2
18.2
26.1
28.2
0.4
99.8
32
31
383
132
59
2105
2742
606
88
4619
.7
.7
8.3
2.9
1.3
45.6
59.4
13.1
1.9
100.0
.5
.5
6.9
1.8
1.3
31.8
42.8
20.7
1.1
99.7
487
761
6713
461
214
16577
25213
14311
587
86326
.6
.9
7.8
0.5
0.2
19.2
29.2
16.6
0.7
99.9
1901
2702
21750
2198
428
69015
97994
5433
1549
152196
1.3
1.8
14.3
1.4
0.3
45.4
64.4
3.6
1.0
100.1
1.0
1.4
11.1
1.0
0.2
32.3
46.8
10.0
0.8
100.0
10
CO
Tali! c S9. Hreakdown of nutrient contributions to Dillon in 1981 and 19S2.
-------
287
secondary treatment during 1981, whereas the standard of operation in
the future, and for most of 1982, is tertiary treatment. Finally, the
performance of the largest wastewater plant (Breckenridge) in 1981 was
unusually poor due to equipment shutdown and malfunction.
The various nutrient sources have been grouped in a number of ways
in Table 60to emphasize different aspects of the distribution of the
load among sources. In 1981 the contribution of wastewater treatment
plants to the total loading was exaggerated by low runoff, which
increased the relative role of these sources, and by the difficulties
at the Breckenridge Wastewater Treatment Plant. The 1982 situation,
which was more characteristic, showed about 16% of the phosphorus and
about 13% of the nitrogen coming from the four major wastewater
treatment plants combined. Package plants made only a minor addition
to this total. Thus it is clear that well over three-quarters of
phosphorus and nitrogen loading under the present land use conditions
comes from nonpoint sources of natural and man-made origin.
Climax Molybdenum contributed about five percent of the total
phosphorus loading in 1981 and about half as much in 1982. The decline
is explained by the reduced pace of operations, and finally the
shutdown in 1982, which drastically reduced the nutrient output froc
that source. Thus for Climax the 1982 data probably underestimate the
equilibrium contribution, assuming that the mines ultimately resume
operation. An equilibrium number would thus be more in the vicinity of
five percent than two percent. Although Climax has not traditicnall..
been thought of as a point source for nutrients, it actually behaves
much the way a point source would, especially in the apparent direct
-------
288
P load, %
Categories
Natural vs. Man-made
Natural
Man-niade
Point vs . Non-point
Wastewater plants
Package Plants and
All other sources
Hunan Was te ( WWT? ' s ,
vs. all others
Human waste
Other sources
1981 1982 Mean
47
53
38
Climax 7
55
Package Plants and
45
55
54
46
16
3
81
Septic)
26
74
50
50
27
5
68
36
64
N
1981
36
64
20
33
47
28
72
load, %
1982 Mean
50 43
50 57
13 16
16 24
71 60
28 23
72 72
Table 60. Percentage contributions to loading of the lake aggregated in
three different ways. The 1982 data are most representative of
a median year.
-------
289
dependency of the nutrient yield from the source on the size of the
workforce. It has already been noted that the nitrogen yield from
Climax is exceptional, and this is shown very clearly in Tables 59 and
60. The nitrogen yield of Climax Molybdenum is greater than all four
wastewater treatment plants combined.
Among the dispersed nonpoint sources, the background contribution,
including everything that would be expected in the absence of human
habitation, is considerably greater than all others combined. Under
present conditions, it accounts for about half the total loading of
nitrogen and phosphorus. Among the dispersed nonpoint sources related
to human activity, septic systems stand out clearly as the highest
contributor. All other sources account for contributions of a fraction
of a percent to a few percent each. Although no single dispersed
source related to human activity is of outstanding importance except
for septic systems, the aggregate effect of the dispersed sources is
significant.
Table 60 shows that the natural background and the sources
associated with human presence share the phosphorus loading almost
equally for phosphorus, and for nitrogen are only slightly skewed
toward the man-made sources. In general Table 60 shows that, under the
present land-use conditions, in a median year hydrologically, both the
phosphorus and nitrogen loading of the lake can be considered to be
divided approximately into four quarters. The first two of these
quarters are taken up by the natural sources. The third quarter is
taken up by sources associated directly with human waste (wastawater
treatment plants, package plants, septic systems), and the fourth
-------
290
quarter is accounted for by human activity not related to the
processing of human waste. In this last category we have a wide
assortment of mechanisms including earth disturbance, change in
vegetative cover, fertilization, and others.
Figure 56 gives a visual impression of the nutrient contributions
from various sources. The 1982 data are shown in the figure because
these are most representative of the median condition at the present
time.
Possibilities for the Reduction of Nutrient Loading
The foregoing analysis allows us to suggest a number of ways the
nutrient load reaching the lake might be reduced. The maximum
imaginable reduction under the current land use conditions would be
about 25%, assuming virtually no limit on financial investment and
that no water is diverted away from the lake. Approximately 50% of the
present loading (i.e., the natural background) is completely
uncontrollable and roughly half of the remaining portion would probably
be uncontrollable even with maximum investment. Thus the following
alternatives in aggregate probably represent no more than 25% of the
total present loading.
A. Major point sources. There are four major wastewater
treatment plants in the drainage. We have assumed up to this point
that all four plants will operate much as did the Breckenridge, Frisco,
and Snake River plants in 1981 and 1982. Since the Copper Mountain
P^ant -.ras changing ever to tertiary treatment during this interval, we
taKe t'r.e ctrier three T?lants as a better indication of the immediate
-------
29:
All other sources
Precipitation
13%
Background Runoff
Major treatment plants
Septic
PHOSPHORUS
All other sources
Background Runoff
Major treatment plants
NITROGEN
ngure
Percentage of loading due to various sources in 1982.
sources over 5% are shown separately.
-------
192
future situation for all four wastewater treatment plants. The three
operating tertiary plants added at least twice as much phosphorus to
the lake in 1981 and 1982 as they might have if they had been able to
operate continuously without equipment failures. Thus a reduction of
something of the order of 8% in total phosphorus loading to the lake
might be achieved if tertiary plant failures could be held at a level
well below the level observed in 1931 and 1982, or if there were some
kind of fail-safe system for each plant. Perhaps the disruptions of
operation will be far less numerous when all of the plants have reached
their capacity and thus are no longer under construction or expansion.
If so, then a significant reduction of nutrient loading from this
source may occur- At any rate, it may be important to pay closer
attention to the occurrence and duration of plant failures and to
embark upon a more vigorous program by which these can be minimized or
avoided.
B. Climax Molybdenum. Climax Molybdenum is a significant source
of phosphorus and behaves essentially like a point source. With the
workforce at 1981 levels, the phosphorus contribution from this source
will be roughly five percent. If tertiary treatment standards were
applied to this source for phosphorus, the contribution could be
reduced considerably, perhaps to as little as one percent of the total
phosphorus loading of the lake.
C. Construction. The 1982 data on the Keystone construction in
the -naVe River bottom illustrates the potential exacerbation of normal
phosphorus loading by heavy construction in or very near the river
beds. Although such construction is not routine, it will probably
-------
293
continue to occur at irregular intervals. Precautions to prevent the
dispersion of disturbed earth and sediments from entering the stream
channel might easily reduce this source by half.
D. Sewered residential areas. Runoff control in sewered
residential areas, including especially the densely settled urbanized
zones, could potentially reduce nutrient input to the lake by one
percent if applied to all sewered areas. Although reductions of this
magnitude may seem scarcely worth bothering with, it is obvious that
any and all sources of nutrient reduction should be considered if
additional development of the watershed is to be expected. The data
show that nutrients associated with sewered residential areas, and
particularly with urban areas, are transported specifically in the
wettest years, and presumably at the time of greatest runoff.
Technology for the ponding of runoff of this type is available and
would be of interest in reduction of the nutrients from this source.
However, removal of the trapped nutrients to prevent later transport to
the lake would be essential if the control program were to be really
successful.
E. Septic Systems. Septic systems are a major contributor of
phosphorus. If septic systems could be replaced by sewer, the yield
from these areas could probably be reduced by half (from about eight
percent to about four percent of the total). Greater reduction than
this is probably unrealistic for several reasons. First, a certain
fraction of the contribution is associated with the presence of a
residences and not with the presence of septic fields. Secondly,
abandoned septic fields will presumably continue to yield phosphorus
-------
294
and nitrogen for some time to come, and the decline in yield may not be
immediate. Finally, tartiary treatment following the addition of sewer
cannot be expected to remove all of the phosphorus.
Present and future yield of septic systems could probably be
reduced significantly by proper maintenance of septic fields. This has
not been a local government priority in the past but perhaps could be
in the future as part of a nonpoint control program.
F. Interstate highway. Interstate highway contributed 2.8
percent of the total phosphorus yield in 1982. This is relatively high
in view of the fact that interstate highway does not cover a very large
percentage of the watershed and does not require the disposal of any
human waste. In other words, interstate highway is a surprisingly rich
source of phosphorus on a unit area basis. It is difficult to
prescribe control measures for this source, since the mechanism by
which phosphorus yield comes from interstate highways is not clear.
The application of phosphorus for revegetation should be computed to
see if it is potentially significant. Possibly some very localized
sampling would be of use in showing more precisely the nature of the
source connected with interstate highway. Containment systems may be
feasible, particularly if the source is related to the transport of
particulates from eroding areas or with heavy use of
phosphorus-containing fertilizers (this may decline with time). Quite
possibly this source could be cut in half, although the expense may be
considerable if erosion control is involved.
G. Ski slopes. The yields from ski slopes accounted for slightly
more than one percent of the phosphorus load in 1982. This is
-------
295
obviously not a major increment, but It Is still vulnerable to some
control because of its localized nature. As with interstate highway,
the prescription is difficult because exact mechanisms of increased
nutrient yield are difficult to pinpoint and in an all probability are
diverse. Fertilizer use, mechanical action of skiers on the slopes,
and erosion aggravated by construction may all play a part. Ponding of
runoff by a rationale similar to that which is used in urban areas is
one possibility and might conceivably reduce the phosphorus yield to
half its present amount.
One final possibility is the addition to Dillon of significant
amounts of water lower in phosphorus than the average runoff currently
entering the lake. This would in effect lower the mean phosphorus
concentration of entering water and thus lower the equilibrium
phosphorus concentrations in the lake. Although this alternative is
almost certainly not feasible solely as a mechanism for the control of
trophic status of Lake Dillon, it may be considered as a potential side
benefit to various water diversion projects, but only if those projects
involve the use of water having very low phosphorus concentrations.
-------
296
The Dillon Clean Lakes Model
A major purpose of the present study was to develop a model that
could assimilate information on changes in land use and from this
predict changes in the lake, and especially those changes of aesthetic
or economic importance. We have developed such a model from the
information on the lake and watershed nutrient yields. We will refer
to this model as the Dillon Clean Lakes Model.
The Dillon Clean Lakes Model consists of three parts, which are
shown in diagrammatic form in Figure 57- These are the land use
component, the trophic status component, and the effects component.
The Land-Use Component of the Model
The land-use component requires as input a matrix of land use
information. The land-use matrix specifies the intensity of different
types of use (on an area or population basis) for each of 19 watershed
segments (Figure 58). The land-use component also requires the amount
of gauged runoff (sun of the four U.S.G.S. gauges) for the year that is
being modelled, the amount of water to be pumped cr diverted from other
watersheds, and the mean concentration of this pumped or diverted water
on a nonth-by-month basis. The land-use component accepts as input the
water yield from ail watershed segments, but will compute these yields
from the total gauged runoff if the yields are not supplied.
The land use computations are organized on a segment-by-segment
basis. The nutrient export from each segment is computed from the
water yield and the land-use matrix by use of the equations that were
developed in the analysis of nutrient yields. The yields are sunmed
-------
CW
c
O rt
M M
TO C
(U O
'J '<
C
f '-I
fa ft,
f<"
n> o
en 11,
INPUT
Gauged runoff —
Water yield by segment^-
Amount of diversion by month—.
Mean concentration of
diversion by month
Land use matrix (by segments) -—"
MODEL
Lake level
Total runoff Input by month
Total outflow by month
Land use component
Amount of diversion
Gauged runoff
by month Annual P load
Trophic status component
Mean P cone July-Oct
Gauged runoff
CHLa
Effects component
Plot f)«i c k a go
OUTPUT
e
-*- Kg yield by segments, P & N
-*- Kg yield by sources, PAN
-»•- Loading by sources, PAN
-»- Total loading, PAN
Mean total P cone
Mean chlorophyll a
^ Secchi (mean, min)
*- AHOD
»- Minimum oxygen
*•• Carlson index
»- Plot % chango
-------
H-
OQ
Ln
CO
tn
(T)
OQ
0
c
en
(T>
O.
H-
a
b
Cb
fD
M
M
H-
3
OQ
miles
to
VO
00
-------
299
for all segments and the inventory change function is applied in order
to correct for changes in the phosphorus or nitrogen inventories in the
valley bottoms according to the amount of runoff. Other sources are
then added, including the three lakeside WWTP's, groundwater,
precipitation, diversions, and irregular construction activities.
Output from the land use component of the Dillon Clean Lakes Model
includes the nutrient yield by watershed segments and by source
categories for both P and N. The total annual P and N reaching the
lake is also part of the output.
The Trophic-Status Component of the Model
The land-use component of the Dillon Clean Lakes Model passes
certain critical information to the trophic status component of the
model. This information includes the monthly schedule of diversions or
pumping, the concentrations of phosphorus and nitrogen en a monthly
basis for pumped or diverted water, the annual phosphorus load by
runoff and precipitation, and the amount of gauged runoff (i.e., su~ of
the four U.S.G.S. gauges). Additional input required at this point but
not supplied by the land use component of the model includes the total
surface water and ground water entering the lake on a monthly basis
(this will be slightly larger than the gauged runoff), the monthly
average lake level, and the amount of water leaving the lake on a
monthly basis. The output includes the mean total phosphorus
concentration of the top 15 in during the period July through October
and the mean chlorophyll a_ in the top 5 m over the sar.e interval.
Because the trophic status of lakes has been the subject of nuch
-------
300
modelling effort, there is considerably more latitude in the choice of
approaches for the trophic-status component of the Dillon Clean Lakes
Model than for the land-use component. Whereas with the land-use
component there is little basis for anything other than a
simple empirical approach to modelling, the trophic-status component
could be based in a more complex and fundamental way on current
knowledge of biological, physical, and chemical processes in lakes. In
discussing the alternative approaches, we can recognize two different
classes of models: process models and empirical models.
A process model assumes that the critical biological, chemical,
and physical processes that have an important bearing on the variable
of interest (in this case, trophic status and its derivatives) are
sufficiently well known to be represented by equations. For example,
in a process model dealing with algal growth, logical functions to be
incorporated in the model would include nutrient uptake, sedimentation
cf algae through the metalimnion, and decomposition of algae to release
nutrients to the water. Since even the simplest natural planktonic
systems are exceedingly complex at the elementary functional level, any
such approach inevitably incorporates very large systems of coupled
equations. An example is the Lake George Model, CLEANER (Park et al.
1979).
Process models have value as exploratory tools for basic ecosystem
research, but their use for prediction, and thus for problem solving,
is probably unwarranted and ill-advised at present. Opinions are by no
means undivided en this subject, but a consensus seems to be developing
about the practical restrictions on process models (e.g., Tailing 1979,
-------
301
Harris 1980, Hobble and Tiwari 1977). The main reasons for the
limitations on usefulness of process models is that the critical
phenomena controlling biological variables in ecosystems are too poorly
understood, too labile, and too complex to be treated in this manner.
It is sometimes stated that the use of process models on biological
systems only awaits the accumulation of further basic information. The
implication is that the ecology of biological systems is in a
relatively primitive state and, as it matures, will logically support
complex models that produce satisfyingly exact answers to practical
problems. This line of reasoning is probably misleading in several
ways. First, the fund of information presently available is quite
vast, especially for planktonic systems. It seems unlikely that any
short-term improvement in this fund of information will drastically
alter the feasibility of process models. Secondly, other disciplines
in which the use of models to deal with large and complex systems has a
long history have encountered much the same problems in the use of
process models as ecosystem biology. Economics and meteorology are two
such disciplines. The motivations for predicting either the economy or
the weather well in advance are obviously enormous, and yet the success
of models to accomplish this is very modest; available models deal best
with short-range predictions that assume nothing unusual happens.
There exists in the study of ecosystems, and particularly of
lakes, an alternative to the process model. Lakes are so numerous that
empirical experience can be accumulated on their responses to widely
varying conditions. This experience can serve as the basis for models.
The philosophy of such models is entirely different from that of
-------
302
process models. The dependent variable of interest is examined from
the viewpoint of a very small number of master variables. While these
master variables are chosen for their known direct and indirect
relationships to the dependent variable, no attempt is made to dissect
the numerous separate functions that contribute to the overall
significance of such a master variable. This approach was first
popularized for lakes by Vollenweider (1968), who related trophic
status to phosphorus concentration and phosphorus concentration to
phosphorus loading and mean depth.
Empirical models inevitably evolve, taking on more complexity -
While the general validity of such a model can be established by
sampling large numbers of systems and comparing the observed to the
predicted values for the variables of interest, there will always be a
certain amount of scatter in the agreement of observations and
predictions. An obvious challenge then is to reduce the scatter by
introduction of new concepts involving the same master variables or
addition of other master variables. The model thus tends to become
more complex, and in principle converges with the process model by
building backward from the level of master variables to the level of
processes. It is doubtful that this convergence will ever be
completed, however, as there are severe practical limits to the
reduction of scatter in the relationship between observed and predicted
values. Even in their most complex forms, present modifications of the
Vcllenweider model are very distant in rationale and complexity from
a process model such as that developed for Lake George.
-------
303
For reasons outlined above, we have accepted as a philosophical
basis for the modelling of the trophic status of Lake Dillon an
empirical modelling approach. We have resolved to use a Vollenweider
type of model, but to consider modifications as appropriate in the
particular circumstances of Lake Dillon.
The first and most obvious improvement in the original
Vollenweider model, which was based solely on phosphorus loading and
mean depth, involved corrections for the flow of water through lakes.
Obviously two lakes with equal phosphorus loadings and mean depths but
very different flushing rates might easily differ in trophic status.
It seems intuitively obvious that a certain amount of the phosphorus
income should be discounted for losses through the outlet. This
discounting concept has been considered by a large number of students
of eutrophication (Welch 1980).
It is a short step from corrections based on flushing rate to
corrections based on measured phosphorus retention coefficients. This
approach has been taken by Larsen and Mercier (1976) and by Dillon and
Rigler (1974). Since phosphorus retention and flushing rate interact,
a number of formulations are possible, as described by Vollenweider et
al. (1980). Underlying all of these approaches is a general mass
balance equation by which a steady-state phosphorus concentration is
defined in terms of phosphorus gains and phosphorus losses. In a lake
such as Dillon, where flowthrough is substantial, there is little doubt
of the need to incorporate some kind of correction for the hydraulic
residence time, or the sedimentation coefficient for phosphorus, or
both of these.
-------
304
The approach to be taken here can be credited to Vollenweider
(1969), although it is a specific derivative of general mass balance
equations. Since this formulation was proposed by Vollenweider
subsequent to his well-known original formulation, which did not take
into account hydraulic residence time, we refer to it as the "modified"
Vollenweider model. The critical equation is as follows:
Cp = Lp/1T(l/tw + s) (16-1)
where C_ equals the phosphorus concentration of the lake, L_ is
phosphorus loading of the lake, ~z is the mean lake depth, s is the
phosphorus sedimentation coefficient and t^ is the residence time for
water (Vollenweider et al. 1980). Vollenweider was able to reduce the
number of parameters in the equation by assuming a sedimentation rate
of 10-20 m per year. The simplified equation thug became
Cp = Lptw/z-(l + tw°'5). (16-2)
Exact meaning of the model parameters is to some degree open to
definition by the user and is not fully standardized. For Lake Dillon,
we defined C as the mean phosphorus concentration of the upper 15 m
of the water column from the interval between July 1 and October 30.
The depth zone 0-15 m corresponds to the maximum extent of the growth
zone. Since changes in transparency and chlorophyll a in the upper
water column are ultimately of interest, this was considered to be the
most advisable choice for the definition of C_• However, had C
been defined in any other reasonable way, major differences would r.ot
have rosuited. It was considered advisable to define the time interval
for -determination of Cp in such a way that the spring runoff, which
incorporates an exceptionally large fraction of heavy particulate
-------
305
material that settles out rapidly, would not unduly influence C -
This prevents the model from being unacceptably sensitive to variations
between years in the amount of runoff.
The hydraulic residence time, t^, can also be defined in a
number of ways. In a lake that remains at steady state, all
*
definitions are essentially equivalent. In the case of a lake such as
Dillon, outflow and inflow are not always equal for a given year. For
present purposes we use as our definition of t^, the ratio of lake
volume to total inflow. This is partly necessitated by discrepancies
between the U.S.G.S. and Denver Water Department figures on inflow and
outflow, which have forced us to choose one or the other of these data
sources. For consistency, we use the U.S.G.S. inflow figures.
Use of the simplified formula (eq. 2) for the modified
Vollenweider model predicts P values that are reasonable but are
consistently too high for the two years of record. Since overestimates
occurred both in a low-water and a high-water year, the implication is
that the simplified formula underestimates the sedimentation
coefficients for Dillon, and that the performance of the model could be
improved by use of a more correct sedimentation coefficient, which in
turn necessitates use of the more complex version of the relationship
(equation 16-1). For this reason, sedimentation velocities (v. where v
= z"'s) were approximated according to the procedure outlined by
Higgins and Kim (1981). This procedure relies on knowledge of the
inflow and outflow concentrations. The resulting estimates of
sedimentation velocities for the two years were between 13 and 15 = per
year. These sedimentation velocities were used to compute values of s
-------
306
(T was 23 m both years), which were in turn used in the Vollenweider
equation (16-1). The results were better but still consistently
somewhat high for both years, implying that the actual sedimentation
coefficient was still being underestimated. Since the estimation
procedure from Higgins and Kim assumes a steady-state condition, which
is not ever literally applicable, the estimate is subject to unknown
error. Furthermore, the outflow concentrations could not be determined
directly for the water going through the Roberts Tunnel, which was not
available for sampling, and this may have resulted in some estimation
error.
Since the sedimentation coefficients were underestimated even
after correction by the procedure outlined by Higgins and Kim, another
approach was taken. Each year was treated independently and the
equation was solved for s using the observed values of the other
parameters for the year in question. The resulting estimates of
sedimentation velocity were 24 m/yr for 1981 and 32 m/yr for 1982. The
larger coefficient for 1982 is quite reasonable in view of the fact
that heavy particulates are moved in nuch larger quantity in years of
high runoff than in years of low runoff (see Figure 22a in the section
on lake phosphorus). The question then arises how the sedimentation
coefficients should be set for purposes of prediction in years of
different runoff. The positive relationship between sedimentation
coefficient and amount of runoff seems justified from the information
at hand but the exact form of the relationship cannot be determined
from two points. It is therefore assumed that the relationship is
linear. Nonlinearity will result in a certain amount of error, but the
-------
307
assumption of linearity is much better than assuming constancy of
sedimentation, which is clearly incorrect.
A second element of the trophic status component of the model is
the relationship between phosphorus and chlorophyll. A number of
statistically derived relationships are presented in the literature.
These show consistently a statistically significant trend toward higher
summer chlorophyll a_ with higher total phosphorus in the mixed layer.
There is, however, quite a bit of scatter and this has been the subject
of considerable discussion and analysis.
The general equations that are available in the literature (e.g.,
Dillon and Rigler 1974, Carlson 1977, Jones and Bachmann 1976, Jones
and Lee 1982, Lambou et al. 1982) do a uniformly poor job of predicting
Lake Dillon chlorophyll a. Observations are 3-6 times higher than
predicted values; Dillon produces considerably more chlorophyll per
unit of P than most lakes. This is unusual but not unique, as shown by
the scatter of points in the above-cited publications, but the
explanation is not immediately obvious.
A recent study by Smith (1982) appears to explain the unusual
phosphorus-chlorophyll relationship of Lake Dillon. Smith assembled
data from several surveys in which both the nitrogen and phosphorus
concentrations were known. He treated nitrogen as an added variable in
studying the relationship between phosphorus and chlorophyll. He
showed that nitrogen is a very important covariate, greatly increasing
the amount of variance that can be explained even when phosphorus
clearly limits growth rates. His conclusion is that increases in the
nitrogen to phosphorus ratio boost the amount of chlorophyll that will
-------
308
be produced for a given amount of phosphorus, even when phosphorus is
limiting. The mechanism by which this occurs is unknown, but the
reduction of variance by incorporation of nitrogen into the equations
is impressive. It is worth noting that OECD (1982) found no added
variance component explained by nitrogen. Smith and OECD did use
different data bases; possibly Smith's was better suited for
statistical illustration of the role of nitrogen. For present
purposes, we accept Smith's conclusions as valid and attempt to apply
then to Lake Dillon.
The lakes studied by Smith varied widely in the ratio of nitrogen
to phosphorus. Smith developed equations for predicting chlorophyll in
lakes of differing nitrogen to phosphorus ratios. His equations show
that the effect of increasing the nitrogen to phosphorus ratio is to
shift the line relating phosphorus to chlorophyll upward (Figure 59).
At low to moderate TN:TP ratios, the upward shift occurs without
changes in slope; the relationship thus takes the form of a family of
parallel lines. In the highest category used by Smith, however
(TN/TP > 35 by weight), the slope increases, and simultaneously becomes
insensitive to further change in the TN/TP ratio. For lakes with low
TN:TP ratios, it is essential to use N as a dependent variable as well
as p. For lakes with a high TNtTP ratio, N can be dropped from the
equation, but the coefficient on P will differ from that of lakes
having low TN:TP ratios.
LaV.3 Dillon, with TN/T? = 54 (weight basis), falls in the very
highest category of nitrogen to phosphorus ratios among the lakes
studied by Smith. In this highest category, as the influence of
-------
3C9
00
to
E
\
e
01
-I
o
0
O.I
DILLON RIGLER
LAKE DILLON-V
SMITH,
TN:TP=IO
SMITH,
TN:TP=25
10 100
TOTAL P, mg/m3
000
Figure 59. Log-log plot of total P and chlorophyll a_, showing lines
for two of Smith's (1982) lower TN:TP categories, the
Dillon-Rigler line, and the Lake Dillon line.
-------
310
further change in TN:TP becomes negligible, the coefficient on ? is
high and convergent with that of the Dillon-Rigler equation (1974).
This convergence is expected, since the Dillon-Rigler equation was
developed without inclusion of N. However, Smith's study shows that
the intercept may continue to be affected as TN:TP is raised to high
levels, even though the slope stabilizes. We thus approach the Dillon
data with the following points in mind: (1) the TN:TP ratio is so high
in Dillon that N need not be used in the equation, (2) the appropriate
slope can be taken from the Dillon-Rigler equation, (3) the intercept
should be determined uniquely for Dillon. We therefore set the slope
of the phosphorus-chlorophyll equation equal to that of the
Dillon-Rigler equation, but assume that the intercept will be
determined by the nitrogen to phosphorus ratio. We determine the
intercept empirically by solving for the intercept using the slope from
the Dillon-Rigler equation and the observed chlorophyll a values. The
resulting equation is as follows:
logB = 1.449 log(Cp) - 0.398 (3)
where B is chlorophyll a_ (July-October, ug/1, 0-5 in) and Cp is total P
(0-15 m, July-October, ug/1). This equation performs well for both
years (predictions will be given below).
Effects Component of the Model
Total phosphorus and chlorophyll a_, as predicted by the trophic
status component of the Lake Dillon Clean Lakes Model, have a number of
correlates of economic and aesthetic importance. Since public
attention and the attention of lake managers and users is likely to be
-------
311
focussed on some of these correlates, it is useful to be able to make
predictions for them.
Transparency is perhaps of greatest concern as a correlate of
eutrophication. This is the aspect of eutrophication most obvious to
the casual observer and most intimately connected with the aesthetic
appeal of a "lake. The most common measure of transparency is the
secchi depth. Since a great deal of information is available from the
literature on the relationship between secchi depth and chlorophyll a,
and since there is a complete set of secchi depth measurements for Lake
Dillon in 1981-1982, we have designed the effects component of the
model to predict secchi depth. The prediction is based on chlorophyll
a, which is the variable most directly responsible for changes in
transparency as eutrophication occurs. The interval of interest is
July 1 through October, corresponding with the post-runoff
stratification season. As in other instances, we have excluded the
month of June from the predictions because all predictions are
complicated in this month by the dominance of inorganic particulate
matter in the water column as a result of runoff. The observed mean
secchi depth over the July-October interval in 1981 was 3.15 m and in
1982 it was 2.76 m.
The literature offers a number of equations relating secchi depth
to chlorophyll a. There are two drawbacks to the use of these very
general equations on Lake Dillon. First, there is a great deal of
scatter around any of the lines taken from the literature, as would be
expected for the comparison of very different kinds of lakes.
Secondly, Lake Dillon is not located in the densest cluster of points
-------
si;
used for any of these equations, since its chlorophyll a_ values are
still relatively low. Thus the general equations, which for
transparency are particularly likely to go astray when extrapolated
from lakes of low transparency to those of high transparency,- may not
be especially appropriate for Lake Dillon.
The relationship between secchi depth and chlorophyll £ is always
nonlinear. Transparency as measured by secchi depth decreases at a
slower rate than chlorophyll increases. The fit of points to an
empirical transparency/chlorophyll a_ relationship is typically rendered
linear by log transformation. The log-transformed relationship thus
has a slope and an intercept. Problems related to the determination of
the slope and the intercept are different. The slope of the
relationship should be much more universal than the intercept. The
slope shows the increment of change in secchi depth for an increment of
change in chlorophyll a_, and lakes cannot be expected to differ widely
with respect to this relationship. It is thought that the package size
of chlorophyll a_ (i.e., the average cell size for phytoplankton)
results in different values of chlorophyll-specific extinction (Harris
1978), and this may be responsible for a certain amount of variation in
the slope. However, we have already shown that Lake Dillon is very
average with respect to its chlorophyll-specific extinction (es =
f)
0.015 m^/mg chlorophyll a_) . We therefore believe it is justified to
use a slope from one of the equations in the literature. We choose for
this purpose the slope of the equation of Carlson (0.68), which is
b.a^d on 147 lakes and has an r value of 0.93 (Carlson 1977). The
Mational Eutrophication Survey data have a somewhat higher slope (0.86:
-------
313
Lambou et al. 1982), but the observed values for Lake Dillon for the
two years of record suggested much closer agreement with the Carlson
coefficient than with the National Eutrophication Survey coefficient.
We therefore adopt the Carlson coefficient.
The intercept of the log-transformed relationship between secchi
«
depth and chlorophyll a_ is likely to vary much more extensively between
lakes than the slope, particularly if one considers reservoirs and
natural lakes together. The intercept for a given lake will be
determined largely by a combination of nonchlorophyll factors
contributing to the extinction of light. In very transparent lakes,
this includes principally the extinction effect of pure water and of
dissolved substances. In lakes that receive substantial amounts of
inorganic particulates, extinction by particulates becomes a major
determinant of the intercept.
It is feasible to customize the intercept for a prediction
equation to a given lake by using a slope from a series of lakes and
the observed values for the lake in question to solve for the
intercept. We did this separately for 1981 and 1982 on the Lake Dillon
data using the Carlson slope. The intercept for 1981 was 2.44 and Che
intercept for 1982 was 2.37. These intercepts were so similar chat a
common intercept of 2.4 was adopted for use with the Dillon data. This
intercept is higher than the intercept of the Carlson equation (2.04),
probably because the Carlson equation is based on natural lakes lacking
the inorganic particulate input that one sees in Lake Dillon. The
intercept is much closer to but slightly lower than the intercept for
-------
314
the National Eutrophicatlon Survey (2.56), which includes a large
number of reservoirs.
The equation resulting from the above analyses is as follows:
InSD = 2.4 - 0.681nB
where SD is secchi depth in meters and B is chlorophyll a_ in ug/1. The
prediction of 1981 and 1982 chlorophyll from this equation is excellent
(see below), suggesting that use of the Carlson slope is realistic.
A rough approximation can also be made of the minimum secchi depth
from the average secchi depth. The minimum secchi depth in this case
applies to the time of maximum phytoplankton standing stock. There is
another seasonal minimum associated with runoff, which is not of
concern here. In 1981 the ratio of minimum secchi depth to average
secchi depth was 0.76 and in 1982 it was 0.65. As an approximation we
take the ratio to be 0.70. There is no extensive literature on this
subject and the prediction is therefore less secure than the other
predictions mentioned up to this point. However, the use of this ratio
will give some indication of the transparency of the lake at that
particular time of the year when transparency problems may be most
aggravated (July).
Oxygen depletion in deep water during the summer is also of
concern in connection with eutrophication. In the section on oxygen,
we presented an analysis of the areal hypolimnetic oxygen deficit and
showed good agreement between the predicted AHOD and an equation
.eve loped by Cornett and Rigler (1980) based on secchi depth and mean
..epch. We use this equation in the effects component of the
Dillon Clean Lakes Model, but we believe it is feasible to improve the
-------
315
prediction by making a. correction for the amount of water entering the
lake in a given year. In 1981 the equation underpredicted the AHOD by
130 mg/m2/day (20%) and in 1982 it overpredicted by 6 mg/m2/day
(1%). The difference in the tvx> years is almost certainly due to the
different amounts of oxygen transported by advection to deep water in
the years of high runoff and low runoff. In a wet year, advective
transport is greater and the areal hypolimnetic oxygen deficit is thus
lower. We correct by making the assumption that runoff is proportional
to the differential between the observed and actual hypolimnetic oxygen
deficits. This results in a small correction that is added to the
model.
Although the AHOD is of some direct interest, the minimum oxygen
level in deep water is of more specific concern. There are no general
equations for prediction of minimum oxygen content, since this
prediction would involve the shape of the lake and the duration of the
stratification period. However, for a given lake, there is obviously a
very close relationship between the AHOD and the amount of oxygen
depletion. The relationship should be very close to linear if the
duration of the stratification period is constant. For Lake Dillon the
initial oxygen can safely be assumed equal to 9.0 mg/1, the saturation
concentration at spring overturn. The oxygen depletion is equal to the
difference between 9.0 mg/1 and the minimum oxygen observed toward the
end of stratification. We deal specifically with a point 5 m over the
bottom of the index station, and use the degree of depletion (mg/1) in
each of the two years as the dependent variable and the AHCD
(mg/m^/day) as the independent variable. This yields a slope of
-------
316
0.00698. This slope, when multiplied times the AHOD, gives the
predicted amount of depletion. The amount of oxygen remaining is then
the difference between the depletion and 9.0. The lower bound of
remaining oxygen is set to zero, even for depletion more than severe
enough to remove all oxygen.
Comparing Predictions and Observations for 1981 and 1982
The complete Dillon Clean Lakes Model was run on land use and
runoff data for 1981 and for 1982. In these runs, the unusual features
of 1981 or 1982 that would not be obvious from runoff or from land use
data of the type accepted by the model for prediction purposes were
ignored. For example, the output of wastewater treatment plants was
computed on the basis of the number of persons served and the
generalized equation described in the section on land use rather than
the known yields from these plants. Similarly, the construction in the
Snake River bottom near Keystone in 1982 was not treated in any special
way in the model for 1982. Thus the agreement between predictions and
observations for the two model runs gives an idea of the deviation
between observed and expected that could be produced by randomly
occurring events whose timing cannot be anticipated.
The predictions and observations are summarized in Table 61. For
all variables the agreement is within 5%, except for the minimum
deep-water oxygen, which is within 10%, The performance of the model
on theso years of known characteristics is thus excellent, despite the
occurrence of a number of events whose timing cannot be anticipated by
such a model.
-------
317
1981
Total P, ug/1*
Chlorophyll a, ug/1*
Mean secchi , m*
Minimum secchi , m
2
AHOD, mg/m /day
Minimum deepwater 0«
Predicted
6.7
6.3
3.2
2.2
706
4.1
Observed
7.0
6.7
3.2
2.4
710
4.4
1982
Predicted
7.1
6.9
3.0
2.1
590
4.9
Observed
7.4
7.3
2.8
1.8
63C
4.6
* Mean, July-October
Table 61. Comparison of predictions from the Dillon Clean Lakes Model
with observations for 1981 and 1982.
-------
313
Predictions for Development Scenarios
The Dillon Clean Lakes Model was applied to 10 different
scenarios provided by the Dillon Clean Lakes Steering Committee. We
report below the results of modelling for each scenario. The 10
scenarios comprise five sets of assumptions for development and land
use, each of which was applied to hydrologic assumptions for a dry year
and for a wet year. Predictions in each case include the total loading
of the lake, the source distribution of the loading, and the response
of the lake to the predicted amount of loading.
The dry-year hydrologic conditions were always set identical to
1981. Since the runoff for 1981 was among the lowest for the period of
record, 1981 provides a good estimate of the limiting hydrologic
conditions for a dry year. For the wet year, we used the hydrologic
conditions of 1982. 1982, although definitely above average, was not
so extreme hydrologically as was 1981. Use of the hydrologic
conditions of 1981 and 1982 is advantageous since the behavior of the
lake in these two years under present land use is very well documented
and can thus be compared easily to future years of similar hydrologic
conditions but very different land uses. We have also used the
observed lake level changes for 1981 and 1982 along with the hydrologic
data for these two years. Thus we have assumed that the schedule of
water release from the lake remains more or less as it is presently.
Predictions might be different if the Denver Water Department began to
release water in a very different way, e.g., by more excessive summer
drawdown or more extended periods of drawdown.
-------
319
The predictions of trophic indicators are translated into a
trophic status category. For this, certain conventions are necessary.
We use chlorophyll as the key trophic indicator, since algal biomass is
the central concern of changing trophic status. We set limits on the
trophic categories from a midrange of boundaries accepted by various
sources. According to the review by Welch (1980), the oligotrophic
boundary will be somewhere between 2 and 4 ug/1 and the eutrophic
boundary will be between 6-10 ug/1- We therefore set the mesotrophic
range for present purposes at 3-8 ug/1 chlorophyll a_ (this is also the
range used by OECD 1982). One other convention is the adoption of an
annual time frame for trophic classification. Trophic classes make no
sense at intervals less than one year, so this is the maximum
resolution possible. There is some merit to the argument that status
should be based on a run of years, or on a year of median conditions.
However, since we are interested in variation between years, we adopt
for modelling purposes an annual interval.
Subsequent to the presentation of model runs is a more generalized
graphical presentation based on the logic of the model.
Model Runs for Present Land Use Patterns
Before any of the scenario data were used, the model was run
twice (wet year/dry year) with a land use input matrix representing
present conditions. The purpose of these model runs was not to check
the behavior of the model, which has already been discussed, but rather
to define the loading and trophic condition of the lake under
standardized conditions subsequent to 1982. These nodel runs differ
-------
320
from the predictions discussed in previous sections in that Copper
Mountain is assumed to have transferred completely to tertiary
treatment, and to serve 2500 persons, slightly more than 1981-1982.
Climax Molybdenum is set at the 1981 employment level (2650 persons).
Also, no allowances are made for any of the special events that
occurred in 1981 and 1982. Thus these two model runs for present
conditions give for comparative purposes the output of the model under
the assumption that present land use stabilizes (by resumption of
activities at Climax and completion of the transfer of Copper Mountain
to tertiary) but that no growth occurs. The results of the two model
runs on stabilized present conditions are shown in Tables 62 and 63.
As expected, the predicted loading and response of the lake are very
similar to those actually observed in 1981 and 1982.
Scenarios 1A, IB; Low Growth
These scenarios specify that wasteload allocations for point
sources will be met, and that septic areas will be built out to 70%.
This means that 70% of the lots will be occupied by dwellings used in
the same way as present dwellings.
The point sources are at present meeting their wasteload
allocations, although there is some random variation from one year to
the next. We therefore leave constant at the 1981-1982 levels the
number of persons served by the Breckenridge, Snake River, and Frisco
wastewa:er treatment plants. We assume that Copper Mountain serves
2:iV persons, which is slightly above the 1981-1982 level, and that
complete tertiary treatment for Copper Mountain is put into effect. We
-------
*
Scenario Conditions
Wet Year (similar to 1982)
Current conditions***
Low Growth (1A)
I/)W Orowth with Diversions (2A)
High Growth (3A)
Hl};h Growth, Best Controls (4A)
High Growth, Cost-effective
Controls (5A)
Dry Year (similar to 1981)
Current conditions***
Low Growth (IB)
Low Growth with Diversions (2B)
High Growth (315)
Hij-,11 Growth, Best Controls (4P>)
High Growth, Cost-effective
Controls (5B)
Total P
Load , kg
4678
5049
50048
6389
3657
5001
2459
2618
47618
3190
2021
2596
Percent
WWTP*
17.4
16.2
1.7
12.8
22.3
16.3
30.0
28.3
1.6
23.2
36.6
28.5
Percent
Septic
8.2
14.7
1.5
31.9
0
20.4
6.6
12.0
0.7
27.2
0
16.7
Percent
Diversion
0
0
89.9
0
0
0
0
0
94.5
0
0
0
Percent
Background**
«
60.5
56.1
6.7
44.4
77.5
56.6
51.7
48.5
2.6
39.9
62.9
49.0
Percent
Other
13.9
13.0
0.2
10.9
0.2
6.7
11.7
11.2
0.6
9.7
0.5
5.8
* Includes pnrkapc- plants
A* Includes hack}*,round runoff and precipitation
*** Assumes Copper Mountain on tertiary, Climax at full employment 1981 levels.
Tabl
e 62. Results of the application of the Dillon Clean Ibices Model to scenario data
-------
Scenario conditions
Wet Year (similar to 1982)
Current conditions***
Low Growth (1A)
Low Crowth with Diversions (2A)
High Growth (3A)
High Crowth, Best Controls (4A)
High Growth, Cost-effective
Controls (5A)
Dry Year (similar to 1981)
Current conditions***
Low Crowth (IB)
Low Growth with Diversions (2B)
High Growth (3B)
High Growth, Best Controls (4B)
High Growth, Cost-effective
Controls (5B)
Lake Total
P, ug/1
7.3
7.9
57.
10.0
5.7
7.8
6.1
6.4
74.
7.9
5.0
6.4
Chlorophyll
ug/1
7.1
8.0
141.
11.2
5.0
7.9
5.4
6.0
206.
7.9
4.1
5.9
Mean
Secchi , m
2.9
2.7
0.38
2.1
3.7
2.7
3.5
3.3
0.29
2.7
4.2
3.3
Minimum
Secchi , m
2.0
1.9
0.27
1.5
2.6
1.9
2.4
2.3
0.21
1.9
4.8
3.0
Minimum
0 , mg/1
2
4.8
4.6
0.0
3.8
5.4
4.6
4.3
4.2
0.0
3.6
4.8
4.2
Trophic Status
Mesotrophic
Mesotrophic
Eutrophic
Eutrophic
Mesotrophic
Mesotrophic
Mesotrophic
Mesotrophic
Eutrophic
Mesotrophic
Mesotrophic
Mesotrophic
* Includes package plants
** Tnr.ludftS harkfrnimrl runoff and nrec \ nitat I o r\
*** Assumes Copper Mountain on tertiary, Climax at full employment 1981 levels.
Table 63. Results of the application of the Dillon Clean I-akes Model to scenario data.
CO
ho
to
-------
323
assume that Climax Molybdenum is operational with 2650 persons in the
work force. The number of persons on septic systems represented by 70%
buildout would be 3052 (full-time equivalent residents). These would
be distributed geographically by proportions identical to septic
distributions for 1981-1982. All other land uses remain unchanged.
The model was run first for a wet year (1A) and then for a dry
year (IB). The results are reported in Tables 62 and 63. Figures 60
and 61 give a graphical representation of the degree of change as
compared with present conditions. In the wet year, the predicted
loading is higher by slightly less than 10% as a result of low growth.
This causes an increase of chlorophyll to 8 ug/1. As explained
previously, we have taken the range of chlorophyll values between
3 ug/1 and 8 ug/1 as indicating a mesotrophic condition. Thus in a wet
year the low growth scenario pushes the chlorophyll a_ exactly to the
upper limit of the mesotrophic span in a wet year. There is an
accompanying decrease in transparency and a decrease in the minimum
oxygen at the bottom of the lake.
For the dry year, the degree of change in total loading is
smaller. Furthermore, the lake is much more toward the middle of the
mesotrophic range for chlorophyll and other trophic indicators during a
dry year, except for deepwater oxygen. Thus increasing the loading
slightly to the degree specified by the low-growth scenarios does not
bring the chlorophyll level so close to the upper boundary of the
mesotrophic range. For a dry year, the lake remains solidly
raesotrophic.
-------
324
WET YR LO GRQWTH(lfi) VS PRESENT
100
CJ
Q£
UJ
0_
UJ
to
60-
2C+
-20--
UJ
UJ
Q_
UJ
UJ
cn
-so-
-JOO
0 l.OCO
TOTflL P
-f-
2.000
CHLfi
-t-
3.000
SECCHI
4.000 5.000
BOTTOM 02
Figure 60. Graphical summary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
DRY YR LO GROWTH!IB) VS PRESENT
100
UJ
(_)
a:
UJ
o_
*
LJ
V3
ce
o
3
SO-
20-
-20-
UJ
tt:
LU
0-
»
ce
UJ
UJ
CD
-so-
-100
•4-
0 1.000
TOTflL P
2.000
CHLfl
3.000
SECCHI
4.000 5.CCO
BOTTOM 02
Figure 61.
Graphical summary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
326
Scenarios 2A, 23; Low Growth with Diversions
In these scenarios it is assumed that 10,000 acre-feet/year is
diverted into Lake Dillon from Straight Creek and 183,000
acre-feet/year is diverted into Dillon by the Eagle-Colorado Project.
The total phosphorus concentrations are set at 30 ug/1 for Straight
Creek and to 200 ug/1 for Eagle-Colorado. The amounts of water
diverted are assumed to apply to years of average or above-average
moisture. For the dry-year run of the model (2B), the amounts of
diversion are scaled down in proportion to the gauged runoff (sum of
the four U.S.G.S. gauges in the watershed). The diverted water is
assumed to enter the lake on a schedule identical to the Tenmile Creek
hydrograph. All other conditions are set identical to scenarios 1A and
IB.
The results of the two model runs (2A, 2B) are summarized in
Tables 62 and 63. Figures 62 and 63 give a graphical representation.
The phosphorus loading coming in by way of diversion is roughly ten
times the loading under present conditions without diversions. The
lake shows drastic increases in total phosphorus concentrations and in
chlorophyll levels. Complete oxygen depletion is predicted for the
hypolimnion, and predicted secchi depth values are extremely low. The
lake under these circumstances would be unequivocally upper eutrophic
(sometimes called hypereutrophic). Nuisance blooms would be almost a
certainty under these conditions.
The chlorophyll levels predicted for the diversion scenarios are
so high that they probably exceed the theoretical maximum chlorophyll
possible in freshwaters. About 300 mg/ni2 of chlorophyll a absorbs
-------
327
WET YR 10 GROWTH W DIVERSION!2fl) VS PRESEN
too
UJ
O
to
01
O
60-
20-
-20-
UJ
LU
Q_
LU
CD
-60-
-100
0 1.000
TOTflL P
2.000
CHLfl
3.000
SECCHI
4.000 5.000
BOTTOM 02
Figure 62. Graphical summary of Pillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines vit'-.
"+" nark the oligotrophic and eutrophic boundaries for
chlorophyll a_.
-------
328
DRY YR LO GROWTH W 01 VERSIONS^23) VS PRESENT
10£h
LJ
CJ
o:
UJ
Q_
UJ
CO
cz
o
SO-
-20-
LU
Qi
UJ
UJ
CD
-SO-
-100
-4-
0 1.000
TOTflL P
2.000
CHLR
3.000
SECCHI
•4-
4.000 5.000
BOTTOM 02
Figure 63. Graphical summary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
329
99% of the light (Margalef 1978, Tailing 1982). At this point light,
rather than nutrients, becomes limiting and further addition of
nutrients does not boost algal biomass. In the upper 5 m of Lake
Dillon, to which algal growth would be essentially confined under the
low transparency conditions of these scenarios, we could expect to find
at a maximum about 60 ug/1 of chlorophyll a_, which would account for
n
300 mg/m* summed up over the 5-m mixed layer. Thus the chlorophyll
predictions, which are in excess of 100 ug/1, should not be taken too
literally. They merely indicate that the lake reaches the theoretical
maximum biomass possible under the light limitation conditions
prevailing in Lake Dillon. The exact concentrations are of course of
little interest once they reach such high levels.
The lake under the conditions of scenarios 2A and 2B would have
the qualities of a sewage lagoon. The lake would thus differ
drastically from its present appearance and would be very different
biologically because of the extreme depletion of oxygen in deep water
and very large crops of algae.
The phosphorus concentrations in the diversion water are the cause
of the extreme trophic response of the lake to diversion. Accurate
knowledge of the total phosphorus concentrations in the diversion water
is absolutely essential if accurate predictions are to be made. For
scenarios 2A and 2B, by far the greatest uncertainty of the prediction
has to do with the phosphorus values that are fed into the model for
the diversion water. The concentrations were supplied by Ton Elncre
and were derived by him from existing measurements not taken in
connection with the Dillon Clean Lakes Study- Although the present
-------
330
modelling exercise certainly shows that the worst-case situation for
phosphorus in diversion water would result in a very undesirable
degradation of the lake, sound predictions beyond this will require
extensive and careful study of total phosphorus in diversion water.
Scenarios 3A, 3B: High Growth
These scenarios assume that current wasteload allocations are met
for the wastewater treatment plants. This implies essentially no
expansion beyond scenarios 1A, IB of the number of persons served by
any one of the four wastewater treatment plants. The number of persons
served by these plants is therefore allowed to remain as for scenarios
1A, IB and Climax Molybdenum is assumed to operate with a workforce of
2650 persons. The high-growth scenarios assume that septic systems
will serve a peak population of 37,000 persons. This corresponds to a
time-weighted average of 8409 full-time equivalent residents. It is
assumed that this number of persons is distributed within the watershed
according to the same pattern of present septic system users. However,
it should be noted that this number of persons cannot be housed on the
existing developments served by septic, even if all of these
developments are built out to 100%. Thus these scenarios assume that
additional developments will be built and served by septic systems, or
that the present developments will have a higher density of housing, or
some combination of these possibilities
Tables 62 and 63 and Figures 64 and 65 summarize the results of
tha -cdel runs for high growth in a wet year (3A) and in a dry year
(3B). For the wet year, the scenarios indicate an increase in total
-------
100
WET YR HI GROWTH (3fl) VS PRESENT
CJ
ce
UJ
o_
• «
UJ
VJ
SO-
20-
-20-
LU
o
0£
UJ
Q_
os:
UJ
CD
-60-
•4-
0 1.000
TOTflL P
2.000
CHLfi
3.000
SECCHI
4.000
BOTTOM
5.000
02
Figure 64. Graphical summary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
322
DRY YR HI GROHTH(3B) VS PRESEN'
100
UJ
CJ
as
UJ
Q_
*
UJ
to
so-
20+
-2C--
UJ
O
Oi
UJ
0-
*k
c:
UJ
UJ
03
-SO--
-100
0 1.000
TOTflL P
•4-
2.000
CHLfl
3.000
SECCHI
4.000 S.OCO
BOTTOM 02
Figure 65. Graphical summary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
333
wet-year loading of approximately 50% over present conditions. There
is a drastic increase in the percentage of loading caused by septic
systems, as would be expected. Chlorophyll is predicted at 11.2 ug/1,
which is well over the boundary from mesotrophic to eutrophic.
Transparencies and minimum oxygen in deep water similarly reflect a
higher trophic status. Although not nearly so drastic as the changes
of scenarios 2A and 2B, the changes of scenario 3A with respect to the
present condition would almost certainly be significant enough for a
casual observer to notice.
For the dry year, the effects of high growth are of smaller
magnitude but still significant. Chlorophyll is predicted to fall at
7.9 ug/1, just below the upper mesotrophic boundary at 8 ug/1. Thus
the lake could probably be classified in the very dryest of years as
mesotrophic, but at all other times would be eutrophic, and in wet
years would be especially so.
Scenarios 4A, 4B; High Growth with State of the Art Controls
This scenario specifies that wasteload allocations are met by
point sources, and that no diversions enter the lake. All sources
except the wastewater treatment plants, package plants, groundwater,
precipitation, and background are set to 0. Thus it is assuned that
all nonpoint sources are completely eliminated.
The results of the model runs are shown in Tables 62 and 63 for a
wet year (^A) and for a dry year (4B). Figures 66 and 67 show the
corresponding graphs. In either a wet or a dry year, the model
predicts significant improvement of the trophic status of the lake,
-------
334
WET YR HI GRO STfiTE OF THE flRT(4fl) VS PRESEN'
lOOi ——
UJ
o
ce
LU
Q_
»
UJ
CO
60--
20--
-20-
UJ
C_)
ct:
UJ
o:
UJ
UJ
CD
-sof
-100
•4-
•4-
0 1.000
TOTflL P
2.000
CHLfi
3.000
SECCHI
4.000 5.000
BOTTOM 02
7igure 66. Graphical summary of Dillon Clean Lakes Model outpucs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
335
YR HI GROWTH STflTE OF THE flRTUB) VS PRESENT
UJ
u
or
UJ
Q_
UJ *
V3
OC
O
sc
-20
ce
UJ
Q_
o:
UJ
UJ
CD
-60
-100
•4-
0 1.000
TOTflL P
2.000
CHLR
3.000
SECCHI
4.000
BOTTOM
s.coo
02
Figure 67- Graphical summary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"T" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
336
despite high growth. Chlorophyll values show lower mesotrophic status
for the lake under this scenario. Transparency, bottom oxygen, and
other indicators are correspondingly improved.
As attractive as this scenario seems, it is almost certainly
unrealistic. It implies that there would be no yield whatever above
background for septic systems, for Climax Molybdenum, and for the many
distributed sources of nutrients associated with human activity. The
high-growth conditions imply that about 8000 persons (annual average)
will be served by septic systems. This is essential because the
wasteload allocations cannot be held constant if additional population
is to be served on sewer. Thus a substantial septic contribution or a
change in the wasteload allocation seems inevitable as an accompaniment
to high growth, but scenarios 4A and 43 allow neither- The results
should therefore be viewed as representing an unrealistically high
degree of control over phosphorus loading, regardless of financial
considerations.
Scenarios 5A, 53: High Growth with Cost-Effective Controls
Scenarios 5A and 53 are intended as counterparts to scenarios 4A
and 4B, but with controls set at levels that are more likely to be
realized. The assumptions are identical to those of scenario AA and
4B, except that all nonpoint sources, which were set to 0 in 4A and 4B,
are allowed to assume half their uncontrolled value. The results
appear :'n Table 62 and 63 and Figures 68 and 69.
ioc a wet year, the model predicts loading that is slightly more
than 5% above present conditions. This causes a slight increase in the
-------
337
WET YR HI GRO COST EFFECTIVE CONTR(Sfl) VS PRESEN
lOCh
LU
CJ
ct:
UJ
D_
UJ
CO
CC
o
60--
20-
-20-
UJ
o
a:
UJ
a_
*
ce
UJ
UJ
in
-60--
-100
0 1.000
TOTflL P
2.000
CHLfl
3.000
SECCHI
4.000 5.000
BOTTOM 02
Figure 68. Graphical summary of Dillon Clean Lakes Model outputs fr.
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines wit:
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
338
DRY YR HI GRG COST EFFECTIVE CONT(53) VS PRESENT
UJ
O
a:
LU
a.
UJ
03
ce:
a
SO--
20f
-20f
2:
UJ
CJ
a:
UJ
a.
&
cc.
UJ
UJ
C3
-50-
-ICO
0 1.000
TOTflL P
2.000
CHIP
3.000
SECCHI
4.000 5.000
BOTTOM 02
Figure 69. Graphical sunmary of Dillon Clean Lakes Model outputs for
four key lake characteristics. Vertical lines show
deviation from present conditions; horizontal lines with
"+" mark the oligotrophic and eutrophic boundaries for
chlorophyll a.
-------
339
expected chlorophyll level, which reaches 7.9, very close to the upper
mesotrophic boundary at 8.0. For a dry year, there would also be a
slight increase in chlorophyll above the present conditions, but the
dry-year increase would not bring the chlorophyll so close to the upper
mesotrophic boundary. Thus the model shows that, if the controls could
*
be accomplished, the lake would in most years fall in the mesotrophic
range.
Although scenarios 5A and 5B are labelled as cost-effective, they
may actually imply intolerable expenditures. For example, the
scenarios assume that phosphorus yield from the expected 8000 persons
on septic systems is somehow reduced by 50%, and they assume the sane
for Climax Molybdenum. They also assume that such diffuse sources as
the interstate highway and ski slopes can be treated in such a manner
as to reduce yields by one-half. Thus the realism of this scenario is
open to some skepticism, but it certainly approaches the reality of
maximum-investment controls much more closely than scenarios 4A and
4B.
Generalized Graphical Predictions
A large number of relationships can be developed in graphical forn
using the logic that is inherent in the model. These graphical
relationships are necessarily less specific in their predictive
capabilities than actual model runs, since simplifying assumptions are
required for a generalized graph that would not be required for a nodel
run. Nevertheless, generalized graphs give an overall impression of
the sensitivity of the lake and its watershed to development of
-------
340
different kinds. We have developed for a graphical presentation three
kinds of relationships: 1) the relationship between total phosphorus
loading and chlorophyll, 2) the relationship between number of persons
on sewer and total phosphorus loading, and 3) the relationship between
the number of persons on septic system and total phosphorus loading.
Figure 70 shows the general relationship between total phosphorus
loading and chlorophyll concentration. As in the modelling exercises,
the chlorophyll concentration referred to in the graph is the average
for the 0-5 meter layer over the post-runoff stratification interval.
The relationship of chlorophyll to phosphorus loading is depicted for
two different sets of conditions. The first set of conditions is based
on the 1981 study year. In the preparation of this curve it was
assumed that the runoff, water level, and lake volume changes were
identical to those of 1981. The second curve is based on 1982 and
involves similar assumptions for that year. Since 1981 was a very dry
year and 1982 was a year of above-average but not extreme wetness, a
median year would fall between the two curves but nearer to the 1982
curve than the 1981 curve. Changes in lake operation (water levels)
for either of these conditions would change the positions of the
curves, but the curves would not move dramatically unless there were
very significant departures from the present mode of lake operation.
Figure 70 indicates along with the two curves the lower limit of
loading for the two hydrologic conditions. These represent the natural
background for the two years depicted in the graph. Also indicated in
Fi:r:rs 70 are the observed loadings in 1981 and 1982 and the trophic
-------
co
c
1-1
O)
I—• T3 cn >•
vo £J" fl>
M cn tn r«
T3 O &1
(u :r 3 TO
0O ID
COO
H en | D*
vo o> M
oo |-™* o
N> O 0 fl
• (11 x—' O
O- T)
H- H- y
OQ
g
Cu
n>
t-h
o
O
O. O
O
O W
W
H- CO
HI 3
n e: o
o 3 HI
3 o HI
O. rt
H- H- in
rt O ft
H- 3 i-l
O (a
Hi
tn rt h1-
H- o O
E;l rt 0)
H- (1) (t
(_• I--1 H-
Co O
M TJ
rt
O
39
36
33
30
37
0
:24
o»
'21
O
o 15
9
6
3
0
4000
Total P load - Ib/yr
8000 12000 16000
18000
Wet year
(1982)
Bsr.^ _ /^ Eutr°phlc
^-Observed load f
Dry year / .^ '^82 Mesotrophic
bac^k^roupd ^^_ I
-Wet year background Oligotrophic
0 1000 3000 5000 7000
Total P loading - kg/yr
9000
-------
classifications that correspond to various chlorophyll levels. It is
clear from the graph, as it was from the modelling exercises, that
minor increases in loading either in a dry or wet year will push the
lake from the upper mesotrophic category where it presently lies into
the eutrophic category.
Figure 71 shows the relationship of number of persons on sewer to
the phosphorus loading in kilograms per year. Number of persons is
annual full-time equivalent (1 full time equivalent equals ca. 100
gallons per day effluent). As with Figure 70, the graph includes two
curves, one of which is based on conditions identical to 1981 and the
second on conditions identical to 1982. The intersection of each of
these lines with the y-axis represents the phosphorus loading to the
lake if there were no contributions whatever from sewered areas. This
can be considered a sort of non-sewer background for the watershed.
Contributions of sewered areas above this background are then added on
the basis of contributions of sewered areas through point and nonpoint
sources. Point sources are assumed to consist of waste water treatment
plants utilizing tertiary treatment. Tertiary treatment is assumed to
function at the same overall efficiency as observed in 1981-1982. In
addition to this point-source contribution, the non-point source
contribution corresponding to urban area on sewer is also added to get
the total per capita contribution from sewered areas. Since the
nonpoint contribution is dependent on amount of runoff, the nonpoint
contributions for years of different wetness are different, but in both
cases nonpoint contributions were below 10% of the point source
contributions.
-------
t—i rt to
vo H a
GO fD a.
H' to
I rt to
h~* 0 0)
vo n> en
oo a c
tj rt B
ft) to
HI rt
to to
ex, ru
o (D
a M
I o.
•o
o n
H- o
a a
rt Co
rt
n to
§s
n
a "0
o o
V<* )-j.
3
C en
to O
M C
rt O
O ft)
(O
rt o
O H-
10 0>
n>
H rt
< (D
ft) >-t
0- rt
H-
n.
c
M
H-
00
cr H
c a1
rt fD
p-
a ^
W O
to fo'
CO f)
CO rt
O H-
n o
H- 3
to co
rt
m H-
o. a
o
C 1 '
H- C
rt O.
rr m
in cr
(D O
*I rt
fo :r
H
ro rt
o- rr
ID
00
C
t-t
(D
"O ?3
a* (o
o H'
en to
_ rt
a* H-
o o
ss
o) ^r
MT)
o
to o
o. H,
H-
a a
OP c
a
O D'
MI ro
to o
?i* Hi
m
O ro
M Co
M O
O 3
3 en
P- O
Ml 3
O 01
rt fD
:r <;
ID n
ft)
cu
in
O O
to
Cu rt
P- 3*
a fD
oo
rt
rn o
O rt
f. ft)
f1 t—'
O
tfi
o 5000
0)
S 4000
ex
en
_v:
i
T3
u
o
o_ 2000-
1000-
20 30 40 50 60 70 80 90
THOUSANDS OF PERSONS ON SEWER
200OO
18000
Wet year
(1982)
Observed load
1982
Observed load
1981
0
100
a
a.
(/>
.a
i
CO
-F-
UJ
-------
344
The observed loadings for 1981 and 1982 are indicated on Figure 71
for reference purposes. The observed loadings correspond to
populations on sewer of about 20,000. The actual number of persons on
sewer in 1981 and 1982 was actually closer to 15,000. The observed
values correspond to a slightly higher population because part of the
population served by sewer (Copper Mountain) was not receiving tertiary
treatment in 1981-1982. In the projections, it is assumed that all
sewered areas are receiving tertiary treatment, which should be true
for future years.
Figure 72 has a similar rationale to that of Figure 71 except that
phosphorus loading is related to contributions from septic systens
rather than sewers. Lines representing dry year and wet year
conditions differ in slope because of the higher yield of phosphorus
from septic sources when runoff is high. In addition, the inventory
change function dictates that a larger percentage of the phosphorus
originating in septic areas will reach the lake in a wet year than in a
dry year. It should be noted that the loading scale (y-axis) covers
three times the range of the scale in Figure 71 for sewer. Observed
loadings in 1981 and 1982 are only a small fraction of the total
potential loading if the population on septic were drastically
increased.
Overview of Scenario Studies
The modelling and graphic studies indicate that Lake Dillon will
r.ove into the eutrophic category if diversion water rich in phosphorus
is added in quantity to the lake or if high growth occurs without the
adoption of nonpoint source controls cr other measures not now in
-------
oq
c
a* n rt pd
to o o m
cn 3 rt M
n cn to to
CX rt M rt
to H-
O 3 T) O
3 rt 3' 3
O cn
to to cn 3*
< rt -a H-
fl> 3" X)
1 it O
to 3- M O
OQ fD C if>
<0 H- tn
t! (-• O
to (-< O T>
It vo CO C
fD 00 C^ H4
1 !-• H- to
3-0 tS H.
•-* IJ uw r**
(D r< O
&• to 3
M cn cn
1 vo cn
g oo c cn
P to Q ft>
O p- i-i
i-h M 3 <
ni n> CTO n>
< ex
M> (D O
O M rt cr1
M cn 3- •<
• ID
OO H h-"U
I-1 3" O rt
fD CU M*
cu a. n
3 -a H-
CX i-1 3 in
o cro "^
M LJ. tn cn
oo o cu
o
3
fD
B
fD tn
2700O
24000 -
21000 -
18000 -
>> 15000-
O)
Q.
cn|2000
9000-
6000-
2000-
0
Wet year
(1982)
Observed
t load
1982
tObserved load
1981
10 20
rt> o
p-
30 40 50 60 70 80
Thousands of persons on septic
60000
54000
'-J48000
- 42000
- 36000
- 30000
*•
- 24000
- 18000
- 12000
-6000
90 100
2
'
•a
o
U)
4>
Ui
-------
246
practice to reduce the phosphorus loading of the lake. Under low
growth or high growth with additional controls, the condition of the
lake could be held within the mesotrophic range, but might suffer some
trophic degradation or remain more or less the same, depending on the
exact conditions.
-------
Literature Cited
Alabaster, J.S., and R. Lloyd. 1980. Water Quality Criteria for Freshwater
Fish. Butterworth, Boston. 297 pp.
American Public Health Assn. 1976. Standard Methods for the Examination of
Water and Wastewater. American Public Health Assn. Washington, D.C.
1193 pp.
Bailey-Watts, A.E. 1978. A nine year study of the phytoplankton of the
eutrophic and non-stratifying Loch Leven (Kinross, Scotland). J. Ecol.
66: 741-771.
Bendschneider, K., and R.J. Robinson. 1952. A new spectrophotometric method
for the determination of nitrate in sea water- J. Mar. Res. 11: 87-96.
Bormann, F.H., G.E. Likens and J.S. Eaton. 1969. Biotic regulation of
particulate and solution losses from a forest ecosystem. BioScience
19: 600-610.
Bormann, F.H. and G.E. Likens. 1979. Pattern and Process in a Forested
Ecosystem. Springer-Verlag, New York. 253 pp.
Carlson, R.E. 1977. A trophic state index for lakes. LLmnol. Oceanogr. 22:
361-368.
Cornett, J.R. and F.H. Rigler. 1980. Areal hypolimnetic oxygen deficit: Ar.
empirical test of the model. Limnol. Oceanogr- 25: 672-679.
Dillon, P.J. and F.H. Rigler. 1974. The phosphorus-chlorophyll relationship
in lakes. Limnol. Oceanogr. 19: 767-73.
Ednondson, W.T. 1972. Nutrients and phytoplankton in Lake Washington, pp.
172-188 _In_ G.E. Likens (ed.) Nutrients and Eutrophication. American
Society of Limnology and Oceanography Special Synposia, Vol. 1.
-------
348
Fee, E.J. 1976. The vertical and seasonal distribution of chlorophyll in
lakes of the Experimental Lakes Area, northwestern Ontario:
implications for primary production estimates. Limnol. Oceanogr. 21:
767-783.
Fogg, G.E., W.D.P. Stewart, P- Fay, and A.E. Walsby. 1973. The Bluegreen
Algae. Academic Press, N.Y. 459 pp.
Fogg, G.E. 1975. Algal Cultures and Phytoplankton Ecology. Univ. of
Wisconsin Press, Madison. 175 pp.
Ganf, G.G. 1975. Photosynthetic production and irradiance-photosynthesis
relationships of the photoplankton from a shallow equatorial lake (Lake
George, Uganda). Oecologia 18: 165-183.
Gibson, C.E. 1978. Field and laboratory observations on the temporal and
spatial variation of carbohydrate content in planktonic blue-green
algae in Lough Neagh, Northern Ireland. J. Ecol. 66: 97-115.
Glooschenko, W.A., J.E. Moore, and R.A. Vollenweider. 1973. Chlorophyll a
distribution in Lake Huron and its relationship to primary
productivity. Great Lakes Res. 16: 40-49.
Goldman, C.R. 1974. Eutrophication of Lake Tahoe emphasizing water quality.
EPA Report 660/3-74-034. 408 pp.
Golterman, H.L. (ed.). 1976. Interactions between Sediments and Fresh Water.
Junk, The Hague. 473 pp.
Grant, M.C. and W.M. Lewis, Jr. 1982. Precipitation chemistry in the
Colorado Rockies. Tellus 34: 74-88.
Grasshoff, K. 1976. Methods of Seawater Analysis. Verlag Chenie, Weinhei=.
317 pp.
-------
249
Harris, G.P. 1978. Photosynthesis, productivity, and growth: the
physiological ecology phytoplankton. Arch. Hydrobiol. Beih. Ergebn.
Limnol. 10: 1-171.
Harris, G.P. 1980. Temporal and spatial scales in phytoplankton ecology.
Can. J. Fish. Aquat. Sci. 37: 877-900.
Heaney, S.I., and J.F. Tailing. 1980. Dynamic aspects of dinoflagellate
distribution patterns in a small productive lake. J. Ecol. 68: 75-94.
Higgins, J.M. and B.R. Kim. 1981. Phosphorus retention models for Tennessee
Valley Authority reservoirs. Water Resour. Res. 17: 571-576.
Robbie, J.E. and J.L. Tiwari. 1977. Ecosystem models vs. biological reality:
Experiences in the systems analysis of an arctic pond. Verh. Internat.
Verein. Limnol. 20: 105-109.
Hutchinson, G.F. 1938. On the relation between oxygen deficit and the
productivity and typology of lakes. Int. Rev. ges. Hydrobiol. Hydrogr -
36: 336-355.
Hutchinson, G.E. 1957. A Treatise on Limnology. Vol. I. Geography, Physics,
and Chemistry. Wiley, New York. 1015 pp.
Hutchinson, G.E. 1967. A Treatise on Limnology, Vol. II. Introduction to
Lake Biology and the Limnoplankton. Wiley; New York. 1115 pp.
Jones, A.R. and G.F- Lee. 1982. Recent advances in assessing impact of
phosphorus loads on eutrophic-related water quality. Water Res. 16:
503-515.
Jones, J.R. and R.W. Bachmann. 1976. Prediction of phosphorus and
chlorophyll levels in lakes. J. Water. Pollut. Cont. Fed. 48:
2176-2182.
-------
350
Kirk, J.T.O. 1975. A theoretical analysis of the contribution of algal cells
to the alteration of light within natural waters. New Phytol. 75:
11-20.
Komarek, J. 1974. The morphology and taxonomy of crucigenoid algae
(Scenedesmaceae Chlorococcales). Arch. Protistenk. 116: 1-75.
Latnbou, V.W., S.C. Hern, W.D. Taylor, and L.R. Williams. 1982. Chlorophyll,
phosphorus, secchi disk and trophic state. Water Resources Bulletin.
18: 807-813.
Larson, D.P. and H.T. Mercier. 1976. Phosphorus retention capacity of lakes.
J. Fish. Res. Bd. Canada 33: 1742-50.
Lasenby, D.C. 1975. Development of oxygen deficits in 14 southern Ontario
lakes. Limnol. Oceanogr. 20: 993-999.
Lean, D.R.S., and F.R. Pick. 1981. Photosynthetic response of lake plankton
to nutrient enrichment: a test of nutrient limitation. Linmol. Oceanogr
26: 1001-1019.
Leopold, L.3., M.G. Wolman, and J.p. Miller- 1964. Fluvial Processes in
Geomorphology. W.H. Freeman, San Francisco. 522 pp.
Levine, S.N., and D.W. Schindler. 1980. Radiochetaical analysis of
orthophosphate concentrations and seasonal changes in the flux of
orthophosphate to seston in Canadian Shield lakes. Can. J. Fish. Aquat.
Sci. 37: 479-437.
Lewis, W.M., Jr. 1974. Primary Production in the plankton community of a
tropical lake. Ecol. Konogr. 44: 377-409.
Lewis, W.M., Jr., 1980. Comparison of spatial and temporal variations in the
zoo plankton of a lake by means of variance components. Ecology 59:
666-671.
-------
351
Lewis, W.M., Jr., and D. Canfield. 1977- Dissolved organic carbon in some
dark Venezuelan waters and a revised equation for spectrophototnetric
determination of dissolved organic carbon. Arch. Hydrobiol. 79:
441-445.
Lewis, W.M., Jr. and F.H. Weibezahn. 1976. Chemistry, energy flow, and
community structure in some Venezuelan fresh waters. Arch. Hydrobiol.
Suppl. 50(2/3): 145-207.
Lewis, W.M., Jr. and M.C. Grant. 1978. Sampling and chemical interpretations
of precipitation for mass balance studies. Water Resour. Res. 14:
1098-1104.
Lewis, W.M., Jr. and M.C. Grant. 1979. Relationships between stream
discharge and yield of dissolved substances from a Colorado mountain
watershed. Soil Science 128: 253-363.
Lewis, W.M., Jr. and M.C. Grant. 1980a. Change in the output of ions from a
watershed as a result of acidification of precipitation. Ecology 60:
1093-1097-
Lewis, W.M., Jr. and M.C. Grant. 1980b. Relationship between snow cover and
winter losses of dissolved substances from a mountain watershed. Arct.
Alp. Res. 12(1): 11-17.
Lewis, W.M., Jr., and J.F. Saunders, III. 1979. Two new integrating samplers
for zooplankton, phytoplankton, and water chemistry. Arch. Hydrobiol.
85: 245-249.
Likens, G.E. 1975. Primary production of inland aquatic ecosystens. l£ H-
Lieth and R.H. Whittaker (eds.) The Primary Productivity of the
Biosphere. Springer-Verlag, New York.
-------
352
Likens, G.E. F.H. Bormann, R.S. Pierce, J.S. Eaton, and N.M. Johnson. 1977.
Biogeochenistry of a Forested Ecosystem. Springer-Verlag, New York. 146
pp.
Lund, J.W.G. 1961. The periodicity of U algae in three English lakes. Verh.
Internat. Verein. Limnol. 14: 147-154.
Lund, J.W.G. 1965. The ecology of freshwater phytoplankton. Biol. Rev. 40:
231-293.
Lund, J.W.G. 1970. Primary Production. Froc. Soc. Wat. Treat. Ex an. 19:
332-358.
Lund, J.W.G. 1971. The seasonal periodicity of three planktonic desmids in
Windermere. Mitt. Internat. Verein. Liamol. 19: 3-25.
Lund, J.W.G. 1979. Changes in the phytoplankton of an English lake,
1945-1977. Hydrobiol. Jour. 14: 6-21.
Margalef, R. 1978. Life-forms of phytoplanktcn as survival alternatives in
an unstable environment. Oceanol. Acta 1: 493-509.
Mortimer, C.H. 1941. The exchange of dissolved substances between mud and
water. J. Ecol. 29: 280-329.
Murphy. J. and J.P. Riley. 1962. A modified single solution method for the
determination of phosphate in natural waters. Anal. Chim. Acta 27:
31-36.
Moss, B. 1972. Studies on Gull Lake, Michigan. Seasonal and depth
distribution of phytoplankton. Freshwat. Biol. 2: 289-307-
Nalewajko, C., and D.R.S. Lean. 1980. Phosphorus, pp. 235-258 In I. Morris
fed.) The Physiological Ecology of Phytoplankton. Univ. Calif. Press,
Rerkelev.
-------
353
Nauwerck, A. 1963. Die Beziehungen zwischen Zooplankton und Phytoplankton
im See Erken. Symb. Bot. Upsal. 17: 1-163.
Nelson, W.C. 1971. Comparative limnology of Colorado-Big Thompson project
reservoirs and lakes. Colorado Division of Wildlife Report (mineo).
OECD. 1982.
Park, A.R. 1979. Modifications to the model CLEANER requiring further
research, pp. 87-108 D. Scavia and A. Robertson (eds.) Perspectives on
Lake Ecosystem Modeling. Ann Arbor Science, Ann Arbor. 326 pp.
Parsons, T.R., M. Takahashi and B. Hargrave. 1977. Biological Oceanographic
Processes. Pergamon, New York. 332 pp.
Rai, Hakumat (ed.) 1980. The measurement of photosynthetic pigments in
freshwaters and standardization of methods. Arch. Hydrobiol. Beih.
Ergeb. Limnol. 14: 1-106.
Ragotzkie, R.A. 1978. Heat budgets of lakes, pp. 1-19. ^n_ A. Lerman (ed.)
Lakes: Chemistry, Geology, Physics. Springer-Verlag, New York. 363 pp.
Ramberg, L. 1976. Relations between phytoplankton and environment in two
Swedish forest lakes. Scripta Limnol. Upsal. 426: 1-97.
Ramberg, L. 1979. Relations between phytoplankton and light climate in two
Swedish forest lakes. Internat. Rev. ges. Hydrobiol. 64: 749-782.
Rieman, B. 1980. A note on the use of methanol as an extraction solvent for
chlorophyll a determination. Arch. Hydrobiol. Beih. Ergebn. Linnol.
14: 70-78.
Rigler, F.H. 1968. Further observations inconsistent with the hypothesis
that molybdenum blue method measures orthophosphate in lake water.
Limnol. Cceanogr. 13: 7-13.
-------
354
Rohlf, J.F. and R. Sokal. 1969. Statistical Tables. W.H. Freeman, San
Francisco. 253 pp.
Saunders, G.W., F.B. Trama, and R.W. Sachmann. 1962. Evaluation of a
modified C-14 technique for shipboard estimation of photosynthesis in
large lakes. Great Lakes Res. Div. Publ. No. 8. 61 p.
Sawyer, C.N. 1947- Fertilization of lakes by agriculture and urban drainage.
J. New England Water Works. Ass. 61: 109-127-
Schindler, D.W., R.W. Newbury, K.G. Beaty, and P. Campbell. 1976. Natural
water and chemical budgets for a small Precambrian lake basin in
Central Canada. J. Fish. Res. Bd. Canada. 33: 2526-2543.
Smith, R.C. 1968. The optical characterization of natural waters by means of
an "extinction coefficient." Limnol. Oceanogr. 13: 423-429.
Smith, V.H. 1982. The nitrogen and phosphorus dependence of algal bioinass in
lakes: an empirical and theoretical analysis. Limnol. Oceanogr. 27:
1101-1112.
Solorzano, L. 1969. Determination of ammonia in natural waters by the
phenol-hypochlorite method. Limnol. Oceanogr. 14: 799-801.
Solorzano, L., and J. Sharp. 1980a. Determination of total dissolved
phosphorus and particulate phosphorus in natural waters. Limnol.
Oceanogr. 25: 754-758.
Solorzano, L., and J. Sharp. 198Cb. Determination of total dissolved
nitrogen in natural waters. Limnol. Oceanogr. 25: 751-754.
Strom, K.M. 1931. Fetovatn: A physiographic and biological study of a
mountain lake. Arch. Eydrobiol. 22: 491-536.
-------
355
Stumm, W. and P. Baccini. 1978. Man-Made Chemical Perturbations of lakes.
pp. 91-123. In_ A. Lerinan (ed.) Lakes - Chemistry, Geology, Physics.
Springer-Verlag, New York. 363 pp.
Stumm, W. and J.J. Morgan. 1981. Aquatic Chemistry. Wiley, New York. 780 pp.
Taguchi, S. 1976. Relationship between photosynthesis and cell size of
marine diatoms. J. Phycol. 12: 185-189.
Tailing, J.F. 1969. Measurement of photosynthesis using illuminated constant
temperature baths, pp. 108-117 J[n_ R.A. Vollenweider (ed.) A Manual on
Methods for Measuring Primary Production in Aquatic Environments.
Blackwell, London.
Tailing, J.F. 1971. The underwater light climate as a controlling factor in
the production ecology of freshwater phytoplankton. Mitt. Internat.
Verein. Limnol. 19: 214-243.
Tailing, J.F. 1976. The depletion of carbon dioxide from lake water by
phytoplankton. J. Ecol. 64: 79-121.
Tailing, J.F. 1979. Factor interactions and implications for the prediction
of lake metabolism. Arch. Hydrobiol. Beih. 13: 96-109.
Tailing, J.F. 1982. Utilization of solar radiation by phytoplankton. pp.
619-631. ^n_ C. Helene, M. Charlier Th. Montenay-Garestier and G.
Laustriat (eds.) Trends in Photobiology. Plenum, New York.
Vollenweider, R.A. 1968. Scientific fundamentals of the eutrophication of
lakes and flowing waters, with particular reference to nitrogen and
phosphorus as factors in eutrophication. Organization for Economic
Cooperation and Development Report. DAS/CSI/68.27, 182 pp.
-------
356
Vollenweider, R.A. 1969. Possibilities and limits of the elementary models
concerning the budget of substance in lakes. Arch. Hydrobiol. 66:
1-36.
Vollenweider, R.A., W. Rast and J. Kerakes. 1980. The phosphorus loading
concept and Great Lakes eutrophication. pp. 207-234 _In_ Loehr, R.C.,
C.S. Martin, and W. Rast (eds.) Phosphorus Management Strategies for
Lakes. Ann Arbor Science, Ann Arbor- 490 pp.
Wahlstrom, E.E. and D.Q. Eornback. 1962. Geology of the H.D. Roberts Tunnel,
Colorado. Geological Society of America Bulletin 73:1477-1498.
Welch, E.B. 1980. Ecological Effects of Wastewater. Cambridge University
Press, New York. 337 pp.
Wolk, C.P. 1973. Physiology and cytological chemistry of blue-green algae.
Bacterioi. Rev. 37: 32-101.
Wright, R.F. 1974. Forest Fire: Impact on the hydrobiology, chemistry and
sediments of small lakes in northeastern Minnesota. Interim Rept. No.
10, Limnology Research Center, University of Minnesota, Minneapolis,
MN. 129 pp.
-------
357
Summary of Public Participation
The Steering Committee of the Lake Dillon Clean Lakes Study was
composed of the following members:
AMAX Inc.
Keystone Corporation
Copper Mountain Water and Sanitation District
Breckenridge Sanitation District
Frisco Sanitation District
Dillon/SiIverthorne Joint Sewer Authority
Denver Water Department
Town of Breckenridge
Town of Frisco
Town of Dillon
Summit County
Northwest Colorado Council of Governments
Colorado Department of Health
Region VIII, U.S. Environmental Protection Agency
During the progress of the study one or more representatives of
the contractor, Western Environmental Analysts, Inc., attended 12
meetings at which discussions or formal presentations were sade
concerning the Study- The date, place and nature of each meeting is
listed below.
Jan. 6, 1981 - Frisco - Steering Committee meeting to discuss work
plan.
Feb. 10, 1981 - Frisco - Steering Committee meeting to discuss addendun
to work plan.
Apr. 8, 1981 - Frisco - Steering Committee meeting to discuss sampling
plan and parameter measurements.
Nov. 23, 1981 - Frisco - Steering Committee meeting to discuss first
year results and second year plans.
Apr- 13, 1982 - Breckenridge - presentation of first year results at
public meeting.
June 18, 1982 - Frisco - description of Study and first year results to
joint meeting of members of Colorado River and Middle Park Water
Conservancy Districts.
July 7, 1982 - Frisco - discussion of comments on first annual report
with Steering Committee and representatives from two cor.s-j-Ltir.g
firns, Engineering Science and Gulp Wessner Gulp.
-------
353
Feb. 16, 1983 - Frisco - status report on analysis of first and second
year results to meeting of Billon watershed discharge permit
holders and others.
Mar- 1, 1983 - Denver - status report on analysis of first and second
year results to meeting of American Waterworks Association members
and guests.
May 19, 1983 - Frisco - Steering Committee meeting to discuss draft
final report.
July 12, 1983 - Frisco - presentation of first and second year results
to Dillon Reservoir Clean Lakes Policy Committee (policy-making
representatives from Steering Committee membership as opposed to
technical representatives described previously as the "Steering
Committee").
July 26, 1983 - Breckenridge - presentation of final (first and second
year) results at public meeting.
-------
359
Appendix I
DILLON CLEAN LAKES MODEL
Input, Code, and Output
For a model run (Scenario 1A: Low growth, no diversions, wet year)
Input - Pages 1-1 - 1-2
Code - Pages 1-3 - 1-15
Output - Pages 1-16 - 1-21
-------
21 1
3572 42-1
2322 83
0
0
0
BR13572.
BR2 424.
BR4 231
BRG 253
BR7 235.
SRI 319.
SR2 310.
SR3 500.
SR4 364.
TH1 433.
TM4 406.
IM7 513.
MC2 325.
SC2 77.
AAA 325.
BBB 325.
CCC 325.
ODD 325.
EEE 325.
235
192
0
0
O
63.
231 253
213 165
0 0
0 0
0 0
25.
11003.
9719.1410
6133. 11
5165. 178
1041 909
2G78.1305
9789.
4 1 22.
G329.
11422.
6447
1000.
2212.
1 143.
1226.
619.
319 310
120 160
O
0
0
5OO
240
O
0
0
364
227
0
0
0
433
259
0
0
0
406 513
151 435
0
0
0
325
126
0
0
0
4263.
33.
1817.
367.
440.
86.
294.
252.2388.
536.
1258.
317.
697
647.
843.
64.
133.
135.
129.
45.
43. 28.
6700.
386.
177.
28.
178.
10.
78. 90.
42.
67
302.
2500.
77
21
O
0
0
325
126
0
0
0
325
126
325
126
325
126
325
126
2650.
3230.
3300.
-------
1
2
3
4
5
6
7
8
9
10
1 1
12
15
45
74
105
135
1 66
19S
227
2S8
288
319
349
8901
8977
8974
8974
8980
9OOO
9018
90)8
901 8
9017
9016
9015
4
3,
5,
7
33,
92,
56,
27,
14.
9.
6,
6,
. 78
.65
, 19
73
, 71
,04
, 30
,48
, 71
93
,23
, 19
12.
1 1
9
4 ,
3,
6,
30,
2G
15,
10,
13.
5,
, 54
13
.07
. 68
, 82
. 15
. 38
53
35
17
, 12
.79
0
0
o
0
0
0
0
0
0
0
0
0
-------
rv or MINNESOTA FORTRAN COMPILER AVERSION 5.4 - 79/03/01) ON THE 6-400 UNDER KRONOS 2.1.0 ON eo/O4/oi AT 16 03
NIVEFtSITY COMPUT I NO CENTER - UNIVERSITY OF COLORADO
MliF, e=
=DILEXP, L =
OOOOOOD
O10335E
OI0335B
010335B
v> .
0.
Ol 1C! IB
Cll IG^bU
Ol 1 G.-'Ul
O I 1 b-l'-
01 I G.l'j
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
PROGRAM DILEXP (INPUT,OUTPUT,TAPE1,TAPE2,TAPE3)
THIS IS THE LAND USE ELEMENT OF THE DILLON CLEAN LAKES MODEL.
IT PREDICTS NUTRIENT YIELD TO THE LAKE GIVEN LAND USE DATA
AS INPUT AND AMOUT OF RUNOFF
DIMENSI ON NAME(19),WATR(19).AREA(19),PTOT(19),NTOT(19),
1PBKG(19).NBKG(19),LUSE(19,8).PYLD(19,6),NYLD(19,8)
2,USTOP(8),USTON(8),USPCP(8),USPCN(8),HI WAT(8),
3LOWAT(8),DI V(12),DVCONP(12),DVCONN(12)
TAPE1 ONE IS THE DATA SOURCE,STORED ON DISK AS DILEX
THE FIRST RECORD ON TAPE1 IS THE RUNOFF(EXCLUDING ANY PUMPED
WATER). THIS IS VARIABLE USGS AND IS GIVEN AS THOUSANDS
OF ACRE FEET PER YEAR. IT IS THE SUM OF THE 4 USGS
GAUGES (SNAKE,KEYSTONE GULCH,BLUE,TENMILE). THE
SECOND RECORD IS A LIST OF THE WATER YIELDS FOR THE
19 SEGMENTS OF THE WATERSHED FOR 1982 IN MM PER YEAR.
THE THIRD RECORD IS THE SAME FOR 1981. THE FOURTH
RECORD IS A LIST OF THE AMOUNT OF WATER DIVERTED
OR PUMPED FROM OUTSIDE THE WATERSHED IN EACH OF THE
12 MONTHS OF THE YEAR, AS THOUSANDS OF ACRE FEET
PER MONTH. THE NEXT RECORD GIVES THE P CONCENTRATION
OF THE PUMPED WATER MONTH BY MONTH AND THE NEXT GIVE
THE N CONCENTRATION. THE LAST 19 RECORDS GIVE THE
AREA, WATER YIELD, AND LAND USE INFORMATION OF EACH OF
THE 19 WATERSHED SEGMENTS.
REAL LUSE,PYLD,NYLD,NTOT,PTOT,PGRAND,NORAND,NBKG,PBKG
1,PRECP,PRECN,PRPCTP,PRPCTN,PCTSTOP,PCTSTON,NETYLDP,NETYLDN
2,HtWAT.LOWAT
I I SETS THE NUMBER OF RUNS
DO 901 11=1,1
THE AREA OF EACH SEGMENT IS ENTERED ON TAPE1 AS IS THE
WATER YIELD. IF THE WATER YIELDS ARE SPECIFIED AH?NO W!TH
THE LAND USE DATA, THE SPECIFIED VALUES ARE USED.IF
THEY ARE NOT SPEC IFlED,THE DISTRIBUTION OF WATER YIELDS
IS SET PROPORTIONAL TO THE 1981 YIELDS FOR A YEAR
BELOW AVERAGE RUNOFF AND PROPORTIONAL TO THE 1982 YIELDS
FOR AVERAGE OR AtiOVE AVERAGE RUNOFF.
THE RUNOFF IS VARIABLE USGS AND IS THE SUM OF THE 4 GAUGES
IN THOUSANDS OF ACRE FEtT PER YEAR. THE MATRIX LUSE HAS 19 ROWS
THAT ARE 19 SEGMENTS OF THE CATCHMENT AND 8 COLUMNS THAT
ARE THE LAND USLS AS FOLLOWS: 1 )RES I DENT IAL SEWER NONPOINT.AS
NUMBER OF PERSONS. 2)URBAN SEWER AS PERSONS,3)SEPT IC AS
PERSONS.4)INTERSTATE HIGHWAY AS HECTARES ROAD PLUS RIGHT OF WAY,
t>)SKI AREA AS HECTARES OF CLEARED AREA, 6)PACKAGE PLANT OUTFALLS,
AS NUMBER OF PERSONS SERVED , 7 ) TE.RT 1 ARY OUTFALLS SAME.8)AMAX
MINING AS WORKFORCE SIZE
THE GROSS YIELDS ARE IINCORRECTED FOR INVENTORY CHANGE IN THE
VALLEY BOTTOMS, THE NET YIELDS ARE CORRECTED FOR THIS.
READ IN THE AMOUNTS OF RUNOFF.
READ (1,260) USGS, ( III WAT(K),K=1 , 19>, (LOWAT(K),K=1 , 19)
2GO FORMAT ( F5 . O , / . 1 9F-1 . 2 , / . 1 9F-1 . 2 )
READ IN THE INFORMATION ON PUMPING AMD DIVERSIONS
READM.297) , ( DVCONP(K ) , K= 1 , 1 2) ,
1 ( DVCONN < K ) , « -- 1 . 1 2 )
297 FORMAT ( 1 2f-b .(),/, 1 21-'o C),/ I^Fb.O)
DO 201 1-1.1 9
-------
10.
1 1
12.
01 1G4GB
01 1670B
01 1G70B
READ (1.202) NAMEC I ) , WATR( I ) , AREA( 1 ) , (LUSE( 1 ,K),K=1 , 8)
202 FORMAT ( A3, F5 . 0, F7 . 0, 8F5 . O)
201 CONTINUE
C COMPUTE THE LOADING BY DIVERSION AND PUMPING AND CONVERT
C DIV TO MILLIONS OF M3 PER MONTH
1 3.
14.
15.
16.
17
is.
19.
20.
21 .
O ft
23.
24.
25.
26.
27
28.
29.
30.
31 .
312.
33.
3-1.
35.
36.
37
38.
39.
40.
41 .
42.
43.
44 .
45.
46.
47,
48.
43.
50.
O1 1672B
01 1672B
01 1G73B
01 1675B
01 1677B
01 1702B
01 1705B
01 17068
01 1710B
01 1713B
0 1 1 7 1 5B
01 1717B
01 1717B
01 1721B
01 1723B
01 1727B
01 1730B
011 73 IB
01 1731B
O1 1733B
01 1740B
01 17468
01 1754B
01 1762B
01 1 770B
01 1774B
0120036
O12012B
012014B
012017D
012021B
O12024B
012026B
01 203 IB
O12033B
01203613
012045B
012054B
DVSUMP=0. 0
DVSUMN--O.O
D8 298 1=1,12
DIV(I)=DIV(I) *1000. *1233. 5* .000001
DVSUMP=DVSUMP+DI V( 1 )*DVCQNP( I )
DVSUMN=DVSUMN+DIV( I )*DVCONN( I )
298 CONTINUE
C SET THE WATER YIELDS
DO 280 1=1,19
IF (WATR(2) .GT. 10. )GO TO 280
IF (US6S.GT. 180. )GO TO 281
WATR( I )=LOWAT( 1 >*(USGS/1 16.5)
GO TO 280
201 CONTINUE
WATR( I )=HIWAT( I ) * ( USGS/21 0. 9)
280 CONTINUE
WATR( 1 )=WATR(2)
PGRAND=0.0
NGRAND=O.O
C APPLY LAND USE EQUATIONS
DO 203 1=1,19
PYLD( I , 1 ) = . 000001 *LUSE<
NYLDC I , 1 )=. 000001 *LUSE(
PYLDC I , 2)=.000001*LUSE(
NYLD( I ,2)=. 000001 *LUSE(
PYLD( I , 3)=. 000001 *LUSE(
NYLD( I , 3) = .000001 *LUSE(
PYLD( I . 4 ) = . 000001 *LUSE(
NYLDC I ,4)=. 000001 *LUSE(
PYLDf I , 5)=. 000001 *LUSE(
NYLD( I ,5)-.O00001*LUSE(
PYLD( I ,6)=. 000001 *LUSE(
NYLD( I ,6)=. 000001 *LUSE(
PYLD( I , 7)=. 000001 *LUSE(
NYLD( I ,7)= . 000001 *LUSE(
PYLD( I , 8) = . 000001 *LUSE(
NYLDC I , 8)= . 000001 *LUSE(
PBKG ( I ) =AREA ( I ) * 1 OOOO . *
NBKG( I )=AREAf I )* I 0000 . *
PTOT( I )=PYLD( I , 1 )+PYLD(
, 1 )*2. 95*WATR( I )**1 . 372
, 1 )*536. *WATR( I )**1 . 154
(2)*1 .50*WATR( I )*»1 . 372
,2)*410. *WATR( 1 )**1 . 154
, 3)*3440. *WATR( I )**0.759
,3)* (44300. *WATR( I )-5205. J
,4)*10000. * , 00209*WATR( 1 )**1 . 799
,45*10000. *1 .71*WATR( I )**! . 13
,5)*6. 50*10000.
, 5)*42. *10000.
,6)*270000.
,6)*1 197000.
, 7)*47000.
, 7)*1 197000.
, 8)*98000.
, 8)*19930000.
000001*. 000782*WATR( I )**1 .372
OOOOO1 * . 0842*WATR( I >**1 . 154
, 2)+PYLD( I , 3)+PYLD( I ,4) +
1PVLD( 1 , 5)+PYLD( I ,6)+PYLD( 1 , 7>+PYLD( I , 8>+PBKG( I )
51
0120650
NTOT( I )=MYLD( I , 1 )+NYLD(
, 2 ) +NYLD ( I , 3 ) +NYLD ( I , 4 ) +
1NYLD( I . b)+NYLD( I , 6)+NYLDf 1 , 7>+NYLD( 1 , 8>+NBKG( I )
52.
S3.
54.
012076B
0121 OOB
012102B
PGRAND=PGRANn+PTQT( I )
NGRAND = NGRANC1*NTOT< 1 )
203 CONTINUE
C ADD 3.5 PERCENT FOR CONSTRUCTION ACTIVITIES P.2.2 PERCENT N
55.
56.
57
50.
59.
CO.
61 .
012103B
01 2104B
01 21 OOB
01 211 OB
O121 12B
0 1 2 1 1 CD
0121 16B
CGNSTP=PGRAND*0. 035
CONSTN=NGRAND*0. 022
PGRAND=PGRAND< CONSTP
NGRAND=MGRAMD i CONSTN
C WFUTE OUT PREDICTED YIELDS
WRITE (2, 190)
19O FORMAT (///////)
WRITE (2,226)
BY SEGMENTS
-------
62.
63.
64 .
65
GO.
67.
G6.
69.
70.
T\
72.
73.
74.
75.
76
77
78.
79.
80.
81 .
82.
83.
84.
85.
86.
87
08.
69.
90.
91
92.
93
94.
95
9G.
97
98.
99.
10O.
101
102.
103.
1.1-1 .
105.
lOf>.
IO/
1OO
109.
110.
1 1 1
012I21B
012I21B
012124B
O1 2124B
012125B
012151B
012151B
O12153B
012157B
012162B
012162B
012165B
012166B
012212B
012214B
012220B
012223B
012223B
012226B
O1222GB
012227B
012241B
012241B
OI2243B
012251B
01 2251B
01 2215 IB
0122'o3B
O12254B
OI2256B
012260B
0122GOB
01 2260B
01226.7B
0122G2B
0122G3B
0122G5B
01226GB
012270B
012271B
01 2273B
0122750
Ol 230 IB
Ol 23CKII)
Ol ?30 IB
01 I-'Jij/h
Ol :>.> I :)B
Ol r.Jl :m
U 1 231 L.Li
oi :'.•> i (.M
22O FORMAT (49H GROSS P YIELD FOR WATERSHED SEGMENTS, KG PER YR )
WRITE (2.215)
215 FORMAT ( 1 5H STATION NAME ,11H MM RUNOFF, 11H AREA HA
111H RES SEW ,11H URB SEW .11H SEPTIC ,11H HI WAY
2 1 1 H SK 1 SLOPE , 1 1 H PKG PLT , 1 1 H TERT PLT , 1 1 H AMAX
31 1H BKGRND )
DO 204 1=1,19
WRITE (2,205) NAME( 1 ) , WATRC I ) . AREA( I ) , (PYLD( 1 , K) , K=l . 6) ,
1PBKG( I )
205 FORMAT ( / , 4X. A3, 5X . 1 1 F1 1 . 1 )
204 CONTINUE
WRITE (2, 190)
WRITE (2,227)
227 FORMAT (49H GROSS N YIELD FOR WATERSHED SEGMENTS, KG PER YR)
WRITE (2,215)
DO 206 1=1,19
WRITE (2,205) NAME( 1 ) , WATR( 1 ) , AREA( 1 ) , (NYLD( 1 , K) , K=1 , 8) .
1NBKG( I )
206 CONTINUE
WRITE (2. 190)
WRITE (2,256)
258 FORMAT (42H GROSS P AND N YIELD BY SEGMENT. KG PER YR)
WRITE (2,257)
257 FORMAT ( 6H NAME , 8H PTOTAL , 8H NTOTAL )
DO 2O7 1=1,19
WRITE (2,208) NAME( I ) , PTOT( I ) , NTOT( 1 )
208 FORMAT ( / , IX, A3, 2F8 . O)
207 CONTINUE
WRITE (2.209) PGRAND, NGRAND
209 FORMAT (/////. 32H GROSS KG PER YR P IN RUNOFF IS.F6.0,/,
132H GROSS KG PER YR N I N RUNOFF IS.F0.O,/)
C COMPUTE THE TOTAL P AND N YIELDS TOTAL FOR EACH LAND USE, GROSS
PRECP=72G.
PRECN=9672.
GRNDP=10.
GRNDN=1066.
OO 210 J = 1 , 8
USTGP( J)=0.0
USTON( J)=0. 0
210 CONTINUE
TBKGf =0.0 .
TBKGN-0.0
DO 250 1=1.19
TBKGP=TBKGP+PBKG( 1 )
TBKGN=TBKGN+NBKG( 1 )
2bO CONTINUE
DO 211 J = 1 . 8
DO 212 1=1.19
USTOP( J)=PYLD( 1 . J)+USTOP(J>
USTON( J)=tlYLO( 1 . J) +USTONI J)
212 CONTINUE
•J; 1 1 CONT 1 NUE
WRITE (2,270)
27O FORMAT ( 5OH GROSS YIELD BY SOURCE KG PER YR P ( ABOVE ), N( BELOW ))
WRITE (2,271)
271 FORMAT f/,1311 RES 1 D . SEWLK'.IOII URB SEW , 1 OH SEPTIC
11 OH III WAY . IOH SKI SLOPE, 1011 PKG PLT ,1011 TERT PLT,
21011 AMAX 1011 BKGRND , 1 OH PRECIP , 1 0H GRNDUAT ,
310II DIVERSION, 1 OH COIIMKIIOT)
UKI re .2,^13) ( us Torn- > , K- i , o) , IBKGP, PUECP, GRNOP, ovsurip, CUNSTP
M
I
-------
i 12.
13.
114.
1 1t>.
lie.
1 17
1 18.
119.
120.
121 .
22.
23.
24 .
25.
26,
127
128.
129.
130.
131
132.
133.
134.
13S.
136.
137.
133.
139.
140.
141 .
142.
143.
144.
145.
146.
147.
148.
149.
150.
151 .
152.
153.
154.
155.
156.
157.
158.
159.
O12343B
01234GB
01 234GB
012350B
012353B
012355B
012360B
O12362B
0123G48
012367B
O12372B
01237GB
012402B
01240GB
012412B
012422B
0124318
012443B
012454B
012455B
012457B
012461B
012-4G3B
0124658
O12467B
012471B
012474B
012476B
012500B
012502B
012504B
012507B
012511B
012513B
012315B
012521B
012524B
012524B
01252VB
012527B
012532B
012557B
012562B
0125G2B
O125G5B
012612B
O12615B
Ol 2G22H
160.
O12G22B
1(USTON(K).K=1,8),TBKGN,PRECN,ORNDN,DVSUMN,CONSTM
WRITE(2,190)
213 FORMAT ( 13F10.1,/,13F1O.1)
C COMPUTE INVENTORY CHANGE AND ADJUST YIELD ACCORDINGLY
PCTSTGP=USGS*(-.40)+87.6
PCTSTON=USGS*(-.57)+112.
DO 290 1=1,6
USTOP(I)=USTOP(I)*(100.-PCTSTOP)*.01
USTON(I)=USTONfI)*(1OO. -PCTSTON)*.Ol
290 CONTINUE
USTOP(8)=USTOP(8)*(100.-PCTSTOP)*.01
USTGN(8)=USTGN(8)*(100.-PCTSTON)*.01
CONSTP=CONSTP*(100.-PCTSTOP)*.Ol
CONSTN=CGNSTN*(100.-PCTSTON)*.Ol
TBKGP=TBKGP*<1OO.-PCTSTOP)*.01
TBKGN=TBKGN*(1OO.-PCTSTON)*.01
WWTPP=PYLD(1,7>+PYLD(16J7)+PYLD(18,7)
1+PYLD(11.7)*(100.-PCTSTOP)*.01
WWTPN=NYLD(1,7)+NYLD(16,7)+NYLD(18,7)
1+NYLD(11,7)*(100.-PCTSTON)*.01
N£TYLDP=(PGRAND-USTOP(7))*(100.-PCTSTGP)*.01+WWTPP+PRECP+GRNDP
1 +DVSUMP
NETYLDN=(NGRAND-USTGN(7))*(100.-PCTSTON)*.Ol+WWTPN+PRECN+GRNDN
1*DVSUMN
USTOPC7)=WWTPP
USTON(7)=WWTPN
C COMPUTE THE PERCENTAGES OF EACH CONTRIBUTION
DO 295 J=O ,8
USPCP(J) = (USTOP(J)/(NETYLDP))* 100.
USPCNC J) = (USTON(J)/NETYLDN)* 100.
235 CONTINUE
PRPCTP=CPRECP/(NETYLDP))*100.
GRPCTP=(GRNDP/(NETYLDP))* 100.
TBKGPPC=(TBKGP/(NETYLDP))* 100.
DVPCP=(DVSUMP/NETYLDP)* 100.
CONSPCP=(CONSTP/NETYLDP)* 100.
TBKGNPC=(TBKGN/NETYLDN)* 100.
PRPCTN=(PRECN/NETYLDN)* 1 00.
GRPCTN=(GRNDN/NETYLDN)* 100.
DVPCN=(DVSUMN/NETYLDN)* 100 .
CONSPCN=(C0NSTN/NETYLDN)* 100.
WRITE (2,190)
WRITE (2,296)
296 FSRMAT(42H VALUES UP TO THIS POINT UNCORRECTED FOR ,/,
)55H INVENTORY CHANGE IN RIVER BOTTOMS, NOW CORRECT THIS.,//)
WRITE (2,370)
370 FORMAT (5OH NET LOADING BY SOURCE KG PER YR P(ABOVE),N(BELOW))
WRITE (2,271)
WRITE (2,213) (USTOP(K),K=1,8),TBKGP,PRECP,GRNDP,DVSUHP,CGNSTP,
1(USTON(K),«=1,8),TBKGN,PRECN,GRNDN.DVSUMN,CONSTN
WRITE (2,275)
275 FORMAT (//,4BH THE NET YIELDS BY PERCENT. ARE P(ABOVE),N(BELOW))
WRITE (2,271)
WRITE (2,213) (USPCP(K),K=1,8),TBKGPPC,PRPCTP,GRPCTP,DVPCP,
1CONSPCP,(USPCN(K),K=1,8),TBKGNPC,PRPCTN,GRPCTN,DVPCN,CONSPCN
WRITE (2,190)
WR1TE(2,265) NETYLDP.NETYLDN
205 FORMAT(48H THE NET P YIELD TO THE LAKE IN KG PER YEAR IS,F6.O,
1./.48H THE NET N YIELD TO THE LAKE IN KG PER YEAR IS.Fa.O)
WRITE (3,902) NETYLDP,PRECP,USGS,(DIV(K),K=1,12),
M
I
CTi
-------
1(DVCONP(K),K=1 , 12)
161 012640B 902 FORMAT (3F7.0,/,12F6.2,/,12F6.2)
162. 012G40B 901 CONTINUE
163. O12642B STOP
164. 012643B END
t-t
I
-------
UNIVERSITY OF MINNESOTA FGRTKAN COMPILER (VERSION 5.4 - 79/03/01) ON THE 6400 UNDER KROMOS 2.1.O ON 83/04/O1
UNIVERSITY COMPUTING CENTER - UNIVERSITY OF COLORADO
AT 16.03
3.
4.
6.
7.
8.
9.
10.
1 1 .
12.
13.
15.
1C.
17.
10.
19.
20.
21 .
22.
23.
24.
OOOOOOB
01241 IB
01241 IB
01241 1 B
0126238
012G44B
0126440
012662Q
012662B
012662B
012664B
0126&4B
012G71B
012G73B
012674B
012676B
OI2677B
012702B
012703B
012704CS
012705B
0127I2B
O12714B
O12715Q
MNF,E = 4,l=OILTP,L=LL.
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
PROGRAM DILTP (INPUT.OUTPUT,TAPE11, TAPE2,TAPE3,TAPE4)
THIS IS THE TROPHIC STATUS ELEMENT OF THE DILLON
CLEAN LAKES MODEL. THIS VERSION IS SET UP TO
INTERFACE WITH LAND USE ON THE INPUT SIDE AND
EFFECTS ON THE OUTPUT SIDE. THE PROGRAM TAKES
USER SUPPLIED INPUT ON TAPE11 AND INPUT SUPPLIED
BY D1LEXP ON TAPE3. KRONOS COMMANDS FOR INTERFACE
ARE ON DILPROC. VARIABLES COMING IN BY MONTH
ON TAPE 11 INCLUDE MONTH NUMBER (M<3N). DAY NUMBER
(DAY), LAKE LEVEL IN FEET (LEV), TOTAL SURFACE
WATER IN MILLIONS OF M3 PER MQNTH (VIN),TOTAL
OUTFLOW IN MILLIONS OF M3 PER MONTH (VUT). VARIABLES
COMING IN ON TAPE3 FROM DILEXP ARE TOTAL P LOAD IN
KG PER YEAR (NETYL.DP), P LOAD DUE TO PRECIPITATION
IN KG PER YR (PRECLO), AMOUNT OF DIVERTED OR
PUMPED WATER IN MILLIONS OF M3 PER MONTH FOR ALL
12 MONTHS, AND P CONCENTRATION MONTHLY IN PUMPED
OR DIVERTED WATER, AS uo PER L.
THE PROGRAM WRITES TO TAPE2 AND OUTPUT IS
PACKED BY DILPROC.
DIMENSION MGN(12>. DAY(12),LEV<12),VIN(12),
1VUT (12), PLD (12), ARE (12), VOL (12), RLDU2)
2,DVCONP(12)
REAL MON, LEV, PLD, PLOTS, PLAKE, LANDLO,
SET NUMBER OF DATA SETS TO BE TREATED
DO 107 11=1,1
READ INPUT DATA FOR EACH OF 12 MONTHS
DIV(12)
PRECLO,NETYLDP
READ (11,3) (MON(K), DAY (K), LEV (K), VIN (K), VUT(K)
1 . K=1,12)
3 FORMAT (3F5.0.2F7.0)
READ (3,101) NETYLDP,PRECLO,USGS,(DIV(K),K-1,12),
1(DVCOMP(K),K=1,12)
101 FORMAT (2F7.0,F7.0,/,12F6.2,/,12FC.2)
DVSUMP=0.0
C COMPUTE LOADING FRACTIONS
DO 150 1=1,12
DVSUMP=DVSUMP+DIV(I)*DVCONP(I)
ISO CONTINUE
LANDLO = NETYLDP - PRECLO-DVSLIMP
C TOTAL UP THE WATER EXCLUDING DIVERSION INPUT AND ADD GROUNDWATER
C VTOT WILL BE TOTAL NONDI VERSION WATER INPUT PER YR IN M3
VTOT=0.0
DO 103 1=1,12
VIN( I )=VIN( I ) t-,575
VTOT=VTOT+VIN(1)*100OOOO.
103 CONTINUE
C COMPUTE MONTHLY P INPUTS IN KG PER MONTH
DO 102 1=1,12
RLD(I)=O.OG3*PRECLO
PLD(I1=LANDLO*(VIN(I)*10OOOOO./VTOT)+DVCONP(I)*DIV(I)
102 CONTINUE
RLD(G)=0.3*PRECLO
C WHITE PRELIMINARY DATA DATA ONTO TAPE 2
WRITE (2,2)
I
oa
-------
O I 2722B
2 FORMAT I//////, ' »MONTH*
I1 »LEVEL FT* ', *MIL M3MO
2' *MIL M3MO OUT* ' ' »KG P
,' «DAY«
ROFF* ',• *M!L M3MO DIVER*
NONPREC* ',' »KG P PREC* ')
20.
27
26.
29.
30.
31
32
33.
34.
35.
36.
37.
38.
39.
40.
41
42.
43.
44 .
4'j.
4G.
47
48.
40.
SO.
51
tj2.
53
b4
65.
«oo .
S7
5U
59
012722B
01 2750B
012750B
012751B
01 27G1B
0127G7B
012775B
013003B
Ol 301 IB
013017B
013025B
01 3O33B
01 304 IB
O13O43B
01 3044B
01 3047B
01 304 VB
013050B
013O51B
01305 IB
01 30t>2B
Ol 3054B
Ol 3O5GB
Ol 3OG1B
01 3OG3B
0130G5B
Ol 3O70B
01 307 IB
O 1 3O72B
Ol 3O74B
Ol 3O76B
Ol 31 03B
Ol 3 1 UUJ
O 1 3 1 1 1 B
WRITE (2,6) (MON(K> , OAY(K> ,LEV(K) , VIN(K),DI V(K)
1VUTCK) . PLD(K) , RLD(K) , K=1 , 12)
6 FORMAT (F7.0, F14.0, F12.0,F12.2, 2F18.2, 2F17.
C COMPUTE LAKE VOLUME At-40 AREA FROM LEVEL
DO 7 I = 1,12
IF (LEV ( 1 ) .GT. 8000. . AND. LEV( I ) . LE . 8950. )ARE( 1 )
1(M -6900. )+1000.
IF (LEV( 1 ) .GT.8900. .AND. LEV( 1 ) .LE. 8925. )VOL( 1 )=
1(1) -8900. )+40.
IF (LEV( 1 ) .GT. 8925. .AND. LEV ( 1 ) . LE. 8950. )VOL( 1 )=
1 ( 1 1-8923. )+67.
IF (LEV ( I ) .GT.6950. . AND. LEV( I ) . LE . 8975. )ARE( 1 )
1(1) - 8950. )+1600.
IF (LEV ( 1 ) .GT. 8950. .AND. LEV ( I ) . LE . 8975. ) VOL
1(1) - 8950. ) +103.
IF (LEV ( 1 ) .GT. 6975. .AND. LEV ( 1 ) . LE . 90OO. ) ARE( 1
1(1) - 8975. ) +2000.
IF (LEV( 1 ) .GT. 6975. .AND. LEV ( 1 ) . LE. 900O. )VOL( I )
1(1) - 8975. )+150.
IF (LEV ( I ) .GT. 9000. . AMD. LEV ( I ) . LE . 9025 . ) ARE( I
1(1) - 9000. )+2700.
IF (LEV ( 1 ) .GT. 9000. . AND. LEV ( I ) . LE . 9025. )VOL( 1 )
1(1) - 900O. )+210.
VOL ( I ) = VOL ( 1 ) * 1 000 . « 1 233 .
ARE( 1 ) = ARE( I )* .4047
7 CONTINUE
C SUM VARIABLES OVER YEAR VOLS IN M3
VUTTd = 0.0
PLDTd =0.0
AR£TO = 0.0
VOLTO = 0.0
VINTd=VTOT
DO 8 1 = 1,12
VUTTO = VUTTO + VUT (I)*1000OOO.
PLDTO = PLDTO + PLD (I) + RLD(I)
ARETO = ARE TO + ARE(I)
VOLTO = VOLTO + VOL ( 1 )
VI NTO = VINTO + DI V( I ) * 1 000000.
8 CONTINUE
C COMPUTE P LOAD PER UNIT AREA IN MG PER M2 PER YR,
C RESIDENCE TIME IN YR,
C VOLUME LOAD I NO IN MG PER M3 PER YR, AMD MEAN DEPTH
PLDTO = K-LOTO/36S.
AREMN = ARETO/ 12.0
VOLMN = VOLTO/ 12.0
ARELD - (PLDTO* 1000. * 1000. *3G5 .) /(AKEMN* 1000O. )
VOLLD = ( PLD1O* 1000. * 1 OOO . *3Gl> . ) /VOLMN
DEPMN = VOLMN/ ( AREMN Jt 1 OOOO. )
RESI D = VOLMN --V INTO
t
1 )
=(580. /50. )*(LEV
(28. 15/25. )*(LEV
( 37./2S)* (LEV
=(400. /25. )*(LEV
( I )-(47. 725. )*(LEV
)=(700. 725. )*(LEV
=(60. 725. )*(LEV
)=( 1050. 725. )*(LEV
=(75.725. )*(LEV
HYDRAULIC
IN M
bl
C COMPUTE THE EFFECT I VL SEDIMENTATION COE1 F 1 Cl EMT , S 1 GMA
0131 I3B SIGMA = USGS*0 0036' . G3
C COMPUTE THE PREDICTED IN-I.AKE P CONCENTRATION
C US I NO VOLLENWEIDER MODI I I LI) INDEX
OlJllbU PLAKE= (ARELD/dlEFMN* ( 1 /KF.blD 'SIGMA)))
C COHPIME THE CHI (li.'OI'HYl.l (IVI'R MIL U 1>M L.AYER,
C STRATI PICA II OI-J SLASON 111 AH
CHLA= 1 . -I-I'J »Al OL'. I O( I I M' h ) - . :)'J13
-------
03. O13126B CHLA=10**CHLA
C WRITE OUT RESULTS
C4. 013131B WRITE (2,1O) AREMN, V8LMN, VINTO
C5. 013140B 10 FORMAT (/////,2X,'MEAN LAKE AREA OVER THE YEAR IN HECTARES IS'
1F6.0./.2X,
2'AVERAGE LAKE VOLUME OVER THE YEAR IN M3 IS ',
3F11.0,/,2X,
4'TOTAL ANNUAL INFLOW IN M3 INCLUDING DIVERSION IS'. F11.0)
66. 013140B WRITE (2,11) VUTTO, PLDTO
67 O13145B 11 FORMAT J/.2X,'TOTAL ANNUAL OUTFLOW IN M3 IS',F11.0,/,2X,
1'TOTAL LOADING IN KG PER DAY IS', F7.3)
68. 013145B WRITE (2,12) ARELD, VOLLD, DEPMN
09. 013153B 12 FORMAT (/,2X,'AREAL LOAD IN M6 PER M2 PER YR IS ',
1F6.1,/,2X,'VOLUME LOAD IN MG PER M3 PER YR IS',
2F6.1,/,2X,'MEAN DEPTH IN M IS',
3F5.1 )
70. O13153B WRITE (2,13)RES ID,PLAKE
71. 013160B 13 FORMAT (/,2X,'TAU IN YR IS', F5.2,/,2X,
1'PREDICTED P CONCENTRATION IN TOP 15 M IS',F5.1,//)
72. 013i60B WRITE (2,14)CHLA
73. 013164B 14 FORMAT (/,2X,'MEAN CHLOROPHYLL A 0-5M DURING STRATIFICATON IS'
1F7.2,//)
C WRITE TO TAPE4
74. 013164B WRITE (4,60) USGS,PLAKE,CHLA,DEPMN
75. 013173B GO FORMAT (F8.O,2F8.2,F8.1)
76. 013173B 107 CONTINUE
77. 013175B STOP
78. 013176B END
H
I
-------
UNIVERSITY OF MINNESOTA FORTRAN COMPILER (VERSION 5.4 - 79/O3/O1> ON THE 64OO UNDER KRONOS 2.1.O ON 63/O4/O1 AT 16.O3
UNIVERSITY COMPUTING CENTER - UNIVERSITY OF COLORADO
MNF,E=4,I=DILEFP,L=LLL.
PROGRAM DILEFP (INPUT,OUTPUT.TAPE4,TAPE2,TAPES)
REAL PCON
DO 61 11=1.1
THIS IS THE EFFECTS ELEMENT OF THE DILLON CLEAN
LAKES MODEL. IT TAKES INPUT FROM THE TROPHIC
STATUS ELEMENT ON TAPE4 AND WRITES TO TAPE2.
THE PURPOSE OF THE PROGRAM IS TO CONVERT
THE PREDICTED CHANGES IN P CONCENTRATON TO
PREDICTED CHANGES IN OTHER VARIABLES OF INTEREST.
READ (4,1) USGS,PCON,CHLA,2BAR
1 FORMAT (F8.O,2F8.2,F8.1)
PREDICT THE MEAN AND MIN SECCHI DEPTH, JULY-OCTOBER
SD=2.4-0.68*ALOG(CHLA)
SD=EXP(SD)
SDMIN=SD».7O
WRITE (2,2) SD.SDMIN
2 FORMAT /////,36H MEAN SECCHI DEPTH IN M JULY-OCT IS.F5.2,//,
135H MIN SECCHI DEPTH IN M JULY-OCT IS.F5.2)
PREDICT THE CARLSON TSI
CTSI=10.*<6.-(SD/ALOG(2.)))
WRITE (2,3) CTSI
3 FORMAT /,32H CARLSON TROPHIC STATE INDEX IS.F5.0)
PREDICT THE AOHD, SUMMER DEPLETION, AND MINIMUM 02
AOIID=-0. 67*ALOG10(SD)-i-0.47»ALOG10(ZBAR)+2.45
AOIID=10. «*AOHD
CORR=-1. 43*USGS + 296.1
1
2.
3.
4.
5.
6
7
8.
9.
10.
1 1
12.
13.
14.
13.
16.
17.
18.
19.
2O.
21
22.
23.
24.
25.
26.
27
26.
29.
30.
31 .
OOOOOOB
010335B
010335B
010337B
010347B
010347B
010353B
01 0355B
O1 0356B
010364B
01O3G4B
010370B
U10376B
010376B
01037GB
01040GB
01041 IB
0104140
0 1 04 1 50
0 1 O4 1 UB
0 1 04 2 1 B
010422B
010431B
01 043 IB
010433B
OI0431JB
0 104 4 013
OI0442B
01 0444B
01 04-1 7B
Ol 04S7B
C
C
C
C
C
C
C
C
C
DEPL=0.00698*AOHD
IF (DEPL.GT.9.0»DEPL=9.0
BOTO2=9.0-DEPL
WRITE(2,4)AOHD,DEPL,BOTO2
4 FORMAT (//,3GH THE AREAL HYPOLIM O2 DEFICIENCY IS,F5.O.//,
132H THE SUMMER DEPLETION IN PPM IS.F5.1,//,
245H THE MINIMUM O2 5M OVER THE BOTTOM IN PPM IS.F5.1)
PCUN- ((PCON/7.3)«100. )- 100.
CHLA=((CHLA/7. 1 )* 100. )- 100.
IF (PCON.OT 100.)PCON-1OO.
IF (CHLA.GT.100.)CHLA=100.
SD=10O.-((SD/2.9)*10O.)
BOT02=100. -( (BOTO2/4.8)* 100. )
WRITE (5,66) PCON, CHLA, SD.BOTOi!
CO FORMAT (3H 1,3H l,Fb.O,/,3H 1,3H
1511 2,311 2,F5.0,/,3H 2, 3H 2,5H
3.F5.0./,3H
4 , 3H
ji? oioiu/n
:>:t 01O4O1I1
D-l OIO.IU.'U
23H
33H
43H
S3H
63H
f. 1 OONT
STOP
KND
3
4
5
G
7
1
t
,
,
(
,
3H
311
3H
3H1
3(11
3,
.1 ,
0,
8,
. 8,
F5
F5
511
5H
5H
.0,
. o,
1
-59
/ , 3H
/ , 3(1
0, /,
3. , / ,
. O, / ,
3
4
311
311
311
NUt
3, 5H
4, 5H
1 ,
0,
0,
0,
511
O,/,
,/.
, / .
/,
5,3H 5,5H 0,,
0, 3112.2, 3H1 3. , / ,
7, 3112. 2, 5H-59. O)
-------
UNIVERSITY COMPUTING CENTER
UNIVERSITY OF COLORADO
SPSS- - STATISTICAL PACKAGE FOR THE SOCIAL SCIENCES
VERSION 8.3 (NOS) -- MAY 04, 1982
070000 CM MAXIMUM FIELD LENGTH REQUEST
63/04/01.
16.03.43.
PAGE
RUN NAME
VARIABLE LIST
N OF CASES
INPUT MEDIUM
INPUT FORMAT
PLOT
VAR001 TO VAR003
14
TAPES
FIXED(F3.0,F3.0,F5.0)
ACCORDING TQ YOUR INPUT FORMAT, VARIABLES ARE TO BE READ AS FOLLOWS
VARIABLE FORMAT RECORD COLUMNS
VAR001
VAROO2
VAR003
F 3. 0
F 3. 0
F 5. 0
1 -
4-
7-
3
6
11
THE INPUT FORMAT PROVIDES FOR 3 VARIABLES.
IT PROVIDES F8R 1 RECORDS (*CARDS*> PER CASE.
A MAXIMUM OF 11 *COLUHNS* ARE USED ON A RECORD.
3 WILL BE READ.
CPU TIME REQUIRED..
.115 SECONDS
GIVEN
PLOT PLOTS=VAR003(-100.100) WITH VAR002(0,5) BY VAR001(1,7)/
TITLE=AS APPROPRIATE/ :
TITLEX= TOTAL P CHLA SECCHI BOTTOM 02/
TITLEY=BETTER,PERCENT WORSE,PERCENT/
SIZE=25,25/
SYMBOLS=0,O,0,0,0,-3,-3/
PLOTS = VAR003(-100, 100)WITH VAR002(0,5) BY VAROO1M,?)/
TITLE=AS APPROPRIATE/
T!TLEX= TOTAL P CHI.A SECCHI
TITLEY^BETTER,PERCENT
SIZE=9,9/
XDIV=10/
YD1V=10/
SYMBOLS=0,0,0,0,0,-3,-3/
READ INPUT DATA
3 VARIABLES, INITIAL CM ALLOWS FOR 271 CASES
MAXIMUM CM ALLOWS FOR 949 CASES
BOTTOM O2/
WORSE,PERCENT/
H
M
t-o
- - - WARNING - - -
MO PLOTS WILL BE PRODUCED UNLESS THE PLOT FILE (TAPE99) IS DISPOSED.
THE FOLLOWING JOB SETUP WILL YIELD MEDIUM L1WF ON PLAIN PAPER-
-------
SPSS.
DISPOSE,TAP£G9,GP=CP3.
EE THE CALCOMP MANUAL FOR OTHER PAPER/PEN COMBINATIONS.
t -\
I
-------
PLOT 63/04/01. 16.03.43. PAGE
OPT!OH - 1
IGNORE MISSING VALUE INDICATORS
(NO MISSING VALUES DEFI NED, . .OPT I ON 1 MAY HAVE BEEN FORCED)
PLOTS WILL NEED APPROXIMATELY 53 INCHES OF PAPER
-------
PLOT
FILE
N.JNAME
(CREATION DATE = 03/04/01 )
83/O4/O1.
16.03.43.
PAGE
'LOT 1
3N THE X-AXIS.
.IN THE Y-AXIS.
VAROO2
VAROO3
1
2
3
4
5
6
7
VALUES
VALUES
VALUES
VALUES
VALUES
VALUES
VALUES
PLOTTED
PLOTTED
PLOTTED
PLOTTED
PLOTTED
PLOTTED
PLOTTED
2
2
2
2
2
2
2
PI OT ;.
pi THE \-AXIS..
O i THE Y-AXIS. .
VAROO2
VAR003
1
2
3
4
5
6
7
VALUES
VALUES
VALUES
VALUES
VALUES
VALUES
VALUES
PLOTTED
PLOTTED
PLOTTED
PLOTTED
PLOTTED
PLOTTED
PLOTTED
2
2
2
2
2
2
2
-------
PLOT
83/04/01.
16.03.43.
PASE
CPU TIME REQUIRED..
.835 SECONDS
FINISH
TOTAL CPU TIME USED.. .965 SECONDS
RUM COMPLETED
NUMBER OF CONTROL CARDS READ 21
NUMBER OF ERRORS DETECTED O
STATION NAME
BR1
BR2
BR4
BRG
SRI
SR2
SR3
SR4
TMI
TM4
TM7
MC2
SC2
AAA
BOB
CCC
ODD
)R WATERSHED SEGMENTS, KG PER YR
1M RUNOFF AREA HA RES SEW URB SEW
424.0
424.0
231 .0
253.0
235.0
313.0
310.0
500. 0
3G4.0
433.0
406. O
513.0
325.0
77.0
325 . 0
325. 0
325. 0
325. O
25.
1 1003.
9719
6133,
S165.
1041 ,
2678.
9789,
4122.
6329.
1 1422.
6447.
1800.
2212.
1 143.
122G.
619.
33,
,0
, 0
, 0
0
0
,0
, 0
0
0
0
0
0
0
. 0
0
0
0
O
0
, 7
7.3
. 1
.9
7.3
10. 1
0
0
0
0
O
0
. 3
2. 1
4. 4
0
a
0
0
11.2
0
O
0
O
0
0
2.3
2. 5
0
.4
0
1O.O
O
0
5.3
SEPTIC
0
107.6
149.2
148.4
182.8
0
17. 1
51 .2
0
46.6
0
0
35.8
O
1 .9
0
12. 5
0
H I WAY
0
0
0
0
0
0
0
0
0
48.6
69.0
0
0
0
0
0
O
O
SKI SLOPE
O
0
25. 1
0
0
0
1 1 .5
0
5. J
0
19.6
O
0
0
0
0
O
O
PKG PLT
O
7.6
O
0
48.1
2.7
0
0
24.3
0
0
0
0
0
0
0
0
0
TERT PLT
314. 9
0
0
0
0
0
0
0
0
0
117.5
0
0
0
0
151 .8
0
155. 1
AMAX
0
0
0
0
0
0
0
0
0
0
0
259.7
0
0
0
0
0
0
1-1
1
BKGRND
.8
346. 3
133.0
S5. 1
72.3
22.2
54.8
386. 3
105.2
205.0
338. 7
263.5
39.3
6.7
25.0
26.8
13.5
.7
-------
i a i 7 . o
19.3
39. 7
GROSS N YIELD FOR WATERSHED SEGMENTS, KG PER YR
iTION NAME
BR1
BR2
BR4
BR6
BR7
SRI
SR2
SR3
SR4
TMJ
TM4
TM7
MC2
SC2
AAA
BOB
CCC
ODD
EEl"
MM RUNOFF
424 . 0
424. O
231 . O
253.0
235.0
319.0
310.0
500. O
364. 0
433. 0
4O6. O
513.0
325. O
//. 0
325. 0
325 0
325. 0
325. 0
325 . 0
AREA HA
25.0
1 1O03.O
9719.0
6133.0
5165.0
1 04 1 . 0
2678. 0
9789. 0
4122. 0
6329. 0
1 1422.0
6447.0
18OO. 0
2212. 0
1 143. O
1226.0
619. 0
33. 0
1817.0
RES SEW
O
36.3
403.6
3.3
52.0
377.7
524.6
0
O
0
O
0
0
23. 7
107. 0
227. 5
0
O
O
URB SEW
0
O
933. 5
0
O
0
0
0
0
165. 9
184.7
0
27. 9
0
775. 4
0
0
408. 5
0
SEPTIC
O
5952.6
7129.0
7248. 1
8771 .7
0
878.6
2945. 3
O
2588. 9
0
0
1856.6
O
100. 7
0
647 7
0
618. 9
HI WAY
0
0
0
0
0
0
0
0
0
664.7
1015.6
0
0
0
0
0
0
0
330. 1
SKI SLOPE
0
O
162. 1
0
0
0
74. 3
0
32.8
0
126. 8
O
O
0
0
0
0
0
O
PKG PLT
0
33.5
O
0
213.1
12.0
0
O
107. 7
0
0
0
O
0
0
0
O
0
0
TERT PLT
8019. 9
0
0
0
0
0
0
0
0
0
2992. 5
0
0
0
0
3066. 3
0
3950. 1
0
AMAX
0
0
0
0
0
0
0
0
0
0
0
52814. 5
O
0
O
0
0
0
0
BKORNO
22.7
9972. 5
4370.6
3063. 3
2369. 2
679.4
1691 .O
10731 . 5
3132. 8
5877. 0
9846. 8
720O 2
1200. 3
200. 0
762. 2
8)7.6
412.8
22. 0
1211 7
OKOb:> P Alii") N YIMD |4Y .MMIIIIT KG ITK Yl?
NAfIL" P10IAI- (IKJTAL
•-11 r, .in i <
-------
BR2
BFM
BR6
BR7
SRI
SR2
SR3
SR4
TH1
TM4
TM7
MC2
SC2
AAA
DBB
CCC
ODD
EEE
-162.
32G .
244.
304 .
32.
94.
437
135.
302.
547
523.
75.
7.
39.
183.
26.
161 .
71 .
15995.
12999.
10315.
1 1406.
1069.
3169.
13677.
3273.
9316.
14166.
60O95.
3005.
304.
1745.
491 1
1 OCO .
4381 .
2161
JRflSS KG PER YR P IN RUNOFF IS, 4435.
JROSS KG PER YR N IN FHJN5FF IS, 185155.
>RCSS YIELD BY SOURCE KG PER YR f(ABOVE),N(BELOW)
:eSIO. SEWER URB SEW SEPTIC HI WAY SKI SLOPE PKG PLT TERT PLT AMAX BKGRND PRECIP GRNDWAT DIVERSION CONSTRUCT
33.2 31.G 765.0 136.9 61.3 82.6 739.3 259.7 2175.1 726.0 10.0 0 150.0
1755.9 2496.0 38738.0 2030.3 396.1 366.3 18828.0 52814.5 63743.4 9872.0 1068.O 0 3985.7
IH
I
M
00
-------
VALUES UP TO THIS POINT UNCORRECTEO FOR
INVENTORY CHANGE IN RIVER BOTTOMS, NOW CORRECT THIS.
ET LOADING BY SOURCE KG PER YR P(ABOVE>,N(BELOW>
•.SID. SEWER URB SEW SEPTIC HI WAY SKI SLOPE PKG PLT
32.2 30.6 740.5 132.5 59.3 80.O
19O1.1 2702.4 41941.6 2198.2 428.8 396.6
TERT PLT AMAX
735.5 251.4
19076.3 57182.3
BKGRND PREC1P GRNDWAT
2105.5 726.0 10.0
69015.0 9872.0 1068.0
DIVERSION CONSTRUCT
0 145.2
0 4315.3
IE NET YIELDS BY PERCENT ARE P(ABOVE),N(BELOW)
.SID. SEWER
.6
. 9
URB SEW
.6
1 .3
SEPTIC
14.7
20. O
H1 WAY
2.6
1 .0
SKI SLOPE
1 .2
.2
PKG PLT TERT PLT AMAX BKGRND PRECIP GRNDWAT
1.6 14.6 5.0 41.7 14.4 .2
.2 9.1 27.2 32.8 4.7 .5
DIVERSION CONSTRUCT
0 2.9
0 2. 1
THE NET P YIELD TO THE LAKE IN KG PER YEAR IS, 5049.
THE NET N YIELD TO THE LAKE IN KG PER YEAR IS, 210098.
•MONTH*
1 .
2.
3.
4 .
5.
6.
7
8.
9.
1 O.
1 1
12.
»DAY*
15.
45.
74.
105.
1 35.
1GG.
190.
?£7
25B.
200
319.
3-1U.
*LEVEL FT* *MIL M3MO ROFF* »MIL M3MO DIVER* *MIL M3MO OUT* *KG P NONPREC* *KG P PREC*
8981 .
8977.
8974.
8974.
8980.
9OOO.
9018.
901 6.
9010.
9017
9O16.
9O15.
5.35
4.42
5.76
8.30
34. 29
92.61
56. 88
28.06
15. 28
10.50
6. 80
6. 76
0
0
0
0
0
0
0
0
0
0
0
0
12.54
11 13
9.07
4.68
3.82
6.15
30. 38
26.53
15.35
10. 17
13. 12
5. 79
84.2
69.6
90.6
130. 5
538.9
1455.7
893. 9
441 . 0
240.2
165. 1
107.0
106. 3
45. 7
45.7
45. 7
45. 7
45. 7
217.8
45.7
45. 7
45.
45.
45.
7
7
7
45.7
MEAN LAKL AREA OVL l< THE YEAR IN HECTARES IS 1130.
AVCKAOL" LAKE VOI.U1L OVER THE YEAR IN M3 IS 2623659GO.
TOTAL ANNUAL UiriOW III M3 INCLUDING DIVERSION IS l'750-IOOOO.
M
I
TOTAL ANNUAL OUHIOU IN M3 IS 148730OOO.
TOTAl lOADIUii IN M) I'EK DAY IS 13.819
AKLAI LOAD IN flu
701 Ill-It I Oft II IN 1-K
t'LK
I-LU
2 Pl-R YR l
in PER YR
44G. 2
19.2
-------
tAN DEPTH IN H IS 23.2
AU III YR IS .95
REDICTFD P CONCENTRATION IN TOP 15 H IS 7.9
MEAN CHLOROPHYLL A 0-5M DURING STRATIFICATON IS 7.97
MEAN SECCHI DEPTH IN H JULY-OCT IS, 2.69
MIN SECCHI DEPTH IN M JULY-OCT IS, 1.38
CARLSON TROPHIC STATE INDEX IS, 21.
THE AREAL HYPOLIM O2 DEFICIENCY IS, 631
THE SUMMER DEPLETION IN PPM IS, A. 4
THE MINIMUM 02 5M OVER THE BOTTOM IN PPM IS, 4.6
I
i-o
O
-------
DATE .
JOB
ACCT
1 983/CM/O1
18GG9 F253CCV
REEL. 140652 START..1641:30 STOP..1641:50
PAGES. .OO0022 STATUS. .8000800O
END-OF-REPORT
ENO-OF-REPORT
END-OF-REPORT
END-OF-REPORT
END-OF-REPORT
END-OK-REPORT
ENO-OF-REPORT
ENO-OF-REPORf
END-OF-REPORT
ENO-OF-REPORT
END-OF-REPORT
END-OF REPORT
ENO-OF-REPORT
END-GF-REPORT
END-OF-liEfJOF
-------