Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Comparison of Tracer-Dilution and Current-Meter Discharge Measurements in A Small Gravel-Bed Stream, Little Lost Man Creek, California

Download as pdf or txt
Download as pdf or txt
You are on page 1of 21

COMPARISON OF TRACER-DILUTION AND

CURRENT-METER DISCHARGE MEASUREMENTS


IN A SMALL GRAVEL-BED STREAM,
LITTLE LOST MAN CREEK, CALIFORNIA

By Gary W. Zellweger, Ronald J. Avanzino, and Kenneth E. Bencala

U. S

GEOLOGICAL SURVEY

Water-Resources Investigations Report 89-4150

Menlo Park, California


1989

TABLE OF CONTENTS

ABSTRACT. . .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

INTRODUCTION .... . ........ . . . ........ .. . .......... . .... . .. . ..

Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Purpose and Scope... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

FIELD SITE

EXPERIMENT

TRACER-DILUTION DISCHARGE METHOD ....... . ............ . ...... .

CURRENT-METER DISCHARGE METHOD ....... . .... ... .. . . . .... . .... .

10

COMPARISON OF METHODS .... . . . .... . . .. . . .. . ..... , .. . ...... . . .. .

10

DISCUSSION

16

CONCLUSION

18

REFERENCES

19

LIST OF ILLUSTRATIONS

Figure 1.
A map of the study ,reach.. . .......... . ........

~ . . .. . ... ..

Figure 3.
Chloride concentration curve at Site "224m" . . .

11

Figure 4.
Discharge from current-meter measurements. .. . .

12

Figure 5.
Longitudinal variation in discharge .... ... ....

14

Figure 6.
Discharge from three tracers .. . . . . . . .. . .... .. .

15

Figure 7.
Summary of discharge at Site "330m" ..... . . . ...

17

Figure 2.
Ion tracer 1imits ....... . ..........

iii

" _..

. . '- --

--

' . --

- - - - - --

._---- - - - -

'.

_._-_ -.- .. .._- ,- - - - -- - - - -


..

Comparison of Tracer-Dilution and Current-Meter


Discharge Measurements in a Small Gravel-Bed Stream,
Little Lost Man Creek, California
By Gary W. Zellweger, Ronald J. Avanzino, and Kenneth E. Bencala
ABSTRACT
Discharge was determined in a small gravel-bed stream, Little Lost Man
Creek, California, using both a current meter and two tracer dilution
measurements: one with a mixing length of. 25 meters and the other with a
mixing length of 300 meters.

Comparison of discharge values obtained

indicated that as much as 25 percent of the channel discharge may have been
flowing as underflow through gravel zones in the channel.

Current-meter

discharge measurements were used to obtain point discharge; tracer-dilution


discharge measurements were used to obtain discharge measurements which
incorporated mixing of surface water and underflow over 25 and 300 meters of
stream reach.
Within the study reach, the confined channel contained gravel and
cobble to a depth of at least one meter.

During summer low flow of 5 to 20

Lis, the surface stream appeared to be I to 2 meters wide, about 20 percent of


the channel width.
Determining discharge in small pool-and-riffle streams using conventional
current-meter techniques is hampered by shallow depths, rough bottoms, and
flow through the gravel bed under and parallel to the stream.

Furthermore,

discharge can vary both diurnally and longitudinally over relatively short
reaches.

Discharge measurement by continuous tracer dilution can accommodate

these factors to some extent.


Chloride was injected for eight days during which the stream water was
sampled along a 300-meter reach below the injection point.

On the seventh

day, a second injection of sodium, chloride, and rhodamine WT dye was made for
24 hours, 305 meters below the first injection, and was sampled 25 meters
downstream.

Current-meter discharge measurements were made along the study

reach three times during the experimental period.

__

....

_ - -

The average discharge of 17.4 Lis calculated from the tracer that
traveled 300 meters was greater than the average of 15.9 Lis calculated for
the tracer that traveled 25 meters.

Both of these tracer-dilution

measurements were greater than the average current-meter discharge measurement


of 13.0 Lis.

The comparison supports the hypothesis that a significant part

of total stream discharge is moving as underflow through gravel zones within


the stream channel.

Figure 1. A sketch of the study reach on Little Lost Man


Creek.
The primary injection was at Site "Om", at the bottom
of the map, the secondary injection was at Site "305m" and was
sampled at
Site "330m". Routine sample sites were at "-2m",
"52m", "99m", "137m" and "224m" and current meter sites at
"62m", "150m", "224m", and "325m" .

Purpose and Scope

This report was written with two purposes.

First, data are presented

showing differences between measurements of discharge in a small stream made


by the current-meter method and the tracer-dilution method.

The second

purpose is to suggest the possibility that the differences between the


measurements may be due to as much as four liters per second, or 25 percent of
discharge, flowing through channel gravel.

FIELD SITE

Little Lost Man Creek is a third-order, pool-and-riffle, coastal stream


within Redwood National Park, in northwestern California, about 40 miles north
of the city of Eureka.
area of 9.4 km

The 10-km, north-northwest flowing stream drains an

2
(3.6 mi ) and is at elevations between 24 and 695 meters above

sea level (Iwatsubo and others, 1975).

Discharge ranges from typical late

summer lows of 6 Lis to a winter high of 5,700 Lis.

Average annual rainfall

is 1,780 mm at the Orick Prairie Creek rain gage, approximately 3 km northwest


of the study reach.

The study reach was 330 meters long, located in the lower

section of the stream where the gradient was 0.018.

The streambed sediments

were poorly sorted and ranged from sand to boulder size.

There was little

clay and silt in the streambed permitting considerable water flow through the
gravel (Bencala and others, 1984).

Of five shallow wells dug into the gravel

bed in and beside the stream, two of which extended a meter below the stream
bottom, none reached the lower limit of the gravel deposit, indicating that
the gravel sediments, in these locations, were at least one meter thick.
precipitation occurred during the study period.

No

Background concentrations of

sodium and chloride just above the study reach were each approximately 6 mg/L.
Figure 1 is a map of the study reach.

The sample sites along the reach

were identified by their distance, in meters, below the lithium chloride


injection site.

Routinely sampled sites were located two meters above the

injection (Site 1-2m") and at 20, 52, 99, 137, and 224 meters below the
injection.

Current-meter discharge measurements were made 62, 150, 224, and

325 meters below the injection site.

EXPERIMENT

The primary tracer was a solution of lithium chloride pumped into the
stream continuously beginning at 2: 30 p . m. PDT on August 15, 1984 and ending
at 10 : 00 a.m. on August 23.
170.1 gi L.

The chloride concentration of the inje c tant was

The injection rate was maintained using a battery-powered, 12

volt, metering pump.

Daily measurements of the injection rate, made by timing

the filling of a 50-mL volumetric flask, averaged 37.29 .32 (1 sd) mL/min.
On the seventh day of the eight-day injection, a secondary injection of
sodium, chloride, and rhodamine WT was made for 24 hours, starting at 11:02
a.m. on August 21, 305 meters below the primary injection.

This secondary

injection was intended for comparison with discharge calculations made from
the primary injection.

The stream was sampled hourly above the injection site

at 300m, and below at 330m, using automatic samplers.


previously been

Rhodamine WT loses have

observed within Little Lost Man Creek (Bencala and others,

1983), however in this study the dye was used over a short mixing r each.
On August 14 , 21 , and 28, personnel of the U.S. Geological Survey field
office in Eureka, California, measured discharge with a current meter
following routine procedures (Buchanan and Somers, 1969).

Two measurements

were made on each date at each site usually three to four hours apart.
Samples collected for chloride and sodium analy sis were filtered through
a 0.45

~m

membrane and stored in darkness until analyzed a few months later.

Samples for rhodamine WT analysis were collected in glass bottles and kept in
darkness until analyzed on August 25, 1984.
Samples were analyzed for chloride using a Dionex ion chromatograph

Sodium was analyzed by atomic absorption spectroscopy using a Perkin-Elmer


model 303 spectrophotometer, following the method described by Skougstad and
others (1975).

Rhodamine WT fluorescence was measured with a Turner Designs

Model 10 fluorometer.

Because lithium is known to adsorb to sediments in this

stream in as few as 62 meters (Bencala and others, 1984) , it was not measured
in this experiment.

1. Use of a product name in this report is for identification purposes only


and does not constitute endorsement by the U.S. Geological Survey
6

--------- -

--

TRACER-DILUTION DISCHARGE METHOD

Discharge can be measured from dilution of a tracer by equating the


increase in the stream tracer mass to the mass being added and solving for
discharge (Kilpatrick and Cobb, 1985).

All tracer-dilution discharge

measurements in this paper were made using continuous injection.

For this

method, the tracer is injected as a concentrated solution at a constant rate


for an extended time.

Downstream of the injection site, the tracer

concentration in the stream increases as the tracer arrives and eventually


stabilizes at a nominally constant concentration referred to as the plateau
concentration.

The plateau concentration of the tracer reflects dilution of

the tracer by all of the water with which the tracer has mixed between the
injection point and the sample point.

This method requires that the tracer is

thoroughly mixed with the stream, and that it is conservative.

A conservative

tracer neither gains nor loses mass by any chemical or physical action within
the study reach.

When this is the case, the tracer dilution reflects the

discharge of the stream.

Discharge below the injection point is calculated as

(C - C )
i
a

(C - C )
b
a
where:

Qb

Stream discharge below the injection point

Qi

Injectant discharge

C.

Tracer concentration in injectant

C
a

Tracer concentration in stream above injection point

C
b

Tracer concentration in stream below inj ection point

l.

Discharge determination is not limited to one site near the injection.


The tracer-dilution method allows discharge to be calculated at any downstream
site where the tracer reaches plateau.

Temporal patterns may be determined by

sampling as long as the injection continues.

1000
0

\\

UJ

C/)

a:
UJ
a..
C/)
a:
UJ

:::J

Na-CI

\\

750

\\

'\
'\

\\

500

'\

'\

~.

'\

'\

', ..........

~Li-CI

UJ
(9

a:
250

"" -----

" " ,CI-Na

".

..........

.....

--

""
CI-LI

............

~--.

C/)

..

-- ..

--- - - -=-=--===-..:.:=::.:.=:="

DURATION, IN DAYS

Figure 2.
Limits on the use of ionic tracers assuming an
injection rate of 1 Llmin, maximum tank capacity of 1400 L, a
$500 budget restriction and a tracer increase equivalent to 5
mglL of chloride. Different analytical sensitivities have been
taken into account. The notation Cl-Li indicates the tracer is
chloride from the salt LiCI. The pump rate is exceeded for GI
as NaGI at discharges above 750 Lis.

Chloride is useful as a tracer in small streams because it behaves


conservatively and is easy to sample and to measure.

In typical streams,

chloride is not appreciably involved in ionic exchange with the streambed


(Kennedy and others, 1984).

As noted by Feth (1981) "The chemical behavior of

chloride in natural water is very tame and subdued

***

an element that is

characterized by its existence in ionic form, and one that is stable with
respect to chemical reactions that greatly influence the behavior of many
other common elements."

When ion chromatography is used to measure chloride,

relative standard deviation is at most 1 to 2 percent at levels of a few


milligrams per liter and analytical interference due to other solutes is rare.
Sample filtration is necessary only to prevent clogging of chromatographic
columns and can be done by the chromatograph at the time of analysis.

There

is normally no need to store the samples chilled.


Sodium can exchange with other cations at charged sites on sediments.
However, it is a monovalent cation and is a weak competitor for exchange
sites.

Sodium sometimes behaves conservatively over short distances (Bencala,

1985).
The usefulness of anion tracers is limited by background concentrations
and stream size.

Calculations used with tracers require subtracting any

background concentration from the total concentration.


the signal concentration nears background.

Precision decreases as

Quality results can be obtained

when the total tracer concentration is at least double background


concentration.
Because stream concentrations must be increased by milligrams per liter,
the rate and total amount of added tracer bec,omes limiting as stream size or
injection duration increases.

If pumping rates are limited to 1 L/min and

storage tanks limited to about 1,400 L, then anion tracers become impractical
3
in streams with flows greater than 600 L/s (50 ft /s) especially for
injections of 24 hours or longer.

These limits can be seen graphically in

figure 2 which includes consideration of salt solubility and measurement


sensitivity.
Rhodamine WT fluorescent dye often is a suitable tracer for streams
without high concentrations of suspended sediment or large amounts of organic
material.

Rhodamine WT can be used in larger streams where anionic tracers

are not practical and where adsorption by stream sediments and organic

materials is minimized.

The dye can be used at concentrations of a few

micrograms per liter; Smart and Laidlaw (1977) reported a minimum detection
level of 0.013 ~g/L in distilled water on a Turner Model 111 fluorometer.
Portable battery-powered fluorometers can be used to measure fluorescence in
the field.

CURRENT-METER DISCHARGE METHOD

Discharge measurements were made on Little Lost Man Creek using routine
U.S. Geological Survey methods (Buchanan and Somers, 1969).

A Price pygmy

meter, which is more suitable for shallow streams than the standard Price type
AA current meter, was used to measure stream velocities.

The stream was

modified at four locations by removing large cobbles from the stream channel,
creating a short reach of uniform width and relatively even flowing water.

measuring tape was secured across the channel of 1 to 2 meters width, and the
depth and average water velocity were measured at 17 to 25 vertical sections
across the width of the channel.

Summing the flows through each subsection

determined stream discharge.

COMPARISON OF METHODS

Figure 3 is a plot of the concentration of chloride measured in samples


from Site "224m".

The rapidly rising left side of the concentration curve

represents samples collected while the chloride concentration


toward the plateau level.

was increasing

That chloride required about two days to reach

plateau indicated that there was a considerable volume of water in the stream
channel that was not in immediate contact with the visible surface stream.
Once plateau was reached, the concentration curve showed a slow steady
increase, as would be expected during a period of no precipitation.
The averages of each pair of current-meter discharge measurements are
plotted against downstream distance in figure 4.

These data show a decrease

in discharge with time and a variation along the length of the stream.

An

estimate of the precision of the current-meter method was made by pooling all
of the 11 sets of two measurements from the three times the stream was
measured.

This estimate assumes that the stream discharge did not change

10

- - - - -- - - - -

cr:
UJ

14

::J
cr:
UJ

8 I
6

I-

10

c..

cr:

Z
0

..-- -

.,-

f-

2
0

Chloride
Site 224m

f--

14

f-

Z
0

~ 4 f--

.. - .. -
I

cr:

zUJ

(!J
.....J

,.
-

12

(f)

16

20

18

22

24

AUGUST

Figure 3.
A plot of the chloride concentration at Site
"224m" . The concentration rises soon after the start of the
injection at 2:30 pm, August 15 and continues to rise reaching
a "plateau" about two days later.

11

20 ,---------,----------,---------,----------,

Aug. 14

15

()

(fJ

oc

Aug. 28

ill

a..

(fJ

oc
w
f- 10

:J

c:>

oc

<{

()
(fJ

(5

o ~--------~--------~--------~--------~
o

100
200
300
DISTANCE DOWNSTREAM, IN METERS

400

Figure 4. Longitudinal discharge for the averages of pairs


of current-meter discharge measurements made on August 14, 21,
and 28, at the four (three on Aug. 14) metering sites.
12
--- ..- . ---_.. - .. . _-_ ....._------------ - - - - - -- ---.-

- .. - - - - .... - _._--_._--_.

substantially during the 3 to 4 hours between the two measurements.

The

measurements ranged from 10 to 19 Lis and the estimated standard deviation of


the method was calculated to be 1.4 Lis for one measurement and 1.0 Lis for
the average of two.

There is no independent measurement of discharge in this

experiment that can be used to determine if there is a systematic under or


over registering of discharge by the current-meter method.
The decrease in discharge between August 14 and the two other dates is
much greater than the precision of the current-meter method and is reasonable
for a small stream recovering from the last rain on July 7.

The longitudinal

variation could easily be due to differences between the prepared gaging


sites.
The average values of current-meter discharge for August 21 are shown
again in figure S, along with discharge, for the same date, calculated for six
sites from dilution of the primary chloride tracer.

The discharges calculated

from tracer dilution were clearly greater than those from current-meter
measurements at each of the four sites.
Samples from the secondary injection of August 21/22, collected at "Site
330m", were analyzed for sodium, chloride, and rhodamine WT.

Discharge was

calculated from each hourly sample and the corresponding background sample
from site "300m".

The average calculated discharges were lS.98 0.2S Lis for

chloride, lS.88 O.SO Lis for sodium, and lS.78 0.S7 Lis for rhodamine WT.
The precision values are one standard deviation and assume that no change in
discharge occurred during the 24-hour injection.

These discharge values are

shown in figure 6 along with the discharge values from the two current-meter
measurements which averaged 13.0 Lis.

Again" as in figure 5, the current

meter discharge values are less than the tracer-dilution discharge values.
The agreement between the three sets of tracer-discharge calculations
indicates that the analytical techniques used were precise.

This does not

necessarily mean that the discharge calculations are accurate.

All three

tracers would have had the same hydraulic behavior, so to the extent that
mixing was not complete, each tracer would have been equally affected.

13

- . - - _.-

- - --

- --

- - -- - - -

._- -

. -- -- - - - -- - -

---

- - - _.- _.

_ _._-

- - -------

, - - - - - --

20

Chloride dilution~

+ + +
0

-t-

15

0
0

ill
(f)

a:

Current meter /

ill
CL

(f)

a:

ill

f-

::::i

10 -

ill

(')

a:

S2

5 I-

OL-____~I------~I------~I~----~

100

200

300

400

DISTANCE DOWNSTREAM, IN METERS

Figure 5. Longitudinal discharge downstream of the primary


chloride injection on August 21. The plus symbols represent
discharge measured by chloride dilution from the primary
injection;
the squares represent discharge measured by current
meter.

14

- - - _.

_ . _ - --

20

0
0
w

(/)

a:

15 I-

+$~~~~~~1~~&~t~~~~~~~Ob.

n..

(/)

a:
w

I- 10 f

:::i

ui
CJ

a:

<{

I
0

51

Chloride

Sodium

Rhodamine

b. Current meter

(/)

0
0
0600

1200

1800

0000

0600

AUGUST 22

AUGUST 21

1200

Figure 6.
Discharge for 24 hours, near "Site 330m", as
calculated from the three tracers of the secondary injection on
August 21/22.
Squares, plus signs, and diamonds represent
discharge by dilution of
Cl,
Na,
and
rhodamine
WT,
respectively.
The triangles represent the two current-meter
discharge measurements.

15

Agreement of the sodium discharge calculations with those of chloride and


rhodamine WT is evidence that sodium can sometimes act as a conservative ion.
For the 2S-meter reach of the secondary injection , in which the observed
travel time appeared to be less than S minutes, the discharges calculated from
sodium were not different from those calculated from chloride or from
rhodamine WT.

Either the rate of adsorption was so low as to be trivial, or

the reaction was fast and reached equilibrium within 13 minutes before the
first sample was collected.

Whichever the case, for this e x periment, sodium

behaved essentially as a conservative tracer.


The hourly background samples from Site "300m" were also used as the
downstream site to calculate discharge from the primary lithium chloride
injection using the chloride analyses from Site "-2m", as the site above the
injection.

For this calculation the chloride tracer traveled 300 meters.

The

average of these values, 17 . 40 O.lS Lis, along with the averages from figure
6, are shown in figure 7 .

DISCUSSION

The discharges calculated from the averages of the three methods are 13.0

Lis for the current-meter measurements, 15.9 Lis for tracer dilution over 25
meters, and 17.4 Lis for tracer dilution over 300 meters.

The differences are

great enough, when compared with their respective standard deviations of 1.0,
0.46, and 0.15 Lis, to indicate actual differences in discharge.
The explanation may be that there is enough water flowing through the
coarse gravel deposits in the stream channel to account for the differences in
discharge measurements.

In this concept, the gravel zone, which is several

times wider than the surface stream during the low flow season and at least
one meter thick in some locations, might carry as much as 25 percent of the
total channel flow (4 Lis).

Water does not travel permanently in the gravel

zone but moves from the gravel zone into the surface stream in some reaches
and from the stream into the gravel zone in other reaches.
distances greater mixing occurs.

Thus, over longer

The current meter, which measures only the

surface flow, would miss all of the underflow water.

The tracers traveling 25

meters would have some opportunity to mix with underflow water, and the tracer
traveling 300 meters would have the greatest opportunity to mix with underflow
water.

16

..

-- ._-_._ - - - - - - - - - -

20

~---------------------------------,

300

25m

o
() 15 f------

ill

(/)

a:

ill
0..

a:

ill
I

:::::i 101---
Z

ill

(9

a:

f------

0"---
CL

Na

CI

Dilution method

Rh

Current
meter
method

Figure 7.
Summary of measurement of discharge near Site
"330m" by three methods.
First, by chloride dilution over 300
meters of reach; second, by dilution of sodium, chloride, and
rhodamine WT over 25 meters of reach; and third, by current
meter method.

17

CONCLUSION

It has long been known that water occupies inter-gravel spaces in steam
channels and presumably moves down the channel as underflow.

In some systems,

that underflow can be large enough to measure.

when such flow

Furthermore,

does occur it can affect discharge measurements to different degrees,


depending on the method used to measure discharge.

Not only can tracer

dilution and current-meter methods result indifferent values, but the tracer
dilution method seems to produce different results when used over different
stream lengths.

18

-- - ----- ---

---- ----- - -- -- -- -- - -

REFERENCES

Bencala,

K. E., Rathbun, R. E., Jackman, A. P., Kennedy, V. C., Zellweger, G.

W., and Avanzino, R. J, 1983, Rhodamine WT dye losses in a mountain stream


environment: Water Resources Bulletin, v. 19, no. 6, p. 943-950.

V.

Bencala, K. E., Kennedy,


Avanzino,

R.

J.,

1984,

C.,

Zellweger ,

G.

W.,

Jackman,

A.

P.,

and

Interactions of solutes and streambed sediment.

Part 1. An experimental analysis

of

cation

and

anion

transport

in

mountain stream: Water Resources Research, v. 20, no. 12, p. 1797-1803.

Bencala,

K.

1985,

E.,

Experimental

and

Performance

simulation

of

sodium

as

transport

tracer-

analysis. Selected Papers in the Hydrologic

Sciences: U.S. Geological Survey Water Supply Paper 2270, p. 83-89.

Buchanan,

T.

stations:

J.,
U.S.

and

Somers, W. P., 1969, Discharge measurements at gaging

Geological

Survey,

Techniques

of

Water-Resources

Investigations, Book 3, Chapter A8, 65 p.

Feth, J. H., 1981, Chloride in natural

continental

water--A

review:

U.

S.

Janda,

R.

Geological Survey Wa-ter-Supply Paper 2176, 30 p.

Iwatsubo, R. T., Nolan, K. M., Harden, D. R., Glysson, G. D.,


J.,

1975,

and

Redwood National Park Studies: Data Release Number 1.

Redwood

Creek Humboldt County, California, September 1, 1973-April 10, 1974:

U.S.

Geological Survey Open-File Report.

Jarrett, R. D., 1984, Hydraulics of


Hydraulics

Division,

American

high-gradient

streams :

Journal

of

the

Society of Civil Engineers, 110, (11), p .

1519-1539.

Kennedy,

V. C., Jackman, A. P., Zand, S. M., and Zel1weger, G. W., 1984/1985,

Transport and concentration controls for


and

lead

chloride,

strontium,

potassium

in Uvas Creek, a small cobble-bed stream in Santa Clara County,

California, U.S.A. 1. Conceptual model: Journal of Hydrology, v. 75 p. 67


llO.

19

~ ..

Kilpatrick, F. A., and Cobb, E.

u.s.

tracers:

D.,

Geological

1985,

Survey,

Measurement

of

Techniques

of

discharge

using

Water-Resources

Investigations,
Book 3, Chapter A16, 52 p .
,
Kondolf,

G.

M., Webb, J. W., Felando, Thomas, 1987, Basic hydrologic studies

for assessing impacts of flow diversions on riparian vegetation:


from

Examples

streams of the eastern Sierra Nevada, California, USA: Environmental

Management, 11, no. 6, p. 757-769.

Marchand,

J.

P.,

water-surface

Jarrett,
slope,

R. D., and Jones, L. L., 1984, Velocity profile,

and

bed-material

size

for

selected

streams

in

Colorado: U.S. Geological Survey Open-File Report 84-733, 82p.

Skougstad, M. W., Fishman, M. J., Friedman, L. C., and Erdmann, D.


Methods

for

sediments:

determination
U.S.

Geological

of

E.,

1979,

inorganic substances in water and fluvial


Survey,

Techniques

of

Water-Resources

Investigations, Book 5, Chapter Al, 626 p.

Smart, P. L., and Laidlaw, I. M. S . , 1977, An evaluation of

some

fluorescent

dyes for water tracing: Water Resources Research, v. 13, no. 1, p. 15-33.

20

- - - - --

You might also like