Spring 2007

 

CE 394K.2

 

Ex 1

 

Prepared by,

 

Ernest To and David Maidment

 

20070130
CE 394K.2  Exercise 1:

 

Analyzing water data from the USGS NWIS and EPA Storet Systems

 

 

1.   Find the names and code numbers of the NWIS and Storet stations at Neuse River near Clayton, North Carolina.  Find the period of data and number of available records for NWIS streamflow  and Storet Nitrate + Nitrite as Nitrogen (note the NWIS streamflow data actually go from the given start date right up to the present). 

 

Table 1:  Names, code numbers and periods of records for stations located at Neuse River near Clayton, North Carolina

 

Source and Network

SiteName

SiteCode

Variable

Period of Record

USGS NWIS

NEUSE RIVER NEAR CLAYTON, NC

02087500

Daily discharge, cubic feet per second

08/01/1927 to present

(Value count = 29036 samples)

EPA Storet

NEUSE RIV AT HWY 42 NR CLAYTON

21NC01WQ:J4170000

Nitrogen, Nitrite (NO2) + Nitrate (NO3) as N

08/17/1972 to 12/30/1996

(Value count = 270 samples)

EPA Storet

NEUSE RIV AT NC 42 NR CLAYTON

21NC02WQ:J4170000

Nitrogen, Nitrite (NO2) + Nitrate (NO3) as N

01/06/1997 to 08/14/2001

(Value count = 124)

 

2.  Make a plot of the time series of the whole record from 1927 to present, and cumulative frequency, frequency histogram and box and whisker plots of these data.   Make screen captures of these plots and turn them in as part of your solution. 

 

Fig 1.  Time series of daily discharge (cfs) at Neuse River near Clayton, NC for entire period of record (08/01/1927 to present).

 

Fig 2.  Cumulative frequency of daily discharge (cfs) at Neuse River near Clayton, NC for entire period of record (08/01/1927 to present).

Fig 3.  Frequency histogram of daily discharge (cfs) at Neuse River near Clayton, NC for entire period of record (08/01/1927 to present).

 

Fig 4.  Box and whiskers plot of daily discharge (cfs) at Neuse River near Clayton, NC for entire period of record (08/01/1927 to present).

 

3. Use HIS Server to identify the period of streamflow record at another USGS station somewhere else in the nation and make similar plots for that station.  

 

The location of the USGS station at Neuse River at Smithfield is shown in the map below:

Fig 5.  Location of the Neuse River at Smithfield, NC gage.

 

The USGS gage at Neuse River near Smithfield,NC (02087570) is located approximately 27 miles downstream from Neuse River near Clayton, NC.  Data are only available from 10/01/1970 to 09/30/1991 for Neuse River near Smithfield, NC.  To make sure that we are not comparing apples to oranges, only data from the same period in time at Clayton, NC were used for comparison between the two gages. 

 

The following graphs show the flow characteristics of the two gages from 10/01/1970 to 09/30/1991:

 

 

 

Fig 6.  Comparison of time series of daily discharge (cfs) at Neuse River at Clayton, NC and Smithfield, NC from 1970 to 1991..

 

Fig 7.  Comparison of cumulative frequencies of daily discharge (cfs) at Neuse River at Clayton, NC and Smithfield, NC from 1970 to 1991.

 

Fig 8.  Comparison of frequency histograms of daily discharge (cfs) at Neuse River at Clayton, NC and Smithfield, NC from 1970 to 1991.

Fig 9.  Comparison of box and whisker plots of daily discharge (cfs) at Neuse River at Clayton, NC and Smithfield, NC from 1970 to 1991.

 

The following table compares the flow characteristics at the two gages:

 

Table 2:  Comparison of flow characteristics at Neuse River nr Clayton and Neuse River at Smithfield.

 

Smithfield, NC (~27 miles downstream)

Clayton, NC (upstream)

Arithmetic Mean

1304

1116

Geometric Mean

695.9

604.5

Maximum

15800

16100

Minimum

65

57

Standard Deviation

1761

1552

Coefficient of Variation

135%

139%

10th percentile

198

178

25th percentile

306

283

Median

587

490

75th percentile

1430

1220

90th percentile

3560

3120

# Observations

7670

7670

 

The Neuse River gains an average of 188 cfs (14% of total flow) as it travels from Clayton to Smithfield (~27 miles).  The shapes of the cumulative frequencies and histograms at both stations are positively skewed which suggests lognormal distributions. The seasonal patterns are similar – both stations experience high flows in March and low flows with scattered storm events in July.  From the map, it is observed that no major tributaries enter the Neuse River between Clayton and Smithfield; and thus the gain in flow is largely attributed to runoff from the surrounding catchment.

 

4.  Study the USGS and EPA variable code lists, as given at http://nwis.waterdata.usgs.gov/usa/nwis/pmcodes and http://www.epa.gov/Storet/legacy/ref_tables.htm for the way that they each characterize nitrogen data.  Describe how each specifies Nitrogen data and which variable code values would have to be used to extract the N data from each of these systems.  Get the Nitrate + Nitrite as N data from NWIS Instantaneous Irregular values using the NWIS Analyst download option.   The complete list of these codes is given as an Excel file at: http://www.ce.utexas.edu/prof/maidment/Ex1/VariableCodes.xls

 

Assumption:  Per the instructions in class on 01/30/2007, it is assumed that “Nitrogen data” refers to the combined concentration of nitrite and nitrate for this problem. 

 

I noticed that the parameters codes for EPA Storet shown in

http://www.epa.gov/STORET/legacy/parameter.zip were different from the Storet codes in VariableCodes.xls.  However, they are similar to the USGS codes in the same spreadsheet.  For the sake of consistency I used information in the VariableCodes.xls spreadsheet to answer this question.

 

The various forms of nitrogen (Total Nitrogen, Ammonia, Nitrite, Nitrate, Organic and Inorganic Nitrogen, etc) are referenced by Storet codes 327 through 336 and USGS codes 600 through 636.  EPA Storet assigns a variable code to each unique chemical type.  For instance, the storet code for nitrogen, nitrite plus nitrate is 336.  USGS is more specific than EPA in that it assigns a variable code not just to each unique chemical type but each unique combinations of chemical type, medium type (e.g. suspended solids and filtered water) and units (e.g deg C versus deg F).  For instance, the USGS code for Nitrite plus nitrate, water, unfiltered, milligrams per liter as nitrogen is 630 while the the USGS code for Nitrite plus nitrate, water, filtered, milligrams per liter as nitrogen is 631.

 

To extract Nitrogen data for the calculation of total nitrite and nitrate load in the Neuse River near Clayton, the most direct method is to use Storet code 336 or USGS code 630.  USGS code 630 is preferred over other USGS measurements in other media (e.g. filtered water [631] and suspended sediment [628]) because the unfiltered water sample accounts for the entire water column and thus the entire load carried by the river. 

 

In the absence of Storet code 336 or USGS code 630, one will have to take into account the relationship between the different variable codes in order to estimate the nitrite and nitrate concentrations:

 

For instance, by definition we know that:

[NO3- and NO2-] as N = [NO3-] as N + [NO2-] as N

Therefore Storet Code 336 =  Storet Code 334 + Storet Code 335

 

By definition, we also know that:

[Nitrite plus nitrate, water, unfiltered] = [Nitrite plus nitrate, suspended sediment] + [Nitrite plus nitrate, water, filtered, milligrams per liter as nitrogen]

Therefore USGS Code 630 = USGS Code 631 + USGS Code 628

 

Based on the definition of inorganic nitrogen we know that:

[NO3- and NO2-] as N =  [Total Inorganic Nitrogen] as N - [NH4+] as N

Therefore Storet Code 336 =  Storet Code 332 - Storet Code 330

 

5.  Nutrient enrichment from the Neuse watershed is leading to algal blooms and contamination of the Albemarle-Pamlico Sound downstream.   By combining the streamflow and water quality data for Nitrate and Nitrite as Nitrogen, make an estimate of the annual Nitrogen load flowing in the Neuse River at Clayton, NC (kg/year).

 

The first step was to develop a rating curve for Nitrogen (Nitrite and Nitrate) Load (kg/yr) vs. Flow (cfs).  The initial intention was to multiply instantaneous flow data at the time of the nitrogen sampling event with the measured nitrite and nitrate concentration to get the instantaneous load in the river.  Unfortunately, USGS instantaneous data collected before the past 31 days were not available.  As a compromise, historical daily discharge flows were used instead to calculate the loads.

 

The nitrogen load kg/yr for each sampling event was calculated using the following formula:

 

Load (kg/yr) = Nitrite and Nitrate Concentration (mg as N/L) * Daily discharge on day of sampling (cfs) * 1e-6 kg/mg * 28.317 L/ft3 *86400 s/d * 365.25 d/yr

 

The resulting loads are shown in the appendix of this report.


The loads calculated for the 81 nitrite and nitrate sampling events were plotted in the load vs flow graph below.  Since this is a simple analysis, a linear model was assumed for the rating curve.  If a more sophisticated model is desired, the reader is encouraged to look into USGS’s LOADEST software(http://pubs.usgs.gov/tm/2005/tm4A5/pdf/508final.pdf) which uses more complex regression analysis to determine the rating curve.

 

Based on linear regression, the formula of the rating curve was determined to be:

 

Load (kg/yr) = 304.78 * Flow (cfs)                  (Y-intercept for was forced to zero to better honor the smaller load values)

Fig 10.   Rating curve for Nitrogen load vs Flow.

 

The next step was to apply the rating curve formula on the daily discharge values at Clayton, NC to estimate the daily load.  Because nitrite and nitrate data were collected between 08/29/1973 and 09/13/2006, I only applied the rating curve to discharge values measured within this period.  A comparison of the estimated daily loads (based on the rating curve) with the measured daily loads is shown in the graph below.  The graph shows fairly reasonable agreement between measured and estimated loads.

 

Annual average =

322,000 kg as N/yr
 

Fig 11.   Comparison of loads estimated using rating curve with measured loads.

 

The annual Nitrogen load flowing in the Neuse River at Clayton, NC was estimated by averaging the daily loads (kg/yr) throughout this period.  The resulting annual Nitrogen load was 322,000 kg as N/yr.

 

 

 

 

 

 

 

 

 

Appendix

 

                                   

Loads estimated at Nitrogen sampling events

Station

Date

Variable Code

Flow (cfs)

Nitrite and Nitrate (mg as N/L)

Flow (L/s)

Load (kg/yr)

2087500

8/29/1973 13:30

630

354

0.9

10024.218

284706.2358

2087500

9/12/1973 15:15

630

192

1.2

5436.864

205889.2552

2087500

10/30/1973 9:15

630

202

0.66

5720.034

119136.9597

2087500

1/15/1974 9:20

630

788

1

22313.796

704169.8486

2087500

1/29/1974 9:45

630

4820

0.57

136487.94

2455122.135

2087500

1/30/1974 9:30

630

5930

0.45

167919.81

2384615.788

2087500

2/1/1974 10:25

630

3000

0.51

84951

1367233.336

2087500

4/4/1974 13:30

630

1490

0.52

42192.33

692374.1101

2087500

5/12/1974 13:45

630

3600

0.6

101941.2

1930211.768

2087500

5/13/1974 15:30

630

4070

0.49

115250.19

1782139.504

2087500

5/16/1974 10:00

630

3070

0.41

86933.19

1124795.163

2087500

5/20/1974 9:30

630

895

1

25343.715

799786.8205

2087500

6/21/1974 9:45

630

343

1.3

9712.731

398463.6237

2087500

8/4/1974 15:30

630

2270

0.9

64279.59

1825658.63

2087500

8/4/1974 18:20

630

2270

0.87

64279.59

1764803.343

2087500

8/5/1974 9:20

630

2270

0.39

64279.59

791118.7399

2087500

8/7/1974 9:25

630

4450

0.38

126010.65

1511105.602

2087500

8/7/1974 14:20

630

4450

0.5

126010.65

1988296.844

2087500

8/10/1974 16:00

630

700

0.54

19821.9

337787.0594

2087500

10/15/1974 10:20

630

202

1

5720.034

180510.545

2087500

1/15/1975 10:30

630

9360

0.08

265047.12

669140.0795

2087500

1/22/1975 12:25

630

3240

0.42

91747.08

1216033.414

2087500

3/19/1975 14:30

630

16100

0.24

455903.7

3452934.385

2087500

4/28/1975 14:45

630

537

0.94

15206.229

451079.7668

2087500

6/9/1975 11:05

630

277

1.7

7843.809

420804.0377

2087500

7/16/1975 14:30

630

11800

0.26

334140.6

2741615.604

2087500

8/26/1975 11:34

630

233

1.7

6597.861

353961.5191

2087500

9/23/1975 9:10

630

592

0.96

16763.664

507860.1629

2087500

9/25/1975 9:45

630

3090

0.43

87499.53

1187348.322

2087500

12/23/1975 13:10

630

380

0.83

10760.46

281846.6628

2087500

2/3/1976 14:05

630

4010

0.38

113551.17

1361692.913

2087500

3/15/1976 9:00

630

710

0.76

20105.07

482195.4953

2087500

4/5/1976 14:20

630

609

0.65

17245.053

353738.115

2087500

5/17/1976 9:45

630

1020

0.63

28883.34

574238.0009

2087500

7/20/1976 12:45

630

123

3.5

3482.991

384701.9287

2087500

7/29/1976 9:10

630

95

4.5

2690.115

382021.0791

2087500

11/30/1976 14:45

630

644

0.71

18236.148

408597.2355

2087500

2/10/1977 9:15

630

383

0.96

10845.411

328564.9365

2087500

3/14/1977 11:00

630

6860

0.34

194254.62

2084271.263

2087500

5/25/1977 7:00

630

992

0.5

28090.464

443233.8134

2087500

7/8/1977 10:30

630

85

1.5

2406.945

113936.1113

2087500

7/28/1977 16:30

630

65

2.1

1840.605

121978.6603

2087500

12/21/1977 14:00

630

2640

0.67

74756.88

1580628.97

2087500

1/9/1978 16:00

630

4620

0.48

130824.54

1981684.082

2087500

1/10/1978 13:30

630

4460

0.42

126293.82

1673922.539

2087500

1/11/1978 14:30

630

4710

0.43

133373.07

1809841.617

2087500

1/14/1978 13:15

630

5860

0.5

165937.62

2618296.518

2087500

1/16/1978 11:30

630

5140

0.48

145549.38

2204730.775

2087500

1/20/1978 14:45

630

8300

0.46

235031.1

3411828.023

2087500

1/30/1978 15:31

630

6860

0.45

194254.62

2758594.318

2087500

4/4/1978 12:00

630

639

0.64

18094.563

365453.4281

2087500

4/26/1978 10:30

630

5350

0.43

151495.95

2055764.894

2087500

4/29/1978 11:15

630

12700

0.3

359625.9

3404679.091

2087500

7/6/1978 10:00

630

169

0.88

4785.573

132898.6547

2087500

10/12/1982 11:15

630

192

3.9

5436.864

669140.0795

2087500

12/9/1982 10:00

630

657

1.5

18604.269

880659.1191

2087500

12/13/1982 14:00

630

3860

0.6

109303.62

2069615.951

2087500

12/16/1982 8:30

630

2910

0.6

82402.47

1560254.512

2087500

1/19/1983 9:10

630

352

1.6

9967.584

503284.8461

2087500

2/16/1983 9:30

630

4770

0.4

135072.09

1705020.395

2087500

3/23/1983 11:40

630

4790

0.2

135638.43

856084.6637

2087500

5/12/1983 12:45

630

401

0.8

11355.117

286672.1922

2087500

9/7/1983 13:00

630

128

2.6

3624.576

297395.5909

2087500

9/29/1983 15:00

630

146

2.5

4134.282

326170.0441

2087500

10/26/1983 12:45

630

193

1.1

5465.181

189714.7955

2087500

11/28/1983 14:55

630

269

2.1

7617.273

504803.9943

2087500

12/7/1983 15:30

630

1720

0.6

48705.24

922212.2891

2087500

1/31/1984 8:30

630

1340

0.6

37944.78

718467.7136

2087500

2/29/1984 11:00

630

2440

0.6

69093.48

1308254.643

2087500

3/9/1984 14:40

630

4730

0.4

133939.41

1690722.53

2087500

3/27/1984 11:15

630

2140

0.4

60598.38

764935.7747

2087500

4/11/1984 12:00

630

4340

0.3

122895.78

1163488.76

2087500

4/25/1984 12:15

630

1800

0.4

50970.6

643403.9226

2087500

5/16/1984 10:00

630

567

1

16055.739

506680.5891

2087500

5/31/1984 13:45

630

2360

0.3

66828.12

632680.5239

2087500

6/19/1984 12:15

630

443

1.3

12544.431

514633.7764

2087500

6/27/1984 9:00

630

398

1.3

11270.166

462357.2077

2087500

7/11/1984 15:15

630

361

1.5

10222.437

483893.3668

2087500

7/26/1984 16:00

630

2480

0.1

70226.16

221616.9067

2087500

8/20/1984 10:30

630

642

0.9

18179.514

516331.6479

2087500

9/13/1996 12:50

630

6240

0.14

176698.08

780663.4261

 

 

 

 

 

 

 

 

There is a tendency for higher N concentrations to occur at lower flows.

 

And if you don’t force the intercept to go to 0, you get  an equation

Load = 236.27Q + 425664  kg/yr, which, with Q = 1116 cfs mean flow for the period, gives a load of 689,341 kg/year, which is probably more reasonable.