Sampling the Big Rivers

Robert H. Meade, John A. Moody, and Herbert H. Stevens, Jr.

History of the Study: 1987-92

Research Vessel

(Click on image for a larger version, 66K)

Figure 16. -- A program to sample a river as large as the Mississippi (specifically, a total length of 2,800 river kilometers between the uppermost station near Minneapolis/St. Paul and the lowermost station below New Orleans) is best carried out aboard a ship that can serve simultaneously as sampling platform, laboratory, and dormitory. The ship that filled all these functions was Research Vessel (R/V) Acadiana, owned and operated by the Louisiana Universities Marine Consortium. Acadiana is 17 meters long, 5.5 meters wide, and has a shallow enough draft (1.2 meters) to operate in most parts of the Mississippi River and in many of its major tributaries. In the configuration shown here, Acadiana has two onboard laboratories and bunk space for seven scientists plus a crew of two.

All this was to change during 1990. During the second half of 1988, after the first three of our sampling cruises had been completed, the Greenpeace ship Beluga conducted a well-publicized trip along the Illinois and Mississippi Rivers and, in December 1989, Greenpeace released its report, "We All Live Downstream: The Mississippi River and the National Toxics Crisis". In January 1990, the USGS was contacted by the office of then Senator Rudy Boschwitz of Minnesota, who asked the USGS to extend its existing study into the Upper Mississippi River, especially into the reach of the river between St. Louis and The Twin Cities of Minneapolis and St. Paul, and produce a report summarizing our results. Soon thereafter, members of Congress from other States along the Mississippi-Arkansas, Illinois, Iowa, Louisiana, Missouri, Tennessee, Wisconsin-joined Senator Boschwitz in his request.

The response to this request was a change in the goals of our program from mainly research to mainly assessment, and an expansion of the scope of the program to include the much different hydrologic setting of the upper river. This expansion required an augmentation of the primary strategy for sampling the contaminants carried by the flowing waters to include a second strategy for sampling the contaminants that were stored with the sediment deposited on the bottoms of the navigation pools of the Upper Mississippi.

Sampled Sites

(Click on image for a larger version, 66K)

Figure 17. -- The flowing waters of the Mississippi River and its principal tributaries were sampled repeatedly at critical sites during our studies.

A

During 1987-90, our studies focused on the lower and middle reaches of the Mississippi River-from the vicinity of its confluences with the Illinois and Missouri Rivers to a point a few kilometers below New Orleans.

B

During 1991-92, the emphasis shifted to the Upper Mississippi River, while the sampling of the Lower Mississippi was continued at a reduced number of sites. Furthermore, time-series samples were collected at three sites on the Mississippi River and at three sites on major tributaries.

Sampling the Flowing Waters

Two principal methods were used to collect water samples for determining concentrations and loads. Both methods take into consideration that the velocity of flow and the concentrations of suspended sediment and even dissolved matter are not distributed uniformly across the widths of large rivers. Flow velocities usually are greatest at the river surface in midstream and they decrease toward the bottom and banks of the river. Sediment concentrations usually are greatest near the river bottom and smallest at the river surface. After large tributaries enter the Mississippi, their waters may not mix thoroughly for 100-200 km downstream. Were it not for these inhomogeneities, one simply could dip a bottle or bucket into the middle of the flowing river, analyze the water collected, and multiply the concentrations times the discharge to calculate the loads. Indeed, this simple method has been used around the world for many years to sample the chemical compositions of rivers large and small. But the results of this sampling approach cannot be evaluated for their accuracy because the point at the centroid of water flow and the point where the contaminant concentration is the representative average of a river cross section are seldom obvious, unless extensive sets of detailed measurements have been collected beforehand.

Our standard method of collecting representative amounts of river water for the analysis of contaminant loads samples the full depth and width of the river. By this method, a sampler is lowered to the bottom of the river and raised back to the surface, collecting water through the full depth of the river, at a number of locations spaced equidistantly across the river from bank to bank. The resulting sample is called a depth-integrated composite. The sampler is designed to admit water at the velocity at which it is flowing in the river, and the sampler travels to the bottom and back at the same vertical speed each time. It collects the most water where the river flows fastest and deepest, and the least water where the river flows slowest and shallowest; the resulting sample is thereby weighted for water discharge. This method has been the standard for many years for sampling smaller rivers. Scaling it up for larger rivers such as the Mississippi is simple in principle but has proven more complicated in actual practice. Working from a moving ship in flowing water requires microwave positioning equipment or some other means of finding and maintaining the appropriate locations for collecting samples in a cross section of the river. A specially designed hydraulic winch and a sampler consisting of an initially collapsed Teflon bag inside a plastic bottle are required to collect the fast and deep waters of the Mississippi. Deploying the navigation equipment, positioning the ship, and collecting a representative sample in sufficient volume for all the required analyses usually kept the ship's crew and half a dozen scientists busy for the better part of a day at most of our sampling cross sections.

Sampling Flowing Waters

(Click on image for a larger version, 99K)

Figure 18. -- Obtaining representative samples of the flowing waters of the Mississippi River and its larger tributaries required special procedures and equipment, some of which are illustrated here.

A

In a typical cross section of the Mississippi River, representative amounts of water were collected, through the full river depth from the surface to the bottom, at locations spaced at approximately equal intervals across the river. All the water collected at the locations across the river was combined into a single composite sample. At many cross sections, two separate composites (shown here as red and yellow) were collected so comparisons could be made to assess the precision of the sampling procedure.

B

Sampling from a freely moving vessel in a large river requires a precise means of determining the location and motion of the vessel. Microwave positioning equipment served that purpose. Shown in the photograph is one of several remote units that were placed at known positions on the riverbanks at or near the sampling sites. Aboard the vessel (not shown in the photograph) was a master microwave unit that measured the distances from the remote units and thereby determined the relative position of the vessel.

C

The array of equipment for measuring water velocity and for sampling the flowing waters and suspended sediment consisted of, from top to bottom,

(1) a current meter, similar in design to the familiar anemometer that is commonly used to measure windspeed,

(2) a sampling bottle, containing a collapsed Teflon bag (shown here partly filled) that expanded as it received the incoming sample, and fitted in front with an isokinetic nozzle that was designed to admit water and suspended sediment in proportion to the velocity at which they were moving in the river, and

(3) a sounding weight of either 150, 200, or 300 pounds-depending on how fast and deep were the waters being sampled.

The hydraulic winch that was used to lower and raise the sampling array is not shown in the photograph.

Sampling Navigation Pools

(Click on image for a larger version, 99K)

Figure 19. -- The contaminants adsorbed to the sediments deposited in the backwater areas of the navigation pools of the Upper Mississippi River were sampled to obtain representative average concentrations for each pool.

A

The Upper Mississippi River between Minneapolis/St. Paul and St. Louis is segmented by 26 principal navigation pools.

B

A "typical" pool (Pool 8, in this example) consists of three parts: an upper section, characterized by riverine channels and emergent flood plains; a middle section, characterized by numerous islands, parts of which are deltaic in origin but most of which are the elevated areas of the flood plains that flanked the river channels before the pool was inundated; a lower section, characterized by a shallow expanse of open water.

C

Sampling was concentrated in the lower sections of the pools (Pool 8 is again the example)-outside the main navigation channel-because this is where fine-grained sediments and their adsorbed contaminants are most likely to be stored. Samples were collected at 15-20 locations per pool (shown here as three linear rows of white dots), which was sufficient to obtain a representative average, without seeking either "hotspots" or pristine areas.

Our second principal method for collecting the flowing waters was by pumping them directly out of the river. Because the analyses for some of the adsorbed contaminants required larger quantities of suspended sediment than we were able to recover by depth-integrated sampling, we adopted the more expedient method of pumping water from the river, usually from a depth halfway between the surface and bottom. The pumping method collected 5 to 10 times the amount of water that could be collected in the same time by the depth-integrated sampling. Consequently, 5 to 10 times the amount of suspended sediment could be recovered, making it increasingly possible to measure many of the industrial and other contaminants that are difficult to detect on small quantities of sediment. The quantities of water pumped in the different parts of the channel were proportional to the water discharges there. The disadvantage of the pumping method is that it does not sample the full depth of the river. However, because depth-integrated samples were being collected at the same time at the same locations, comparative studies were made to show that, for the suspended particles with which contaminants are preferentially associated, the pumped sample was accurately representative of what was being transported in the full depth of the river. A typical pumped sample was somewhere between 500 and 1,000 liters (one-half ton to a ton, that is) of river water, which required 6 to 10 hours to collect and many more hours of shipboard processing to separate all the water from the suspended sediment.

Despite its disadvantages, a secondary method of dipping bottles into the surfaces of the rivers was used to obtain some samples. This sampling method was used most consistently in the tributaries of the Upper Mississippi River that were too shallow to be navigated by R/V Acadiana and had to be sampled from a small boat. The procedure consisted simply of dipping bottles into the tributaries near their estimated centroids of water discharge. The concentrations of contaminants in these surface-dipped samples are useful for computing approximate downriver loads, so long as the contaminant is transported in solution and the river waters being sampled are well mixed from top to bottom and from bank to bank. In other circumstances, the concentrations of surface-dipped samples may be more indicative than definitive.

Sampling Bed Sediments

(Click on image for a larger version, 99K)

Figure 20. -- The sediments and their adsorbed contaminants on the bottoms of the navigational pools of the Upper Mississippi River were sampled by fairly traditional means.

A and B

The shallow backwater areas of most of the navigation pools were sampled from a small boat. Sampling equipment consisted of a clamshell grab (midboat) and gravity corer (forward). Also measured were current velocity (aft) and the position of the boat (microwave unit atop the pole).

C

In the deeper pools such as Pool 19 (shown here) or Lake Pepin, bottom-sediment samples were collected directly from Research Vessel Acadiana.

D

From the sediment recovered in the clamshell grab, separate subsamples were taken for the analysis of organic constituents such as PCBs and coprostanol. Samples for heavy-metal analysis were taken separately with a plastic gravity corer so as to avoid any error caused by the introduction of extraneous metal from the sampling equipment itself.

Sampling the Stored Sediments in the Pools of the Upper River

The contaminants adsorbed onto the sediments stored in the pools of the Upper Mississippi River were sampled by conventional means. A clamshell grab and a small plastic corer were used to collect samples of the uppermost 10 cm of the silt or mud that lay on the bottom at 15 to 20 places in the shallow or backwater areas of the navigation pools. In only the two largest pools, Lake Pepin and Pool 19, were the waters sufficiently deep to allow the bed samples to be collected directly from Acadiana. In the remainder of the pools, samples were collected over the side of a 14-foot-long boat. The locations in the pools where the samples were collected were fixed by their distances (determined by microwave-distancing equipment) and bearings from objects or landmarks whose positions were known. Sampling the stored bed sediments of a pool took four people most of a day. Equal quantities of each of the 15 to 20 grab samples collected in each pool were combined into a single batch sample for analysis of organic compounds such as PCBs and sewage contaminants such as coprostanol. Core samples were kept separate for individual analyses for heavy metals.

Sampling the Length of the River

Time-Series Sampling at Fixed Locations

Further Information

More detailed accounts of the procedures used to sample the Mississippi River are given in the following reports.

Leenheer, J.A., Meade, R.H., Taylor, H.E., and Pereira, W.E., 1989,

Sampling, fractionation, and dewatering of suspended sediment from the Mississippi River for geochemical and trace-contaminant analysis, in Mallard, G.E., and Ragone, S.E., eds., U.S. Geological Survey Toxic Substances Technical Meeting, Phoenix, Arizona: U.S. Geological Survey Water-Resources Investigations Report 88-4220, p. 501-511.

Meade, R.H., and Stevens, H.H., Jr., 1990,

Strategies and equipment for sampling suspended sediment and associated toxic chemicals in large rivers-with emphasis on the Mississippi River: The Science of the Total Environment, v. 97/98, p. 125-135.

Moody, J.A., 1993,

Evaluation of the Lagrangian scheme for sampling the Mississippi River during 1987-90: U.S. Geological Survey Water-Resources Investigations Report 93-4042, 31 p.

___ ed., 1995,

Chemical data for water samples collected during four upriver cruises on the Mississippi River between New Orleans, Louisiana, and Minneapolis, Minnesota, May 1990-April 1992: U.S. Geological Survey Open-File Report 94-523, 297 p.

___ ed., 1995,

Hydrologic, sedimentologic, and chemical data describing surficial bed sediments in the navigation pools of the Upper Mississippi River, July 1991-April 1992: U.S. Geological Survey Open-File Report 95-708.

Moody, J.A., and Meade, R.H., 1992,

Hydrologic and sedimentologic data collected during three cruises at low water on the Mississippi River and some of its tributaries, July 1987 through June 1988: U.S. Geological Survey Open-File Report 91-485, 143 p.

___ 1993,

Hydrologic and sedimentologic data collected during four cruises at high water on the Mississippi River and some of its tributaries, March 1989 through June 1990: U.S. Geological Survey Open-File Report 92-651, 227 p.

___ 1994,

Evaluation of the method of collecting suspended sediment from large rivers by discharge-weighted pumping and separating it by continuous-flow centrifugation: Hydrological Processes, v. 8, p. 513-530.

___ 1995,

Hydrologic and sedimentologic data collected during three cruises on the Mississippi River and some of its tributaries from Minneapolis, Minnesota, to New Orleans, Louisiana, July 1991-May 1992: U.S. Geological Survey Open-File Report 94-474, 159 p.

Moody, J.A., and Troutman, B.M., 1992,

Evaluation of the depth-integration method of measuring water discharge in large rivers: Journal of Hydrology (Amsterdam), v. 135, p. 201-236.

Rees, T.F., Leenheer, J.A., and Ranville, J.F., 1991,

Use of a single-bowl continuous-flow centrifuge for dewatering suspended sediments-Effect on sediment physical and chemical characteristics: Hydrological Processes, v. 5, p. 201-214.