ANSWER TO U.S. CORPS OF ENGINEERS
CONCERN NUMBER 5

The extensive channel modifications could result in channel instability which could also extend up tributaries of the Pearl River.

 

    The lake level is supposed to provide a much slower flow rate during all of the time that the river is not on flood. It will prevent the constant movement and redistribution of sand that occurs continuously in a steeply flowing river channel. The main flow coming into the lake will be relatively clean water from the reservoir. Silting will occur in the tributary estuaries in the same manner that silting is occurring now in the tributaries along the lower reservoir. This silt can be recognized from original lake bottom. A measurement of the amount of silting coming from a given watershed with heavy residential development can be determined in the reservoir. From the 30 year history of silting on reservoir tributaries it can be determined how much extra dredging will be needed in the estuaries to accommodate this silt for a long time. An amount could be removed that would be calculated to provide a 100 year maintenance free flow channel in the flood control lakes. Most of the sand that exists between the levees and bridges south of downtown Jackson would be removed leaving no sand to be redistributed in the lowest most lake. Preliminary estimates are that about four million yards of additional dredgings would be required.

SEDIMENT ENTERING THE LAKE

     Creeks below the Barnett Reservoir dam will carry sediment. The flood control lake as a 140 square mile watershed form these sources. The 1998 USGS quadrangles show approximately 70% of the area developed into mature residential subdivision and business areas. It is estimated that 30% of the watershed remains subject to development. Very little farming occurs in this area. Observation of creeks during flash floods reveals that the metropolitan creeks do not carry high concentrationa of sediment.

     One of the most extensive areas for recent development is in the area along Mill Creek which flows into the reservoir. Subdivisions extend all the way to its headwaters near Brandon. Over 20% of this upland area not in subdivision has, until recently, been formed in wheat and soybeans. Much of this farming has recently been discontinued. Erosion from this area during the past 15 years has deposited approximately 100,000 years of sediment in the reservoir. This amount of sediment has been dredged by the reservoir wngineers to keep boat docks open in the creek estuary. Complete development of this watershed could result in as much as 200,000 additional yards of sediment settling in the reservoir over the next 30 to 50 years. Other studies conducted on residential lakes reveal that similar amounts of sediments collect in them during development and that very lettle sedimentation accurs after the development has ended.

     The Mill Creek watershed is appriximately one-tenth the cumulative sediment bearing watershed of the flood control lake. If all of the lakes' watershed were subject to development and farming, such as the Mill Creek watershed, approximately three to four million years of sediment could be estimated to enter the lake over the next 100 years. However, since only 30 percent of the area remains to be developed and little re-development is expected in the mature areas, as little as one million yards could enter the lake in 100 years. Additionally, Department of Environmental Control regulations now reqire that sediment be contained on construction sights. This should reduce the amount of sediment that future construction will contribute to the flood control lake.

     This plan calls for overdredging the creek estuaries to accommodate heavy sediment from the tributary watershed such that it will not enter the flood control lake. Two to four million yards should accomplish this with a two to four safety factor for 100 years. Very fine sand and colloidal material will enter the lake.

SEDIMENT TRANSPORTATION IN THE LAKE

The flood control lake will occupy the lower meander belt (riverbottom land) of the two mile wide Pearl River flood plain. This belt of land averages 3000' wide and is at a level approximately nine feet lower than the upper level terraces (flatwoods) of the flood plain. Material in the lower level meander belt consists of sandbars (fine to coarse grain clean to silty sand) that rests upon a hard blue clay (see cross-section below). Sand thickness ranges from 25 fet at the crest of sandbars to five feet in the lowest troughs. This sand averages 15 feet thick.

     Dredging will remove approximately 40 jpercent of this sand. The rest will be planed off and remain on the bottom of the lake. The sandbar material dredged from the lake will be fine to coarse grain quartz sand. This material will remain also on the lake bottom. We have made observations of sand movement in high enerby flow over sandbars in a effort ot develop judgement as to how this sand will act during times when major floods pass though the lake. See map of the Pearl River channel below.

POSITION 1

     Stream flow 2-1/2 inches deep over medium grain (.25 to .50 mm) sand was measured at a velocity of 0.80 ft/sec at the surface. No movement of sand was occuring. The surface of the sand was smooth and had an orange color. Some areas had a greenish, pulp like coating approximately 1/16 inch thick. Breaking the ripples revealed that the sand was gray in color and that the orange coating was on the surface only. The underlying sand would not move when the surface is broken. A 4 oz cup of clean gray sand was deposited on the sand bottom surface and observed. Stream flow flattened the pile of sand out over an area about 6 inches wide and 10 inches long in the downstream direction. No further movement occurred. Observed for 2 hours. Sand grains placed on a metal measuring tape 1/4 inch above bottom were observed. They would not roll on the tape, indicating that velocity in the bottom 1/4 inch was not sufficient to move the sand grains.

POSITION 2

     Stream flow 2-1/2 inches deep was measured at a velocity of 1.25 ft/sec at the surface. No movement of sand was occuring.  Sand size was measured to be .25 to .50 mm. The surface of the sand bottom had "flagstone" ripples and an orange stain. Sand movement is by sporatic movement f individual grains. Two 4 oz cups of clean gray sand were deposited on the sand bottom surface. These piles were flattened by the stream flow into dune shaped flagstone ripples covering an area approximately 6 inches wide and 10 inches long. Sand grains continued to move by sporatic dislodging and lodging of individual sand grains.

     One sand dune was scattered to break up its shape. In its flattened condition, sand movement occurred at a more rapid pace (still individual lodging and dislodging of individual sand grains). After 2 hours, sand had trailed our in a downstream direction 3 feet, moving over the orange stained encrusted sand surface in a laminate 1/16' thick or less.

     The sand dune that had not been flattened retained its shape and moved approximately 4 inches downstream in 30 minutes. It then began to become incrusted with a greenish orange material and after 1 hour stopped moving.

     Sand was scooped up along with a layer of this greenish pulp coating and deposited in a 4 oz. pile. This pile flattened and retained a shallow mound shape. It developed a partial coating of the greenish mineral matter and stopped moving after 5 minutes. After 1 hour this mound developed a slight ripple mound shape.

     This same area was observed 24 hours later. The depth had decreased to 1-1/4 inches. Velocity of flow had only slightly decreased to 1.10 ft/sec at the surface. Individual sand grains were moving on the sand bottom surface over the general area in trace amounts. The greenish pulp surface was observed to be eroding in small areas with increased sand movement in the eroded areas. Clean sand in 4 oz. piled would flatten and move by lodging  and dislodging of sand grains. These piles string out in the downstream direction 3 feet in 30 minutes. Flow appeared to be more erosive than the previous day with the only known change being that the depth was reduced form 2-1/2 inches to 1-1/4 inches . After observing the shallow bar where velocities remained close to 1 foot per second for 7 days, the bar did not change shape and moved downstream about 1 inch.

POSITION 3

     This position on the outside of the channel bend was 32 inches deep. Stream flow velocity was measured at 1.90 ft/sec. Bottom could not be seen at this depth. Material on the bottom was measured to ve medium (.2 to .6 mm) grain quartz sand. A 4 oz. cup was buried flush with the bottom and observed for full. After 1 hour no sand settled out in the cup. Flow around the leading edge of the cup and sides, dished out sand to a depth of 1/4 inch. It was determined that at this depth and velocity, stream flow could not suspend .2 to .6 mm sand grains. By holding a 5 foot cane rod in one hand perpendicular to the stream flow and gauging the pressure created by the flow it could be determined that much less occured when the cane rod was held on the bottom surface, and pressure increased dramatically as the rod was lifted to a level 10 inches above the bottom surface. By suspending a 12 oz. plastic bottle filled with water on a fishing cork, flow velocity was gauged a various depths. It could be seen that flow velocity decreased to 1.5 feet per second in the bottom foot of depth. Repeats of this test demonstrated a wide range of results and the test was not considered to be very accurate. The results, however, did show consistently lower velocities near bottom. On the 7th day of testing, a 3/4 inch PVC tube was used to gauge stream velocities at various depth by gauging flow pressure in a clear static tube. Depth had increased to to 38 inches. Velocity at the surface was 2.1 ft/sec. Velocities decreased slightly with depth and dramatically in the bottom 6 inches. Velocity in the bottom 1 inch was approximately 1 ft/sec. The sand bottom could not be seen but felt slightly firm on top and it was determined that sand grains were not suspended in the flow.

     A history of reservoir discharge was compiled to determine what flow rates could be expected in the lake over the next 100 years (see tables and graphs attached). A 200 year event was included in this 34 year record and its effect was tripled in the 100 year estimation. This tended to distort the graph at higher magnitude floods. From the resulting table, it is estimated that flow rates in the lake will exceed 1.25 ft/sec 47 days during the next 100 years. At this rate, direct observation shows that the medium grain sand base in the lake will not, as a practical matter, appreciable move. If the dand could be kept perfecly clean it would develop a resistance shape (flagstone ripples) and move as a thin laminate approximately up to 75 feet per day, 3500 feet in 100 years.

DEPOSITION OF COLLOIDAL MUD

     Colloidal material will enter the lake from the strem estuaries and dehydrate into the medium grain sand bottom of the lake. By observing the build-up of sediment in residential lakes, beaver ponds and the reservoir it can be estimated that a thick muddy bottom will develop in the lake over the next 100 years. Colloidal mud with predominant particle size of .005 to .015 mm settles out in beaver ponds which, when constructed across creeks, experience frequent high velocity flow. Between periods of high velocity flow, thse ponds do not dry up, slight flow continues and colloidal particles are, nevertheless, able to combine, attract and attach themselves to the bottom. It is expected the this will also happen in the flood control lake during periods when very little movement occurs. (See table showing mean lake velocities.)

EFFECT OF MAJOR FLOODS ON A MUD BOTTOM

     In an effort to develop judgement about how this mud bottom will react to flow during major floods in the lake, we returned to the previous test site. a one quarter inch thick layer of this material was placed on spread metal sheets and placed in the Pearl River stream flow at positions 2 and 3 on the sandbar to observe how it holds up and resists erosion. Mud, about 1/4 inch thick, was laso spread upon the sand bottom surface.

POSITION 2

     After 4 days no change was in the condition of the mud on the sand surface or on the spread metal grid. At this time, the depth had increased from 2-1/2 inches and a velocity of 1.25 ft/sec and the velocity to 1.67 ft/sec.

POSITION 3

     This area, 32 inches deep, was flowing at a velocity of 1.90 ft/sec at 1 foot of depth. A spread metal grid with 0.005 to 0.015 mm mud spread upon it was placed at the bottom surface. The mud was oserved for 4 days and remained intact. At the end of the test, depth had increased to 38 inches and velocity to 2.1 ft/sec. This is approximately the same velocity as will occure in the lake during a 100 year flood. The time period (4 days) is also the same as the period that a 100 year flood can be expected to flow at these rates. We used 3/4" PVC tubing to gauge stream pressure at varying depth and estimated flow velocity in the inch next to bottom to be approximately 1 ft/sec.

TRANSPORTATION OF SEDIMENT THROUGH BRIDGES

     At the present time, major floods make effective use of the channel span of the bridges only. These spans average approximately 350 feet width. The other 2/3rds of the bridge opening is blocked by sandbars and willow growth. This subjects the channel span to extreme flow which erodes down to and into the blue clay. By opening up and deepening the entire bridge section, average flow rates through the bridge will be greatly reduced. In many cases, the removal of these sandbars would free little more piling than exists at present on the most exposed piling in the low elevation back swamp areas of the bridge sections. Improved flow through the intire bridge section must be studied by someone with experience in hydraulic analysis of bridges to determine what, if any, provisions need to be made to adapt the bridges to these new conditions.

CONCLUSION

     Direct observation of high energy flow over flat sandbars in the river indicates that the clean medium grain sand bottom forms a resistance shape (flagstone ripples) and a packed surface which resists erosion. Erosion begins to occure when flow velocity approaches 1 ft/sec in the depth interval that is 1 inch above the sand bottom. One ft/sec flow was observed in the 1 inch interval above the sand bottom when 3 feet of water was flowing at 2.1 ft/sec at the surface. (The mean velocity was calculated to be 1.8 ft/sec). Assuming that similar velocities will occure in the lake (15 feet deep) these rates may be expected to occur 26 days during the next 100 years. Nevertheless, it appears that the lake will experience deposition of colloidal mud during close to 20,000 days over the next 100 years. Since the lake will develop a mud bottom and, since no flow is projected to occur in 100 years that will erode this type of bottom, the lake should be maintenance free for 100 years,

The Diagram below shows threshold velocities of sediment in stream flow. The upper curve shows the velocity necessary for a stream to pick up and move a particle of a given size. This is a zone on the graph, not a line because of variations resulting from stream depth, etc. The lower curve indicated the velocity at which a particle of a given size will settle out and be deposited. Note that fine particles will stay in suspension at velocities much lower than those required to lift them from the stream bed surface.