BATHMASTER Technical Note

Contact: Tom D'Avello or Roger Windhorn

• General Article

• Design notes for battery/solar power source

Purpose:

Mapping the current depth of lakes and reservoirs, and measuring sediment to estimate original depth and sediment accumulation are common inventories supporting watershed planning activities.

History of Sediment Surveys:

In-lake sediment surveys have been conducted by NRCS for a number of years, to both monitor PL-566 projects and to contribute to the National Sedimentation Study. These traditional methods involved sampling along cables stretched across the lake, originating from permanent concrete monuments. These surveys were time consuming and required a large number of individuals to conduct properly. Several small boats were necessary, as was surveying equipment. The ropes or cables used were subject to stretching and, of course, caused logistic problems if stretched across boat-busy sections of a lake. These surveys were also limited to small and moderate sized lakes, due to the physical limitations of stretching cable or rope an extended distance. Sediment and water depth sampling at designated points along the cable was dependent on the size of the boat used and the skill of the boat operator in keeping the boat from excessive movement. This methods also was limited as to the number of points that could be sampled as the cost per sample was high.

New needs:

In Illinois, there is now an increased demand for sediment information on all sizes of lakes. With the increased boat traffic on many lakes, too, a method was necessary to measure sediment that was easier, quicker, safer, less time consuming, less personnel demanding, and less costly than traditional methods. The use of Global Positioning System (GPS) in support of sediment surveys was started in the summer of 1995. This involves capturing the coordinates of each sample point and keeping a record of water depth and sediment thickness for each point. The sample pattern and number of samples that can be obtained are restrained only by time. The point data is input into a GIS and subsurface elevations are generated. Volumes for both water and sediment are determined from the GIS.

Equipment Used:

• 486 or Pentium Laptop Computer with Serial Port PCMCIA Card
• Rockwell PLGR+ Global Positioning Receiver with serial cable
• Sonar capable of outputting NMEA 0183 ( In our case, the Lowrance 350C)
• There will probably have to be modification of cabling and port, dependent on make and model of sonar. This information can be obtained from the manufacturer.
• GeoLink XDS (XDS) software and hardware key
   

Procedure:

Pre-Field:

1. Selection of the proper base map is essential. A 7.5 minute quadrangle map supplemented with a recent aerial photo are probably the best choices. This map, in hard copy form, serves as the field guide and is essential for taking field notes. The map also serves as the base for digitizing the lake boundary in a GIS and for use as a "background" map in GeoLink XDS (XDS). It is important to use the most recent data when digitizing the lake boundary. This usually requires acquisition of current aerial photographs, as most 7.5 minute quadrangles are relatively old. Also, when the lake was originally constructed, it had an initial surface area of a certain acreage. Through time, and now, at the beginning of the sediment survey, that original surface area will be less, even though "local" sources will still quote the documents that say "their" lake has a certain surface area. It is important to know as precisely as possible what the current lake surface area is.

2. All of the equipment used in this procedure requires a 12-volt supply of power. It is recommended that all batteries are charged sufficiently before leaving for the field. Also, it is appropriate to have an additional supply of "AA" batteries along. The re-chargeable ones work fine, but generally life-span is not as long.

3. If sediment samples are to be selected for lab analysis or for general data collection, the proper sampling equipment must be selected. A standard AMC sampling tube with the proper number of extensions works well for small samples or to cross-check sediment depth. To actually obtain a sample to analyze, I use an AMC bucket sampler with a "butterfly" closure. To obtain an "undisturbed" sample for soil mechanics testing, I generally hand-push a 4-inch Shelby tube (or its "equivalent"). All this eqipment and any accompanying tools must be gathered together also.

4. This procedure measures water volume in the lake at the current time. Sediment volume is estimated when we compare current volume of water to that volume that was prsent when the lake was first constructed. To obtain the "initial volume" figure, we need to look to several reliable sources. The best source is the original volume estimated from the orginal cross-sections during construction of the lake. If not available, the national Dams Inventory will provide an estimate, and if that is not suitable, several state agencies publish estimates that they use for their work, such as evaluating fishery habitat, etc. Unfortunately, very seldom do these estimates agree!

5. It must also be determined what the current lake level elevation is. The easiest way is to take the elevation of the principal spillway from the most up-to-date quadrangle map. If the year has been particularly dry, however, this elevation may not reflect the most accurate estimate. The local water plant often monitors the lake level on a daily basis, and this would be the best place to obtain an accurate lake level.

6. As soon as a particular day has been selected to run this survey, it is a good idea to find out the availability of satellites throughout the day. To do this, run a Number Svs and PDOP to determine if there are "bad spots" in the satellite coverage for that particular lake on that particular day. If these "bad spots" extend through the time the survey was planned to be conducted, a re-adjustment in the scheduling is needed. Generally, these poor coverage times exist for a few hours at most, and can be avoided, if known about in advance.

In-Field:

1. The boat used to conduct the survey does not need to be "fancy" but must be, above all, stable in the water. A "john-boat" or pontoon-type boat seem to be the best for this type of survey. In both cases, samples of the sediment can be taken while leaning over the side of the boat without causing excessive tipping. The motor size must simply be sufficient to propel the boat at a reasonable speed of 5 to 10 miles per hour. A 15 to 25 horsepower motor is generally a good size to use. If a much larger motor is used, then it must be "throttled-down" to the extent that fouling of the plug(s) can occur.

2. Once the equipment is on board, the depth finder should be hooked up first. The transducer is a portable type, utilizing a large suction cup to hold it in place. Ideally, the transducer should be attached to the boat hull on the side, not the rear. If it is attached too close to the motor, "prop wash" will cause cavitation and the transducer will not produce accurate readings. If mounted on the side of the boat, a "deflector shield" that can be bolted on to the rail is essential. It holds the transducer in place, when the waves push against it.

3. With the depth finder in place and giving readings, check the water depth manually, with a marked pole or graduated line to determine if the readings on the depth finder depict the actual water depth. Different mounting locations on the hull can produce different depths. It is a good practice to check or "calibrate" the depth finder several times over the course of the day.

4. Turn on the GPS unit on-site so that it can acquire satellites. This could take up to 20 minutes, depending on satellite availability and where the GPS was last used. Once 4 or more satellites have been acquired, the unit can be put in Standby Mode to conserve power.

5. The instrumentation set-up for the procedure is as follows:

Sonar: Set output depth units to meters. Meters are REQUIRED by XDS.

Additional settings may be required depending on device

GPS: Continuous Mode

Turn timer OFF

Coordinate system and datum are controlled by XDS, but we usually set both

to agree with the digital "background" data

Standard serial port settings

Let PLGR acquire satellites
 

Laptop: We use a DOS boot disk with minimal settings to AUTOEXEC.BAT and CONFIG.SYS to insure no background processes interfere with GeoLink. Quickerboot-up is another advantage.

Contents of AUTOEXEC.BAT:
PATH C:\DOS;C:\geolink
 

Contents of CONFIG.SYS:

FILES=65
BUFFER=10
STACK=9,256
REM LASTDRIVE=J
DEVICE=C:\DOS\HIMEM.SYS
DEVICE=C:\DOS\EMM386.EXE NOEMS X=D000-DFFF
DOS=HIGH,UMB
SHELL=C:\DOS\COMMAND.COM /P /E:1024
REM CardSoft (TM) 3.1 PCMCIA DRIVERS
DEVICEHIGH=C:\CARDSOFT\SSVLSI.EXE
DEVICEHIGH=C:\CARDSOFT\CS.EXE
DEVICEHIGH=C:\CARDSOFT\CSALLOC.EXE
DEVICEHIGH=C:\CARDSOFT\ATADRV.EXE
DEVICEHIGH=C:\CARDSOFT\SRAMDRV.EXE
DEVICEHIGH=C:\CARDSOFT\CARDID.EXE
DEVICE=C:\EPP\EPPDRV.EXE
LASTDRIVE=M
 

Boot-up laptop with boot disk
 

XDS: Type "glxds" at prompt to start XDS program
 

Select SETUP from menu

Select Logging

select NMEA

select collection interval. The following chart is used as a reference:

Select Port for GPS with following settings

Baud rate = 9600

Parity = None

Data Bits = 8

Stop Bits = 1
 

Select Port for Sonar with following settings

Baud rate = 4800

Parity = None

Data Bits = 8

Stop Bits = 1

Select Map Ctl

Select Background Map File with corresponding format

Select Datum & Projection and select appropriate datum and

projection for real-time and background display

Select Map Width to adjust the map scale, which controls the amount

of background map displayed at one time

Select Mode

Select View a Map file and indicate corresponding background map to display

Select Log to GPS and enter file name. Do NOT touch keyboard during logging

session.
 

Hit ESC button on keyboard to end session
 

If the inventory is incomplete, a new session can be created, with a new log file, or the original file can have new values appended to it.

6. Once the instruments have been set-up and calibrated accordingly, begin the actual in-lake procedure by the principal spillway, if possible. Often times this is where the boat launching area is located. The sampling process is relatively simple. A random, zig-zagging pattern that includes all major bays and inlets is probably the best way. With the outline of the lake displayed on the monitor, it is easy to see where the boat is and what course it must take to "see" the majority of the lake bottom. Special care must be taken to sample all major sediment-producing streams and inlets so that a true representation of sediment distribution in the lake is possible.

7. In the upper reaches of the lake, where large sediment "flats" may exist, it is important to reach as far up in to these as is possible. The better the distribution and the greater the number of samples, the more accurate will be the resulting surface. These relatively shallow "flats" are often the best places to obtain sediment samples for analyses. A tube can often be pushed into these by hand here, without falling out into deep water!

8. It is recommended, but not required, that the random sampling pattern be concluded near the point of origin. Although this is not essential, it does tend to "draw" the survey together and to more easily determine if all the important areas of the lake have been covered.

9. When it has been determined that the inventory is complete, removed the transducer disk from the side of the boat first. This insures that it doesn't get damaged as the boat enters shallow water near the boat ramp.

10. All data gathered during the inventory, into the laptop, should be "saved" to the document file set up at the beginning of the survey. Make sure all equipment is then turned off and store it all back inside the "Bath-Master" box.

Post-Field:

1. Any samples that have been collected need to be wrapped and labelled completely. These sediment samples will be saturated and a "waterproof wrap" is essential. Often times, smaller samples will need to be air-dried once they are back at the office, to make for easier mailing. Any "undisturbed" samples, however, may need to be retained in their "field" condition.

2. To prepare for the "next" survey, all batteries should be re-charged or replaced as appropriate. All equipment should be cleaned and dried, and all electrical connections should be taken apart, dried, cleaned, and re-connected so that no rust or corrosion can form. It is essential that the laptop is allow to thoroughly dry out!

Post Processing

XDS creates several files with various extensions, which can be used for import into software packages for creation of surfaces. Typical GIS packages such as Arc/Info and GRASS have several methods available for interpolation of random points into surfaces, as does the software package "SURFER", which is not a GIS package, but a specialized surface generation package. The two critical files will have a .PTS and .PTX extension. If the data will be processed in a Unix environment, run the dos2unix command prior to processing. The contents of the .PTS file are: id,x,y and the contents of the .PTX file are: id,attribute.
 

For example:

"FILE.PTS"
1,423000,1700234
2,423010,1700200
3,423040,1700180
., . , .
n,x,y
 

"FILE.PTX"
1,10
2,12
3,12
.,.
n,depth
 

These files can be manipulated to accomodate the import requirements of the respective post-processing software. To insure a good intepolation, points of "zero depth" should be appended to both the .PTS and .PTX files. The "zero depth" points are assumed to be the current lake boundary and can be exported from most GIS packages using the digital lake boundary as a source. The export properties and behavior are unique to each GIS package. Be sure to offset the "ids" for these points to insure that all points are unique.

Once imported, creation of an acceptable surface follows. This is typically an iterative process using any number of methods including, but not limited to:

Inverse Distance Weighting
Spline
Kriging
Triangulated Irregular Network (TIN)
 

Review of the resulting surface by those familiar with the area is recommended.
 

Current Volume can be determined from a GIS with the following formula:

SUM( (resolution)² x (depth) x (number of pixels of given depth) ).
 

Final Calculation: The above calculations give us a Current Lake Volume, and it is assumed that an Initial Volume, before the dam was closed, was known or pre-determined (See Procedure; Pre-Field, #4). The difference between these two values should give us a good estimate of the volume of sediment now deposited in the lake.

A yearly rate of sedimentation can easily be calculated, since we know the year the dam was closed and water first started to pool. A per-acre and a per-square- mile-of-watershed rate can also be easily calculated. Another advantage of this method is that the pattern of sediment deposition in the lake can also be tracked. Sometimes this is almost more important in reservoir management than the actual rate itself.

The accuracy of this method increases when the lake is "round" or "oval" with few shallow, "fingers" or long, narrow tributaries in the upper reaches of the lake. These shallow areas trap and hold a tremendous amount of sediment, but are difficult to maneuver in a boat and therefore offer less of a chance for inventory and the software to "account" for them. Also, some of these areas only contain water during the wettest times of the year. This sometimes leads to questions as to whether or not they should even be included in the "reservoir" totals!

Finally, any sediment values obtained MUST be checked for "reasonableness"! This method is just another way to estimate a rather abstract value - it it not an infallible procedure! If the estimated rates appear to be too high for the location in the state, soil types, slopes, etc., be sure and "validate" them! One of the best ways to do this is to compare the rates of sedimentation measured here with the rates of erosion and sediment delivery in the entire watershed. Could the erosion rates predicted actually produce this volume of sediment? If not, then we must backtrack until we can discover where the discrepancy exists!