As Federal, state, and local government budgets shrink, monies once devoted to collection and analysis of sediment data have become scarce. Consequently, many of the U.S. Geological Survey (USGS), Water Resources Division (WRD) Districts have seen a decline in their basic sediment-data collection program. This loss of record is unfortunate. Basic data are, for many geographic areas and time periods, the only data from which temporal trends in sediment discharge can be determined. At the same time, the water science community is losing valuable information with which to assess the transport of many other contaminants (such as phosphorus and trace metals) because of the close association of those contaminants with sediment. Faced with the reality of shrinking funding and the need for sediment data, the USGS and its cooperators must decide how best to spend limited resources to gain the greatest benefit.
Many WRD districts use virtually the same data collection and analysis methods today that were used 30 or more years ago. Although these methods are still scientifically sound, they are manually intensive and are consequently time consuming and expensive. Potential gains in efficiency could be realized in almost every phase of the data collection and analysis process. Efforts are currently underway within the USGS and other agencies to seek out or develop more efficient methods and technologies; examples include the use of robotics to assist in laboratory sediment concentration determinations and the use of laser and acoustic techniques to measure suspended-sediment concentration in-situ. The remainder of this abstract will deal with methods for the computation of daily suspended-sediment discharge and the potential for improvements and efficiency gains.
In the 1980's, several computer programs for calculating sediment discharge were developed independently in District offices of the USGS, WRD, to meet the needs of the respective offices. Several of these programs were combined to form a program called SEDCALC (Koltun and others, 1994), which was accepted for use by the WRD in March 1992.
The use of SEDCALC imparts several advantages over manual computation methods. With respect to efficiency, elimination of the need to perform manual subdivision for periods with changing streamflow and (or) concentration is probably the greatest advantage realized when SEDCALC is used. The process of manual subdivision is not only one of the most time consuming but also one of the most error-fraught steps in the manual sediment-record computation process. Slightly less important efficiencies gained by use of SEDCALC include the ability to (1) plot simple instantaneous transport curves, (2) export concentration data to an external statistics package so that more complex transport relations can be explored and developed, (3) estimate concentrations for periods of missing concentration record by means of bias corrected transport equations, (4) apply simple time-varying cross-section coefficients to the raw concentration data, and (5) import data from and export data to the National Water Information System's water-quality data base by use of a card-image format.
Although SEDCALC can be used to compute daily sediment discharge more efficiently than by manual methods, selected enhancements to SEDCALC could yield further gains in efficiency and utility. Currently, the hydrographer using SEDCALC must enter sediment concentration data obtained from the laboratory, plot hydrographs and raw-data sediment pollutographs, and analyze those plots along with ancillary information to estimate the times and concentrations of additional data needed to make the raw-data pollutograph approximate the shape of the true sediment pollutograph. Those estimates of sediment concentration must then be manually entered and the hydrograph and pollutograph must then be replotted and checked to ensure that the pollutograph looks reasonable and that no transcription or data entry errors were introduced. Depending on how many estimates are required to approximate the true sediment pollutograph (a function of the timing and frequency of sample collection), this phase of the record-computation process can require significant effort and time.
The WRD's Hydrologic Analysis Software Support group is overseeing development of a graphical user interface (GUI) for SEDCALC. Although details of the GUI (which will be written in the Java programming language) have not been finalized, it likely will permit the hydrographer to interactively review and edit the sediment pollutograph at the computer or terminal, thereby eliminating the need for iterative plotting and data entry. This interactive feature will not only make the daily sediment-record computation process more efficient; it also will likely result in a more accurate record because of reductions in transcription and data-entry errors. As an added benefit, an improved user interface should promote more widespread use of SEDCALC as a computational tool, thereby leading to a more uniform approach to computation of daily sediment-discharge records within the WRD.
Other enhancements are planned for SEDCALC. From a utility and efficiency standpoint, the most important of the planned enhancements is the replacement of the current program for applying time-varying cross-section coefficients with one that will allow the hydrographer to apply cross-section coefficients as a function of streamflow and (or) time. Other planned enhancements include the addition of more sample qualification codes and support for computing discharge of water-quality constituents other than sediment.
Ideally, SEDCALC or programs with equivalent or improved functionality will eventually become integrated with the WRD's National Water Information System (NWIS). When that happens, additional efficiencies will be gained because the hydrographer will be relieved of the burden of handling the many data and report files used or produced by SEDCALC. Further, an NWIS-integrated SEDCALC will be able to take advantage of the NWIS infrastructure to add important features that are presently missing such as data aging, data security, and data archiving.
AutobiographyKoltun, Greg F., U.S. Geological Survey, Columbus, Ohio: Currently serves as surface-water and sediment specialist for the USGS in Ohio and provides national technical support for the SEDCALC sediment-record computation programs. Work experience includes 2 years at the Ohio Environmental Protection Agency and 16 years with the USGS, WRD. Areas of interest and work include sediment transport, statistical hydrology, numerical modeling, and computer applications.
Contributions from Other Federal Agencies
Contribution from the USGS