Computer Program for Smoothing and Correcting Hourly Time Series Data

by Samuel G. Bryson, Hydrologist; Resources Consultants, Inc, United States,
Robert L. Barkau, Hydrologist; Resources Consultants, Inc, United States,

Document Type: Proceeding Paper

Part of: Computerized Decision Support Systems for Water Managers


A FORTRAN computer program has been developed for the U.S. Army Corps of Engineers St. Louis District that automatically screens and corrects hourly time series data for gross errors and inconsistencies. The program is referred to as AUTOSCRN, for Automatic Screening. The type of data errors considered are 1) missing or struck gage data, 2) abrupt stage shifts, 3) outliers not detected in range checks, and 4) excessive oscillations. The program both reads from and writes to files using the Corps' -Hydrologic Engineering Center Data Storage System (HECDSS). Use is made of a previously available smoothing algorithm developed by AT&T Bell Labs, known as LOWESS, for Locally Weighted Scatterplot Smoother. This algorithm fits a least squares regression to locally weighted data points to determine the best estimate for any given data point in the time series.

Subject Headings: Computer software | Time series analysis | Hydrologic data | Information management | Automation | Errors (statistics) | Computer analysis | Federal government

Services: Buy this book/Buy this article


Return to search