Estimating peak discharges of small rural streams in Massachusetts

Abstract

Floodflows on natural-flow streams in Massachusetts with drainage areas between 0.25 square miles and 260 square miles may be estimated from drainage area, main-channel slope, mean basin elevation, and the area of swamps, lakes, and ponds. Multiple-regression techniques were used to define the relationship between a suite of basin and climatic characteristics and flood peaks in three flood-frequency regions at a total of 95 sites. Station flood-frequency data were computed following guidelines in Bulletin I7A of the U.S. Water Resources Council. The frequency analyses are based upon weighted skew values, and adjustments for high and low outliers, and historic peak data. Regression equations for estimation of peak discharges for 0.5, 0.2, 0.1, 0.04, 0.02, and 0.01 exceedance probabilities are provided for ungaged sites. An improved sample of flood peaks and gaging stations and the definition of three flood-frequency regions reduced the standard error of estimate by about 5 percent over those for the 1977 relations. Included in this analysis were the synthetic flood-frequency data at eight sites computed using historic climatic data and 10 parameters optimized by calibration of the U.S. Geological Survey's rainfall-runoff model with storm data observed over II years. The equations are applicable to streams unaffected by regulation where the usable manmade storage is less than 4.5 million cubic feet per square mile, or by diversions or urbanization. The equations are restricted to sites where the basin indices are within a specified range outside of eastern Plymouth, Barnstable, Dukes or Nantucket Counties. In these areas, the available data do not adequately define the influence of high infiltration and storage capacities of drainage basins on floodflows.