Skip Links

USGS Home Contact USGS Search USGS

Science for Watershed Decisions on Abandoned Mine Lands: Review of Preliminary Results, Denver, Colorado, February 4-5, 1998

Recurrence Intervals, Probability, and Annual Duration of Dissolved-Zinc Concentrations Using Flood Analysis Techniques in the Upper Animas River Watershed, Colorado

By Kenneth J. Leib,¹ M. Alisa Mast,² and Winfield G. Wright³

The recurrence interval (in days), probability (in percent), and duration (in days) of dissolved-zinc concentrations at the gage on the Animas River below Silverton, Colorado, were estimated by using flood-analysis techniques. Recurrence, probability, and duration are needed to describe exposure of fish to different zinc-concentration ranges and flow regimes in this mining-impacted river. These parameters also are useful for monitoring the progress of remediation in the basin because pre-, during-, and post-remediation models can be applied to this flood-analysis approach. First, using samples collected during 1992-93, a relation was developed between dissolved-zinc concentrations and streamflow using a hyperbolic regression model. A hyperbolic regression model is a flow-modification technique used to improve the relation between the dependent (dissolved-zinc concentration) and independent variables (mean-daily discharge). This data set was chosen to reflect pre-remediation conditions in the basin. Dissolved-zinc concentrations then were generated for the entire period of mean-daily-discharge record (water years 1992-96) using the pre-remediation model. No major flow alterations were known to have occurred in the basin upstream from the sample site; therefore, as long as the pre-remediation model was used, all mean-daily-discharge record at the gage could be used to predict pre-remediation dissolved-zinc concentrations. Next, a log-Pearson type III flood-analysis technique was applied to the computed dissolved-zinc concentrations to estimate recurrence intervals and exceedance probabilities for concentrations that range from 200 to 550 micrograms per liter (µg/L). The log-Pearson type III distribution also was used to normalize the predicted dissolved-zinc data set. Normalizing the data set reduced the tendency of the predicted dissolved-zinc values to bias extreme events.

Durations for dissolved-zinc concentrations ranging from 200 to 550 µg/L were estimated by using another flood-analysis tool known as the streamflow-duration curve (duration curve). The duration curve was constructed using the mean-daily discharge record mentioned previously. The pre-remediation model was then used in conjunction with the duration curve to determine dissolved-zinc durations for given flow regimes. A duration curve is used because each mean-daily-discharge value is used in the construction of the curve, as opposed to information obtained from a log-Pearson type III distribution, which uses the average of mean-daily discharges for each year. The duration curve and corresponding dissolved-zinc duration values represent the period of record (1992-97). Recurrence intervals for dissolved zinc ranged from 7 days for concentrations of 550 µg/L or more to 1.25 days for concentrations of 200 µg/L or more.

Daily exceedance probabilities for dissolved zinc ranged from 90 percent for 200 µg/L and 18 percent for 550 µg/L. For dissolved-zinc durations, 200 µg/L was equaled or exceeded 328 days per year and 550 µg/L was equaled or exceeded 50 days per year.

¹U.S. Geological Survey, P.O. Box 3367, Durango, CO 81302 (kjleib@usgs.gov)

²U.S. Geological Survey, MS 415, P.O. Box 25046, Denver Federal Center, Denver, CO 80225 (mamast@usgs.gov)

³U.S. Geological Survey, P.O. Box 3367, Durango, CO 81302 (wgwright@usgs.gov)

OFR 98-297 Home | OFR 98-297 Table of Contents | Previous | Next

USGS Abandoned Mine Lands Intiative Home Page