Charles M. Brendecke

Comparison of two daily streamflow simulation models of an alpine watershed

Abstract A comparison of daily streamflow simulation by two different computer models has been completed for an alpine basin in central Colorado, U.S.A. The catchment is part of a 33-km2 area providing most of the City of Boulders raw water supply. It is located between 2964 and 4103 m a.s.l. in elevation along the Continental Divide 24 km west of Boulder, Colorado. The SSARR model simulation, which makes generous use of lumpedtype parameters, was compared with the PRMS model simulation, which makes greater use of individually defined parameters, by examining their respective ability to fit observed hydrograph volumes and shape. Both computer models were found to simulate streamflow equally well. The PRMS model was found to be more suitable overall because of its ability to handle small catchments and the generous use of individual physically-based parameters.

Nutrient Enrichment Studies

The purpose of a nutrient enrichment study is to determine the actual or incipient limitation of phytoplankton by specific nutrients. The literature on nutrient enrichment contains a wide variety of experimental designs and methods for such studies. This variation in methodology is in part explained by differences in the exact purpose of the studies. In general, nutrient enrichment studies can involve enrichment of standard algal cultures, or enrichment of natural assemblages over a few hours, a few days, or many days.

Design of the Study

Variation in Lake Dillon and its watershed was quantified by a dual sampling program, one part of which was designed primarily to provide detailed information for a few sites on many dates and the second to provide a detailed picture of spatial variation on a smaller number of dates. These two approaches were supplemented by special data collection programs of more limited scope whose purpose was to provide information that would not necessarily be forthcoming from the routine studies. The Lake Dillon Study is thus supported by three data sets: (1) time series, (2) spatial survey, and (3) special studies.

Horizontal Spatial Variation in the Lake

Up to this point most of the analysis has dealt with samples taken from the index station in the middle of the lake. The question arises how typical the index station is of the lake as a whole. This question is answered here in two stages. First, an analysis is made of the five-station heterogeneity series. As described in Chapters 2 and 3, this series consisted of analyses made on a set of samples taken on 32 different dates at the four main stations and the index station. The top (0–5 m) and bottom of the water column were sampled at each station. Since each one of the four main stations was located at the mouth of one of the four main arms of the lake, this sample series gives information about the degree of variation to be expected over the deep-water section of the lake.

Particulates and Phytoplankton Biomass

The total particulate load of a water body can be divided into living and nonliving fractions. The living fraction is influenced by nutrients. The nonliving fraction, however, is jointly determined by biological factors such as the rate of production of biological detritus and by abiotic factors such as silt content of the lake’s water supply. It is essential that the particulate functions be disentangled as well as possible before any causal analysis is attempted. Here we show the patterns of particulate concentrations in Lake Dillon and the contributions of identifiable fractions to these patterns.

Total Nutrient Loading of the Lake

The nutrient chemistry data for rivers and effluents must be expressed in terms of transport or loading rates to support an analysis of lake-watershed coupling. Total transport is dealt with in this chapter, which is followed by a dissection of sources and mechanisms of transport in subsequent chapters.

Physical Variables and Major Ion Chemistry of the Lake

The geology of the Lake Dillon region has been well documented in connection with the construction of the Roberts Tunnel and the exploitation of mineral resources in the watershed. The publication of Wahlstrom and Hornback (1962) gives details and contains references to earlier studies.

Using the Model for Prediction

The Lake Dillon Model was applied to eight different scenarios considered to encompass the range of possibilities that might be realized over the next 20–25 years. The scenarios comprise five sets of assumptions for development and land use, each of which was applied to hydrologic assumptions for a dry year and for a wet year. Predictions in each case include the total loading of the lake, the source distribution of the loading, and the response of the lake to the predicted amount of loading. The dry-year hydrologic conditions were always set identical to 1981, and the wet-year conditions were set identical to 1982. Use of the hydrologic conditions of 1981 and 1982 is advantageous since the behavior of the lake in these 2 years under present land use is well documented and can thus be compared easily to future years of similar hydrologic conditions but very different land uses.

Phosphorus and Nitrogen in Lake Water and Sediments

Phosphorus and nitrogen are the two elements most likely to limit phytoplankton growth. Detailed information on the concentrations, vertical distributions, and chemical fractions of these elements is therefore useful in interpreting average phytoplankton abundances and seasonal changes in abundance. We present here an overview of the phosphorus and nitrogen chemistry that can be referenced in later chapters dealing with phytoplankton.

Chemistry of Nutrient Sources as They Enter the Lake

Pathways followed by water and nutrients entering Lake Dillon can be divided for present purposes into six categories: (1) major rivers, (2) streams not joining a major river prior to reaching the lake, (3) sewage effluents, (4) miscellaneous surface drainage not accounted for in other categories, (5) precipitation, and (6) groundwater. All of these categories except miscellaneous surface drainage and groundwater, which made minor contributions, were studied chemically at or very near their point of entry to the lake. In the first category are the Snake River, the Blue River, and Tenmile Creek. These will take up most of the attention of this chapter, since they account for the bulk of surface transport to the lake. In the second category are Soda Creek and Miner’s Creek, both of which were sampled above the point sources that enter them near their mouths. In the third category are the effluents of the Frisco WWTP (which enters Miner’s Creek near the lake) and Snake River WWTP (which enters Soda Creek near the lake). Both of these effluents were sampled prior to entering streams. Breckenridge effluent was part of the Blue River sample, since the effluent enters above the river mouth, and Copper Mountain effluent was part of the Tenmile Creek sample for the same reason. Discharge data for rivers were obtained from USGS data records. For the effluents, discharge data were obtained from plant operators, and for Miner’s Creek and Soda Creek the data were based on our own measurements. All chemistry data were from our own analyses.