USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2008-5070

Back to Table of Contents

Summary

Water temperatures in the Roza–Prosser Reach affect aquatic habitat and the viability of native fish populations. However, water demand for increasing population and agriculture in the Yakima River basin has caused the Bureau of Reclamation to consider water-allocation management plans that could potentially alter water temperatures in the Roza–Prosser Reach. A mechanistic water-temperature model was constructed by the U.S. Geological Survey (USGS) for use by the Bureau of Reclamation to study the relative effects that various water-management plans might have on water temperature during a growing season (April 1 through October 31). The DOS-based Stream Network Temperature model (SNTEMP version 2.0) was selected for the simulation because input data for the model were readily available, and the model has been successfully used in water temperature studies since the mid-1980s. Five SNTEMP modeled reaches were linked in series to simulate the daily maximum water temperatures at locations selected by the Bureau of Reclamation for input into the Ecosystem Diagnosis and Treatment analytical model. Meteorological data for model input were collected from an AgriMet weather station (Harrah). Streamflow data collected from 15 Bureau of Reclamation Hydromet gaging stations and field measurements collected by the USGS in September 2005 were used to estimate flow in the Roza–Prosser Reach. Regression equations were used to estimate water temperature data for flow entering the Roza–Prosser Reach from the upper Yakima River upstream of the Roza Dam and the Naches River and for all inflowing laterals under current operating conditions. Shading data for model input were collected by the USGS in October 2006. The model was calibrated using a simulation period, April 1 through October 31, 2005, and tested using a simulation period, April 1 through October 31, 2006. A sensitivity analysis indicated that daily maximum water temperatures are most influenced by daily mean air temperature and solar radiation.

Graphs comparing daily simulated and measured water temperatures at calibration and testing sites show that the model simulated seasonal patterns of measured water temperature. Root mean squared error (RMSE) for the five sites used for model calibration ranged from 1.3 to 1.9 degrees Celsius (°C), and mean error ranged from –1.3 to 1.6°C. For the five sites used for testing simulation, the RMSE ranged from 1.6 to 2.2°C, and mean error ranged from 0.1 to 1.3°C. Similar RMSE and mean error for the calibration and testing simulations indicate the model simulated differing flow and climate conditions with the same level of accuracy and that the model is suitable for studying the effect of potential water management decisions on water temperature in the Roza–Prosser Reach.

Back to Table of Contents