Air is entrained in water as it is flows through the spillways of dams, which causes an increase in the concentration of total dissolved gas in the water downstream from the dams. The elevated concentrations of total dissolved gas can adversely affect fish and other freshwater aquatic life. An analysis of total-dissolved-gas and water-temperature data collected at eight monitoring stations on the lower Columbia River in Oregon and Washington in 2011 indicated the following:

• During the spill season of April–August 2011, hourly values of total dissolved gas (TDG) were larger than 115-percent saturation for the forebay (John Day navigation lock, The Dalles forebay, and Bonneville forebay) and Camas stations. Hourly values of total dissolved gas were larger than 120-percent saturation for the tailwater stations (John Day Dam tailwater, The Dalles tailwater, Cascade Island, and Warrendale).
• During parts of August and September 2011, hourly water temperatures were greater than 20°C (degrees Celsius) at the eight stations on the lower Columbia River. According to the State of Oregon water-temperature standard, the 7-day average maximum temperature of the lower Columbia River should not exceed 20°C; Washington regulations state that the 1-day maximum should not exceed 20°C as a result of human activities.
• Of the 79 laboratory TDG checks that were performed on instruments after field deployment, all were within ± 0.5-percent saturation and only 2 checks were out of calibration by more than 2 mm of Hg.
• All but 4 of the 66 field checks of TDG sensors with a secondary standard were within ± 1.0-percent saturation after 3–4 weeks of deployment in the river. All 67 of the field checks of barometric pressure were within ±1 millimeter of mercury of a primary standard, and all 66 water-temperature field checks were within ±0.2°C of a secondary standard.
• For the eight monitoring stations in water year 2011, a total of 93.5 percent of the TDG data were received in real time and were within 1-percent saturation of the expected value on the basis of calibration data, replicate quality-control measurements in the river, and comparison to ambient river conditions at adjacent sites. Data received from the Cascade Island site were only 34.9% complete because the equipment was destroyed by high water. The other stations ranged from 99.6 to 100 percent complete.