Making Sense of Sensors: Stream Carbon Flux Determination at the Five USGS WEBB Watersheds

Thursday, 18 December 2014: 5:45 PM
James B Shanley1, JohnFranco Saraceno2, Mark Dornblaser3, Brent T Aulenbach4, Alisa Mast5, David W Clow5, John F Walker6, Krista Hood6, Kimberly Wickland7, Brian A Pellerin8, George Aiken9, John T Crawford10 and Robert G Striegl11, (1)USGS, Montpelier, VT, United States, (2)USGS CAWSC, Sacramento, CA, United States, (3)US Geological Survey, Boulder, CO, United States, (4)USGS, Georgia Water Science Center, Norcross, GA, United States, (5)USGS Colorado Water Science Center Denver, Denver, CO, United States, (6)USGS, Middleton, WI, United States, (7)USGS, Baltimore, MD, United States, (8)USGS California Water Science Center Sacramento, Sacramento, CA, United States, (9)USGS Colorado Water Science Center Boulder, Boulder, CO, United States, (10)University of Wisconsin Madison, Center for Limnology, Madison, WI, United States, (11)USGS WRD, Boulder, CO, United States
At the five diverse forested and/or alpine headwater sites of the U.S. Geological Survey (USGS) Water, Energy, and Biogeochemical Budgets (WEBB) program, we measure stream concentrations of all forms of carbon – DOC, PC (assumed to be POC), CO2, CH4, and bicarbonate alkalinity. One goal of this work is to estimate DIC, DOC, and POC fluxes from headwater landscapes. For DOC, POC, and CO2, we utilize high-frequency measurements with in situ sensors. We also take discrete samples over a range of hydrologic conditions for independent flux calculations, and to verify the sensor measurements and test proxy relations. In situ fluorescing dissolved organic matter (FDOM) serves as a proxy for DOC. Turbidity is needed to adjust FDOM, and also serves as a proxy for POC. DIC flux is computed from the discrete alkalinity analyses.

Sensor measurements at the five sites – in Colorado, Wisconsin, Vermont, Georgia, and Puerto Rico -- began either in 2011 or 2012. The sensors generally functioned well, but occasional down time required gap filling, typically using strong relations with stream discharge. We computed at least one full year of carbon flux at each site. In this paper we compare sensor-based and sampling-based carbon flux estimates. We interpret the differences among sites in total carbon flux and its relative apportionment among DIC, DOC, and POC, with respect to climate and ecosystem type. Finally, we evaluate the significance of stream carbon flux as a component of the ecosystem carbon balance.