A tower located at the base of the watershed, was extended from28 to 37 min the Fall of 2004 It was instrumented in May 2005 with ten shielded thermistors (Model 107 temperature probe, Campbell-Scientific Inc., Logan, UT), a net radiometer (Q7 REBS, Campbell-Scientific Inc.), a temperature/RH probe (HMP45c, Campbell-Scientific Inc.), eight 2-D sonic anemometers (WS425, Vaisala, Helsinki, Finland) and two 3-D sonic anemometers (81,000, RM Young, Traverse City, MI) (Table 1). For most periods the 2- and 3-D sonic anemometers were recorded at 0.1 and 1 Hz, respectively, and values were averaged over 1 min intervals. During select periods of CO2 or SF6 monitoring, the 3-D sonic anemometers were recorded at 10 Hz. The remaining instruments were recorded every 1 s and averaged over 15 min intervals (using CR10X and CR23X data loggers, Campbell-Scientific Inc.).
A second tower, 19.8 m tall with a 3.7 m telescoping pole at the top, was constructed near the ridge of the watershed (Ridge tower; Fig. 1a). It was instrumented in May 2005 with a cup anemometer and wind vane at 23.5 m (05103-L RM Young Wind Monitor, Campbell Scientific Inc.), a temperature/RH probe at 19.8 m (HMP45c, Campbell-Scientific Inc.) and a thermistor at 3 m (Model 107 temperature probe, Campbell-Scientific Inc.). The instruments were recorded by a datalogger (CR10X, Campbell-Scientific Inc.) every 1 s and averaged over 15 min intervals.
The spatial variability in air temperature within the watershed was measured using 30 thermistors located on a transect that stretched along the (east–west) axis of the watershed (Fig. 1a). Each of the 30 thermistors (HOBO-Pro model #H8-031-08, Onset, Bourne, MA) was installed in a radiation shield (part# RS1, Onset) 0.5 m from the ground and recorded the air temperature once every 15 min on its individual datalogger.
During an intensive sampling period in August 2006, the vertical profile of air temperature above the canopy top was measured using a tethersonde balloon carrying a sensor consisting of a copper–constantan thermocouple (TC6-T, Onset) and a pressure transducer (SBPA- CM10, Onset) monitored at 1 Hz by a small datalogger (H12, Onset). The tethersonde was raised and lowered through a small clearing approximately 100 m from the tower. The vertical profile of wind speed and direction was monitored simultaneously using a sodar (Scintec, MFAS) co-located at the tethersonde site. The sodar provided averaged 15 min wind speed and direction up to 105 m; data were stored on a laptop computer.
The variation of CO2 concentration with height was monitored 1 day per week at the base tower using a profiling system controlled by a data logger (CR21x, Campbell-Scientific Inc.). Concentrations of CO2 were measured every second at one of five heights (3, 10, 20, 30 and 37 m) using an infra-red gas analyzer (IRGA) (Li-6262, Li-Cor Inc., Lincoln, NE, USA). The concentrations were monitored at each height for 1 min, with the average CO2 concentration for the last 30 s being recorded by the data logger. Inlets at the remaining heights that were not being monitored by the IRGA were purged continuously by a second pump.
