Increasing demands on water supplies, along with concerns about non-point source pollution, and water quality-based ecological factors all point to the need for observing stream flow perturbations and pollutant discharges at higher resolution than has been practical until now. This work presents a rapidly deployable Networked Infomechanical System (NIMS RD) technology for observing spatiotemporal variability in hydraulic and chemical properties across stream channels. NIMS RD is comprised of two supporting towers and a suspension cable delivering power and Internet connectivity for controlling and actuating the tram-like NIMS unit. The NIMS unit is capable of raising and lowering a payload of sensors, allowing a preprogrammed or data-actuated adaptive scan to be completed across a stream channel. This work examines velocity vectors in a real river system using an acoustic Doppler velocimeter (ADV) sensor with positioning controlled by the NIMS RD apparatus. Raster scans providing velocity fields for river cross sections at the San Joaquin River-Merced River confluence are presented for low flow conditions in August of 2006 and 2007. Sensor dwell time is examined in an effort to determine optimal sampling rates for balancing experimental efficiency with the precision of observations. The observed velocity profiles are compared to theoretical vertical and transverse velocity profiles. While some aspects of the real and theoretical profiles agreed qualitatively, the real system exhibited more complicated behavior than the simplistic approach was capable of describing.