Until recently, water level measuring systems typically used an analog-to-digital recorder (ADR) driven by a float within a stilling well. A "stilling" well quiets the water surface in the immediate vicinity of the water level sensor. Water level data were recorded on punched paper tape at 6-minute intervals (5-minutes for the Great Lakes). A schematic of a typical installation is presented below although there might be any number of variations in an actual station's configuration. ADR timing was controlled by solid-state timers accurate to one minute per month. However, typical operational timing accuracies were to the nearest tenth of an hour because tide observers did not reset the timers unless they were more than 6 minutes off watch time. Backup gauges were nitrogen gas pressure driven bubblers gauges recording on pen and ink strip charts with timing controlled by spring wound clock mechanisms accurate to several minutes per month (other types of float-driven analog gauges were used as backup on the Great Lakes). Data were typically manually prorated for timer drifts prior to processing. For tide stations a typical stilling well consisted of a 12-inch diameter pipe at the bottom of which would be attached a 1-inch diameter cone orifice (Great Lakes stations used a sump well located inland with a horizontal pipe connecting the sump to the open water). Inside the stilling well would be an 8-inch diameter float suspended by wire from the ADR unit above.
Each measurement was a discrete instantaneous value measured when the wire leading to the float was mechanically locked in place while the ADR unit punched the paper tape with a binary code representing the value of the water level. Measurements were recorded with 0.01 foot resolution. Local tide observers made timer checks and resets, and recorded from three to six tide staff/electric-tape-gauge (ETG) reading per week. The tide staff/ETG readings served as independent checks of water level and were used to calculate monthly setting which were applied to the gauge reading to refer them to station datum. Second-order class I geodetic levels were run annually from the tide staff/ETG to a nearby network of bench marks to monitor vertical stability of the tide station and its support structures (i.e., the pier and pilings) relative to land. Field groups completed yearly performance tests on the gauges along with required corrective and preventive maintenance procedures (e.g., clean biological fouling from the orifice and stilling well surfaces).