Low-Cost Sensor Network for Stream Flow Monitoring in the Alto Beni Region of Bolivia

Lack of data is a persistent problem in hydrology and other field work in developing countries. Low cost monitoring devices allow investigators to maximize spatial coverage on a limited budget, as well as to minimize the financial risk of loss of instruments placed in vulnerable locations. This work contributes to an ongoing project to evaluate the sustainability of discharge from springs supplying gravity-fed potable water systems in the Alto Beni region of Bolivia where land use is rapidly changing from forest to agriculture. The approach is to estimate ground water recharge as a function of land use variables using a water balance model in several representative watersheds. Monthly stream discharge is currently estimated using monthly manual measurements of water level by a local technician. Continuous water level measurements will allow an improved estimate of the cumulative discharge, and generate data on statistical distribution of daily flow that may be useful to estimate discharge in similar, ungaged watersheds. Continuous water level measurements, along with available precipitation data, will allow analysis and comparison of the response of watersheds to individual precipitation events as a function of land use variables. We assembled a low cost level logging system for stream flow monitoring that measures and records distance up to 6 m to the nearest 25 mm every ten minutes, and runs for a month on six rechargeable AA batteries. The system consists of a sonic range finder sensor (MaxSonar-EZ2, Maxbotix Inc., Baxter, MN, 30), a temperature sensor (MCP9701, Microchip Technology Inc., Chandler, AZ, 0.25), and a datalogger (Hobo U12, Onset Computer Corp., Pocasset, MA, 104) along with a weather-resistant enclosure and common items for a total cost of 230 per unit. The level loggers were attached to bridges over three subject streams. A local technician visits the sites monthly to download data, replace the rechargeable batteries, and take a manual water level measurement. Simultaneous recording of temperature is necessary to correct the distance measurement for the temperature dependence of the speed of sound in air. We report the data processing, temperature correction procedure, and performance of the system relative to manual measurements and discuss suitability of the system for hydrological applications in other settings.