Scaling Sensor Networks to Estimate Horticultural Crop Water Use in a Watershed in Ecuador
Scaling Sensor Networks to Estimate Horticultural Crop Water Use in a Watershed in Ecuador
Monday, July 28, 2014: 9:15 AM
Salon 5 (Rosen Plaza Hotel)
We have deployed four weather stations in the Dulcepamba watershed in Bolivar province, southern Ecuador, to estimate the water demand of horticultural crops for farmers, who are faced with restrictions on irrigation water use concessions. This effort is part of a larger project that is also gathering stream-flow and socio-economic data from farmers in this area. The situation has arisen as a private hydroelectric project has been granted a concession of 6.3 m3 per second of water at the confluence of the two major rivers at the base of the watershed, for the next 50 years. Each weather station is deployed in a different climatic zone in the watershed; each consists of an EM50G data logger (Decagon Devices, Inc.), and sensors measuring air temperature, relative humidity, vapor-pressure deficit, wind speed and direction, total radiation and rainfall. One-minute environmental data are averaged and logged every 5 minutes. The logged data are transmitted over the internet to a cloud server via a subscriber identity module (SIM) card in the EM50G logger, which is configured for the local cellular service provider. All data from the server are then downloaded into Sensorweb software (Carnegie Mellon) to a computer located in College Park, MD. Individual crop water demand is then computed using the FAO Penman-Monteith model, which has been incorporated as a “growing tool” within Sensorweb. The daily rainfall totals and crop water demands are calculated, based on average crop coefficient (Kc) values, and used to estimate the weekly irrigation water need for cacao, banana, blackberry, bean in each section of the watershed. The difference between weekly crop water demand and rainfall data are being aggregated in a simple, open access website so the community can easily access this information, to quickly determine whether they need to irrigate their crops or not. The study is also quantifying total acres of irrigated land for each crop (using satellite imagery and ground-truth data in a geographic information system database), so the seasonal data can be scaled for total crop irrigation water demand. Periodic stream flow measurements are also being made on the major rivers in the watershed to quantify water availability. This will be used to estimate whether the total water flow will be adequate to support the prioritized needs of farmers for crop irrigation water concessions, in relation to the water concession awarded to the hydroelectric company.