4th UF Water Institute Symposium Abstract

Submitter's Name Nathan Holt
Session Name Poster Session: Water Security: the water-energy-food nexus
Poster Number 64
Author(s) Nathan Holt,  University of Florida, ABE (Presenting Author)
  Sanjay Shukla,  University of Florida
  James  Knowles, University of Florida
  Developing water and nutrient efficient bed geometries for raised bed, plastic mulch crop production systems using a vadose zone model.
  The traditional bed geometry currently used in the raised bed, plastic mulch crop production system common to Florida tends to be short and wide. The dimensions of the beds were initially developed for seepage based systems and the geometries have remained relatively constant over the years even with the increased usage of drip irrigation/fertigation to supply water and nutrients. This study attempts to challenge the status quo by testing narrower and taller alternative bed geometries believing these alternatives can provide multiple benefits over the current traditional bed geometry including extra water storage capacity, protection against nutrient loss and root damage from flooding that can occur after a tropical rainfall event, better disease protection, decreased fumigation cost, yield protection, and plastic mulch savings. This presentation aims to highlight a HYDRUS (2D) vadose zone model that in conjunction with measured water and nutrient data, evaluated alternative bed geometries with respect to water fluxes. Measured data was collected throughout a growing season at a tomato production farm near Immokalee, Florida. Three alternative bed geometries were monitored along with the standard bed used by the grower. The model was calibrated to real time soil moisture levels measured at various depths near the root zone in each of the four different bed geometries. Soil moisture characteristic properties, irrigation, weather, and water table depth were collected throughout and after the growing season and used in both model design and simulation. The model was designed to represent the cross section of the four different bed geometries tested. The finite element scheme employed by HYDRUS (2D) allowed for the full representation of the cross sections of the different bed geometries. The boundaries of the model domain were assigned as a no flux boundary on the surface representing plastic mulch covering the bed, atmospheric boundary conditions representing row middles, a variable flux boundary representing the drip tape, a variable head boundary representing the water table, and free drainage along the sides of the model domain. An hourly time step was employed by the model to capture rapid changes in water table levels.