Calibration and verification of an irrigation return flow hydrosalinity model

Abstract

A modified irrigation project model (first developed and tested in its original form by Tanji in 1977) is presented which evaluates the impacts of present as well as future alternative land and water resource management on the volume and salt concentration in the irrigation return flow. Modifications were necessary to fit more closely into the prevailing hydrologic and salinity conditions in the Ebro River Basin (Spain), and include the initial and final stored soil water, rim inflows from adjacent hydrologic systems and contributions of gypsum and soluble salts to salt loadings.

The model (which requires a total of 25 input variables) was calibrated in a 4,000 ha irrigation project where the salinity problem is not serious (average EC e is 2.2 dS/m), although gypsum is present in large quantities. Results show that model calibration and verification were, respectively, within −2% and +13% of measured surface irrigation return flows (volume and TDS, total dissolved solids) from the 1982 (calibration) and 1983 (verification) hydrological years.

The performed sensitivity analysis shows that the key parameter for minimizing salt loading (at present one ha yields about 20 t of salt per year) in irrigation return flows is an increase in water application efficiency (WAE), which presently is only 48%. This increase in WAE will not lead to a substantial rise in the TDS of return flows because of gypsiferous character of these soils.

The purpose of this paper is to describe modifications of an irrigation project hydrosalinity model (Tanji 1977) introduced to adapt it more closely to the hydrologic and salinity conditions prevailing in the Ebro River Basin, and to present their calibratio/a and validation steps obtained with data gathered in the Violada Irrigation District (Faci et al. 1985). This model has been programmed in BASIC language on a small Hewlett-Packard computer (HP-85) with 32 K of random access memory (RAM).