Calibration and Global Sensitivity Analysis for a Salinity Model Used in Evaluating Fields Irrigated with Treated Wastewater in the Salinas Valley

Treated wastewater irrigation began two decades ago in the Salinas Valley of California and provides a unique opportunity to evaluate the long-term effects of this strategy on soil salinization. We used data from a long-term field experiment that included application of a range of blended water salinity on vegetables, strawberries and artichoke crops using surface and pressurized irrigation systems to calibrate and validate a root zone salinity model. We first applied the method of Morris to screen model parameters that have negligible influence on the output (soil-water electrical conductivity (EC_sw)), and then the variance-based method of Sobol to select parameter values and complete model calibration and validation. While model simulations successfully captured long-term trends in soil salinity, model predictions underestimated EC_sw for high EC_sw samples. The model prediction error for the validation case ranged from 2.6% to 39%. The degree of soil salinization due to continuous application of water with electrical conductivity (EC_w) of 0.57 dS/m to 1.76 dS/m depends on multiple factors; EC_w and actual crop evapotranspiration had a positive effect on EC_sw, while rainfall amounts and fallow had a negative effect. A 50-year simulation indicated that soil water equilibrium (EC_sw ≤ 2dS/m, the initial EC_sw) was reached after 8 to 14 years for vegetable crops irrigated with EC_w of 0.95 to 1.76. Annual salt output loads for the 50-year simulation with runoff was a magnitude greater (from 305 to 1028 kg/ha/year) than that in deep percolation (up to 64 kg/ha/year). However, for all sites throughout the 50-year simulation, seasonal root zone salinity (saturated paste extract) did not exceed thresholds for salt tolerance for the selected crop rotations for the range of blended applied water salinities.

Abstract licence: CC BY

38 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.