Soil water management: evaluation of infiltration in furrow irrigarion systems, assessing water and salt content spatially and temporally in the Parc Agrari del Baix Llobregat area.

Sustainability of irrigated agriculture is a growing concern in the Baix Llobregat area. Although irrigated land accounts for a substantial proportion of food supply to the local market, it has been, and still is increasingly degraded by poor agricultural management. This dissertation focuses on ways to evaluate furrow irrigation and to assess soil water content and soil salinity (temporally and spatially) under usual farmers's management practices. This dissertation meets these goals through an extensive study of relevant literature and the implementation of practical research. The latter was carried out with a case study on representative fields of the area. Empirical and stochastic models were applied to evaluate furrow irrigation as well as to monitor water flow and solute transport in the root zone. This research produced a number of key findings: first, evaluating furrow irrigation confirmed that 40-43 % of the applied water would have been saved in the study fields if irrigation was stopped as soon as soil water deficit was fully recharge taking the amount of water needed for salt leaching into account, and that the application efficiency (AE) would increase from 48% to 84% and from 41% to 68% (Field 1 and Field 2, respectively). Second, the predictions of soil water content using ARIMA models were logical, and the next irrigation time and its effect on soil water content at the depth of interest were correctly estimated. Third, considering the linear relationship eb-sb, by transforming the Hilhorst (2000) model, which is based on the deterministic linear relationship eb-sb, into a time- varying Dynamic Linear Model (DLM) enabled us to validate this relationship under field conditions. An offset esb=0 value was derived that would ensure the accurate prediction of sp from measurements of sb. It was shown that the offset esb=0 varied for each depth in the same soil profile. A reason for this might be changes in soil temperature along the soil profile. The sp was then calculated for each depth in the root zone. Fourth, by using a (multiple input--single output) transfer function model, the results showed that soil water content and soil temperature had a significant impact on soil salinity, and soil salinity, predicted as a function of soil water and soil temperature, was correctly estimated. Finally, applying the analysis of variance (ANOVA), the results showed that the irrigation frequency, according to the farmer's usual management practice, had statistically significant effects on soil salinity behaviour, depending on soil depth and position (furrow, ridge). Moreover, it was shown that at the end of the crop's cycle the farmers left the field with less soil salinity, for each depth, than at the beginning of the crop's agricultural cycle