Evaluating the potential of desalinated irrigation in water-stressed regions through optimized planting dates and irrigation strategies

With the looming water crisis and its adverse impacts on the food and agriculture industry, exploring the potential of unconventional resources like desalinated water with a fresh perspective is essential. The idea of irrigation using desalinated water has gained increasing traction recently. To maximize the potential of this idea, is essential to appreciate the nuances of using desalinated water for irrigation. In particular, using desalinated water can help avoid dependency on the hydrological cycle, for instance, it may be possible to optimize planting dates and irrigation scheduling to increase agriculture production beyond what would otherwise be possible. As a proof-of-concept, this paper showcases that idea using a case study of tomato production in the Atacama region. A python-based crop growth model, AquaCrop-OSPy, represents the crop responses to irrigation strategies under local hydro-climatic conditions. The self-tuning multi-layer (STML) algorithm was linked to this model to identify the optimal deficit irrigation scheme. This was repeated for different initial planting dates. The impact of soil property uncertainties on the results was analyzed using a Monte Carlo simulation. The results highlight that adjusting the planting dates along with optimal deficit irrigation could indeed improve water productivity, which could, in turn, help offset the costs of using desalinated water. Leveraging economies of scale, both in agriculture and in desalinated water production, could further offset some of the costs. • A novel framework optimizes desalinated irrigation strategies using the self-tuning multi-layer (STML) algorithm • Coupling of STML algorithm and AquaCrop-OSPy enables autonomous schedule optimization • Flexible planting dates improve water productivity and crop yield by leveraging desalination • Economic analysis shows viability under variable market and desalination costs • Monte Carlo simulations assess robustness of crop performance under soil variability