Increased water demand and climate change impacts have recently enhanced the need to improve water resources management, even in those areas which traditionally have an abundant supply of water, such as the Po Valley in northern Italy. The highest consumption of water is devoted to irrigation for agricultural production, and so it is in this area that efforts have to be focused to study possible interventions. Meeting and optimizing the consumption of water for irrigation also means making more resources available for drinking water and industrial use, and maintaining an optimal state of the environment. In this study we show the effectiveness of the combined use of numerical weather predictions and hydrological modelling to forecast soil moisture and crop water requirement in order to optimize irrigation scheduling. This system combines state of the art mathematical models and new technologies for environmental monitoring, merging ground observed data with Earth observations from space and unconventional information from the cyberspace through crowdsourcing.