A spatially explicit risk model to evaluate future drought and flood impacts on ecosystems services provided by flow-irrigation systems: a case study in northeast Italy

Traditional flow-irrigation systems relying on water canals – primarily used for supporting agriculture purposes – supply multiple ecosystem services (ES). However, their capacity to deliver ES is threatened by climate change. The Veneto region, located in the northeast of Italy, is experiencing severe increases in drought periods followed by intense rainfall, which are undermining its dense and complex network of flow-irrigation canals.  Despite the urgency of this situation, the exposure of risk and the consequences on the irrigation systems remains unknown, and so its impact on ES. Spatially explicit models become prominent to evaluate future climate-induced events and potential consequences on ES provided by flow-irrigation systems. Results from those models can inform decision makers and planners to prepare better and efficient adaptation strategies, which will include protecting and maintaining ES.

The aim of this study is to identify and localize areas where ES are more likely to be affected by flood and drought risk in future scenarios (years 2050 and 2100). The model has been built by using k.LAB technology of ARIES (Artificial Intelligence for Environment and Sustainability), an open-source artificial intelligence (AI) modeling framework. By leveraging semantics and machine reasoning, k.LAB enables the integration of independent models and datasets. Moreover, it  automatically assembles spatially explicit models into the spatial scale most appropriate for the context of analysis. By conceptualizing risk as a function of hazard, exposure and vulnerability, our methodology uses spatial multi-criteria analysis to aggregate multi-dimensional information into a single parameter output map.

The study resulted in three major findings. First, the model outputs predicted the impact of droughts and floods on ES provided by the irrigation system. Second, risk maps show the future distribution of both hazards at the level of the water canal spatial unit. Third, hotspot maps identify where ES will be more likely threatened by floods and droughts. We conclude our study by discussing how policy makers and planners can effectively use these analyses to guide better plans.