Achieving the Sustainable Development Goal (SDG) 6 is a major challenge as ensuring sustainable and universal access to water faces considerable demographic and climatological uncertainties. Limited data on water availability and water use is a key impediment to evidence- based decision making. Emerging hydrological monitoring technologies as well as improved modelling tools offer an opportunity to bridge this gap. However, hydrological evidence is not always perceived as conducive to sustainable water resources management and governance. Instead, it can lead to complex political reactions and increased conflict. I hypothesized that whilst hydrological data are a necessary pre-condition to improved decision-making, shaping such data into actionable evidence requires a combination of effective tools and broader policy considerations to ensure sustainable water resources management and governance. To test this hypothesis, I focused on the tropical Andes, a region facing major water-related challenges, significant anthropogenic pressures, and limited governance capacity. First, I designed a procedure to derive spatially explicit maps of anthropogenic water demand in data-scarce regions, a key element of the water cycle with substantially limited information across the globe. I then coupled this dataset with the physics-based land-surface model JULES to simulate the 2016 drying of Lake Poopó in Bolivia. I found a clear impact of anthropogenic water demand on the evolution of the lake, highlighting the importance of hydrological data to water management. I therefore subsequently examined the underlying assumptions on data availability in various water resources governance arrangements. I studied the Ecuadorian capital Quito to derive criteria to diagnose potential governance deficits when considering an increase in hydrological data availability. Lastly, I identified a paradox between apparently sub-optimal level of monitoring activities and their potentially high economic returns. I attributed this paradox to key limitations in the methods used to value monitoring benefits, which I attempted to address by developing a risk-based framework, illustrated using the case of the Rímac river basin in Lima, Peru. The outcomes of this thesis help overcome potential bottlenecks in data availability to support evidence-based decision making in order to achieve the SDGs.