Accurate environmental and climate modelling depends on the availability of large observational datasets, yet the generation of such data is often costly and logistically challenging for individual research groups. This limitation is particularly acute in New Zealand, where geographic isolation, complex topography, and strong climatic gradients require dense observational networks for robust national and regional scale modelling. Here we present New Zealand's national precipitation isotope database, compiled through a community data-collection effort. The database integrates historical and contemporary datasets with varied analytical methods, sampling strategies, and temporal coverage. To enable combined use of these data, we describe a standardisation framework, including weighting using gridded precipitation data, and alignment of sampling periods to calendar months. We then show how the database updates and improves the Precipitation Isotopes New Zealand (PINZ) machine-learning isoscapes. Incorporation of newly contributed data, particularly from inland and high-elevation regions, reduces overall root mean square error relative to earlier model versions and substantially expands the model's area of applicability in New Zealand. Our results highlight that further advances will benefit from the availability of submonthly precipitation isotope measurements and improved coverage of alpine environments. We show that community participation offers a practical pathway to strengthening national precipitation isotope products and their application across hydrology, climate science, and related environmental science disciplines.