Pitfalls and possible solutions for using geo-referenced site data to inform vegetation models

摘要: Abstract Most predictive models rely on ‘the known’ to infer unknown’. Geo-referenced, on-ground observational data are the ‘point of truth’ upon which many vegetation built. We focus some enigmatic errors that we have uncovered when using plot data. Using a case study, sourced 9362 sites examine prevalence spatial errors. found an incorrect datum was recorded for 5% sites; less than 2% were duplicated and up 34% located within 1000 m each other. Whilst neighbourhood not necessarily errors, they do need be considered context environmental layers modelling. offer solutions identifying managing locations point ensure information-rich resource held in repositories is compromised by unidentified error.