Abstract: Invasive species, especially plant pests, pose growing threats to ecosystems and food security. Geospatial predictive models and early alerts help manage these risks, but gaps, biases, and delays in global pest observation data introduce uncertainty into forecasts. Web news and social media offer abundant text information, potentially documenting past and emerging pest outbreaks in near-real time. We assess how uncertainty is quantified in spatiotemporal forecasts of invasive species and address data gaps by creating an open dataset that consolidates official pest records from disparate sources. Through two case studies, we show how web media can provide valuable geospatial data on spreading pests. Our findings offer practical guidance, new datasets, and a flexible text processing workflow to support global pest surveillance.