Uncertainty is the quantification of the unknown, a key process in science. Quantifying uncertainty and how it propagates through our stack of data to our prediction layers is important to identify the source of uncertainty. SPUN research is guided by uncertainty, helping us map the Earth’s least explored ecosystems: namely, where environments and ecoregions are not covered by our current database, where our predictions are have high uncertainty, but are not environmentally unique, and sub-pixel error where we lack information below our prediction resolution

A mycorrhizal fungi uncertainty map is a spatial or predictive model used to visualize the statistical margin of error, data gaps, or reliability of global fungal biodiversity predictions.

We get this done by combining large geo-located databases of mycorrhizal diversity and ecological variables to generate spatial predictions of mycorrhizal diversity. This allows us to quantify and map the uncertainty of these model predictions and identify under-sampled ecoregions and help guide future mycorrhizal research across the globe.

Mapping mycorrhizal fungi involves feeding billions of DNA sequences and environmental data into machine-learning algorithms. Mapping these hidden networks involves significant uncertainty because over 70% of global ecosystems remain unsampled. SPUN is mapping this statistical uncertainty, with metrics like the coefficient of variation, to identify sampling gaps and guide conservation priorities.