The USA-NPN's Extended Spring Indices, produced as map layers for the United States, combine modeled leaf out and bloom dates for lilacs and honeysuckles, resulting in a day of year for the start of spring. To assess the plausibility of the species sub-models that comprise the Indices, we compare modeled output to observations of breaking leaf buds and open flowers on lilac and honeysuckle plants submitted by observers.

The dashboard below shows the difference, in days, between the date of breaking leaf buds or open flowers reported for individual lilac and honeysuckle plants via Nature's Notebook to the date the event was predicted for the plant's location by the corresponding Extended Spring Index sub-model.  

Negative values (observed-predicted) indicate that the Spring Index sub-model predicted the event (leaf out or bloom) later than an observer reported it. Positive values (observed-predicted) indicate that the Spring Index sub-model predicted the event earlier than it was reported by an observer.

• Top dashboard: Modeled vs observed plant phenology (models use PRISM temperature data; available 1981-2020)
• Bottom dashboard: Modeled vs observed plant phenology (models use NCEP URMA temperature data; available 2016-2021)
• Zabelii lilac observations ended in 2008.

For more details on our approaches and plans for estimating uncertainty in our gridded products, see Crimmins et al. (2017).