False springs and spring phenology: Sensitivity to downscaling techniques and training data

Abstract:

Spring plant and animal life cycles (otherwise called phenology) and false springs are of particular interest with respect to the impacts of climate change on numerous species, such as migratory birds in the central flyway. Using a suite of maximum and minimum temperature projections created for the south-central United States, we examine the sensitivity of projections for spring phenology and false springs to the multiple components used to create the climate projections. Multiple statistical downscaling techniques and their associated training datasets, along with three GCMs and representative concentration pathway (RCP) 8.5, are used as inputs to calculate the Extended Spring Indices (SI-x) and the False Spring Exposure Index (FSEI) during historical (1981-2005) and future (2070-2099) periods. The FSEI is also divided further into an early FSEI (EFSEI) and late FSEI (LFSEI) for false springs occurring after the first leaf and first bloom, respectively. As one would expect, the first leaf and first bloom dates are projected to occur more than a week earlier in the year in all the projections. However, different downscaling techniques can have a profound impact on projected changes for the FSEI, particularly for the LFSEI. While first leaf and first bloom are primarily connected with the accumulation of heat over time, the FSEI is sensitive to the timing of the occurrence of freeze events. These cold temperature events, which occur in the tail of the minimum temperature distribution for this region, can be difficult for some downscaled projections to represent adequately. User’s assessing future changes in spring phenology or false springs for impacts to species and habitats should carefully consider if available projections adequately represent the frequency of cold temperature events prior to using the projections for impact assessments.