You are here

Harvard Forest >

Harvard Forest Symposium Abstract 2018

  • Title: Multi-Scale Analysis of Trends in Northeastern Temperate Forest Springtime Phenology
  • Primary Author: Minkyu Moon (Boston University)
  • Additional Authors: Mark Friedl (Boston University); Eli Melaas (Boston University)
  • Abstract:

    The timing of spring leaf emergence is highly variable in many ecosystems, exerts first-order control growing season length, and significantly modulates seasonally-integrated photosynthesis. Numerous studies have reported trends toward earlier spring phenology in temperate forests, with some papers indicating that this trend is also leading to increased carbon uptake. At broad spatial scales, however, most of these studies have used data from coarse spatial resolution instruments such as MODIS, which does not resolve ecologically important landscape-scale patterns in phenology. In this work, we examine how long-term trends in spring phenology differ across three data sources acquired at different scales of measurements at the Harvard Forest in central Massachusetts. Specifically, we compared trends in the timing of phenology based on long-term in-situ measurements of phenology, estimates based on eddy-covariance measurements of net carbon uptake transition dates, and from two sources of satellite-based remote sensing (MODIS and Landsat) land surface phenology (LSP) data. Our analysis focused on the flux footprint surrounding the Harvard Forest Environmental Measurements (EMS) tower. Our results reveal clearly defined trends toward earlier springtime phenology in Landsat LSP and in the timing of tower-based net carbon uptake. However, we find no statistically significant trend in springtime phenology measured from MODIS LSP data products, possibly because the time series of MODIS observations is relatively short (13 years). The trend in tower-based transition data exhibited a larger negative value than the trend derived from Landsat LSP data (-0.42 and -0.28 days per year for 21 and 28 years, respectively). More importantly, these results have two key implications regarding how changes in spring phenology are impacting carbon uptake at landscape-scale. First, long-term trends in spring phenology can be quite different, depending on what data source is used to estimate the trend, and 2) the response of carbon uptake to climate change may be more sensitive than the response of land surface phenology itself.

  • Research Category: Regional Studies