TY - JOUR
T1 - Using time series of MODIS land surface phenology to model temperature and photoperiod controls on spring greenup in North American deciduous forests
AU - Moon, Minkyu
AU - Seyednasrollah, Bijan
AU - Richardson, Andrew D.
AU - Friedl, Mark A.
N1 - Publisher Copyright:
© 2021 Elsevier Inc.
PY - 2021/7
Y1 - 2021/7
N2 - The timing of leaf emergence in temperate and boreal forests is changing, which has profound implications for a wide array of ecosystem processes and services. Spring phenology models, which have been widely used to predict the timing of leaf emergence, generally assume that a combination of photoperiod and thermal forcing control when leaves emerge. However, the exact nature and magnitude of how photoperiod and temperature individually and jointly control leaf emergence is the subject of ongoing debate. Here we use a continuous development model in combination with time series of land surface phenology measurements from MODIS to quantify the relative importance of photoperiod and thermal forcing in controlling the timing of canopy greenup in eastern temperate and boreal forests of North America. The model accurately predicts biogeographic and interannual variation in the timing of greenup across the study region (median RMSE = 4.6 days, median bias = 0.30 days). Results reveal strong biogeographic variation in the period prior to greenup when temperature and photoperiod influence greenup that covaries with the importance of photoperiod versus thermal controls. Photoperiod control on leaf emergence is dominant in warmer climates, but exerts only modest influence on the timing of leaf emergence in colder climates. Results from models estimated using ground-based observations of cloned lilac are consistent with those from remote sensing, which supports the realism of remote sensing-based models. Overall, results from this study suggest that apparent changes in the sensitivity of trees to temperature are modest and reflect a trade-off between decreased sensitivity to temperature and increased photoperiod control, and identify a transition in the relative importance of temperature versus photoperiod near the 10 °C isotherm in mean annual temperature. This suggests that the timing of leaf emergence will continue to move earlier as the climate warms, and that the magnitude of change will be more pronounced in colder regions with mean annual temperatures below 10 °C.
AB - The timing of leaf emergence in temperate and boreal forests is changing, which has profound implications for a wide array of ecosystem processes and services. Spring phenology models, which have been widely used to predict the timing of leaf emergence, generally assume that a combination of photoperiod and thermal forcing control when leaves emerge. However, the exact nature and magnitude of how photoperiod and temperature individually and jointly control leaf emergence is the subject of ongoing debate. Here we use a continuous development model in combination with time series of land surface phenology measurements from MODIS to quantify the relative importance of photoperiod and thermal forcing in controlling the timing of canopy greenup in eastern temperate and boreal forests of North America. The model accurately predicts biogeographic and interannual variation in the timing of greenup across the study region (median RMSE = 4.6 days, median bias = 0.30 days). Results reveal strong biogeographic variation in the period prior to greenup when temperature and photoperiod influence greenup that covaries with the importance of photoperiod versus thermal controls. Photoperiod control on leaf emergence is dominant in warmer climates, but exerts only modest influence on the timing of leaf emergence in colder climates. Results from models estimated using ground-based observations of cloned lilac are consistent with those from remote sensing, which supports the realism of remote sensing-based models. Overall, results from this study suggest that apparent changes in the sensitivity of trees to temperature are modest and reflect a trade-off between decreased sensitivity to temperature and increased photoperiod control, and identify a transition in the relative importance of temperature versus photoperiod near the 10 °C isotherm in mean annual temperature. This suggests that the timing of leaf emergence will continue to move earlier as the climate warms, and that the magnitude of change will be more pronounced in colder regions with mean annual temperatures below 10 °C.
KW - Bayesian
KW - Climate change
KW - Deciduous forests
KW - Hierarchical modeling
KW - Land surface phenology
KW - MODIS
KW - Photoperiod
KW - Spring phenology
KW - Temperature sensitivity
UR - http://www.scopus.com/inward/record.url?scp=85104652400&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85104652400&partnerID=8YFLogxK
U2 - 10.1016/j.rse.2021.112466
DO - 10.1016/j.rse.2021.112466
M3 - Article
AN - SCOPUS:85104652400
SN - 0034-4257
VL - 260
JO - Remote Sensing of Environment
JF - Remote Sensing of Environment
M1 - 112466
ER -