Abstract: Studying the start (SOS) and end (EOS) of the vegetation growing season can improve vegetation prediction under climate change. Climatic factors have temporal effects on vegetation growth, including the no time effect (no), time-lag effect (lag), time-accumulation effect (acc), and both time-lag and -accumulation effects (lagacc). The linear regression equations between SOS/EOS and climatic factors were constructed for the Loess Plateau (LP). Subsequently, we analyzed the effects of single and multiple climatic factors on vegetation phenology under four temporal effect scenarios and investigated the response of vegetation phenology to the time-lag and time-accumulation effects of climatic factors, under the lagacc scenario. Among the four temporal effects, lagacc explained the effects of climate on vegetation phenology to the greatest degree, and it is the optimal temporal effect for simulating the relationship between vegetation phenology and climate on the LP. Moreover, the explanation degrees of multiple climatic factors were higher than those of single climatic factors across different temporal effects and vegetation types. Simultaneously considering multiple climatic factors improved predictability of their impact on vegetation phenology. Under lagacc, responses of SOS to temperature (TMP) and precipitation (PRE) exhibited 1.44 ± 0.43/3.49 ± 0.85 and 1.38 ± 0.30/3.38 ± 0.71 months lag/accumulation, respectively, and responses of EOS to TMP and PRE exhibited 1.35 ± 0.17/2.37 ± 0.34 and 1.59 ± 0.19/4.16 ± 0.50 months lag/accumulation, respectively, across the entire LP. The results show that both lagacc and multiple climatic factors require consideration when establishing relationships between climatic factors and vegetation phenology.

Key words: climatic factor, start of the growing season, end of the growing season, temporal effect, vegetation phenology, Loess Plateau