Research On Spring Phenology And Its Sensitivity To Climate In Boreal Forests

Research on the response of vegetation phenology changes to global climate change is an important current research direction.This response involves a variety of factors such as climate change,land use change,atmospheric pollution and vegetation type.Specifically,inter-annual variation in vegetation phenology reflects not only the ability of plants to adapt to environmental change,but also their own growth,thus providing a more accurate ecosystem regulation for terrestrial ecosystems.The special characteristics of boreal forests make highly sensitive to climate change,and the traditional vegetation indices,represented by NDVI,invert the phenological period parameters,which mainly reflect the degree of greening of vegetation and cannot accurately reflect the influence of environmental factors such as moisture and heat on photosynthetic rates.In this study,Solar-Induced Chlorophyll Fluorescence(SIF)was compared with traditional vegetation index to explore the applicability of SIF to inversion of phenological parameters in boreal forest areas,analyze the spatio-temporal variation characteristics and trend of spring phenology of boreal forest,and quantify the lag effect of climate on spring phenology based on CRU climate data set.The characteristics of climate sensitivity changing with precipitation gradient and temperature gradient were explored by correlation analysis.The main conclusions of this study are as follows:(1)The NEE＿SOS extracted from FLUXNET NEE observation data was compared with SIF＿SOS and NDVI＿SOS to conduct reliability verification and comparative analysis.Both of them are significantly correlated with FLUXNET NEE＿SOS(p<0.0001),and the correlation between SIF＿SOS and NEE＿SOS is high,with R~2reaching 0.4491.The SOS estimated by different remote sensing data lagged behind NEE＿SOS observed by FLUXNET,but the phenology extracted by SIF could better capture the growth stage of vegetation,and SIF＿SOS was closer to the results of FLUXNET observation than NDVI＿SOS.(2)The average distribution of SIF＿SOS in boreal forests during the past 20years was on the 138th day,from mid-April to the end of June.The start of growing season based on NDVI data and SIF data showed obvious latitude correlation in space,and SOS was gradually delayed with the increase of latitude.In the study area,86.7%of the pixels SIF＿SOS were earlier than NDVI＿SOS.In the variation of SOS mean value in different years,SOS obtained based on SIF data reflecting vegetation photosynthesis characteristics lags behind that obtained based on NDVI data.(3)The strongest correlation between preseason temperature and spring phenology was in the month before the beginning of SIF＿SOS,which fully indicated that temperature had a strong lag effect on spring phenology.NDVI＿SOS had the strongest correlation with the mean temperature in the month when spring phenology occurred,and the proportion of pixel was the largest,about 32.13%.According to the spatial pattern of preseason precipitation length in SIF＿SOS,the response of preseason cumulative precipitation to spring phenology was more dispersed than that of preseason mean temperature,and the correlation was greatest in the current month and the preseason January.However,the number of pixels decreased with the increase of preseason length of boreal forest precipitation calculated by NDVI data.The NDVI＿SOS of vegetation and the accumulated precipitation of last month were the main factors determining the phenological period in this region.(4)In the study area,96.67%of SIF＿SOS temperature sensitivity was negative,that is,temperature increase promoted the start of spring phenology of vegetation,and the temperature sensitivity of vegetation SIF＿SOS showed spatial heterogeneity.In the northeast of North America,the temperature sensitivity of NDVI＿SOS is higher than that of SIF＿SOS,but the temperature sensitivity of NDVI＿SOS is less than that of SIF＿SOS,which shows that SOS advance with preseason temperature increase.The preseason mean temperature is the main factor driving the change of boreal forest SOS.In the study area,32.71%of SOS precipitation sensitivity was negative,and67.29%of pixel SOS precipitation sensitivity was positive.SIF＿SOS showed a significantly delayed trend with the increase of preseason accumulated precipitation.Compared with the spatial distribution of precipitation sensitivity of SIF＿SOS,the spatial variation of sensitivity of NDVI＿SOS to preseason accumulated precipitation was more dispersed.The influence of precipitation on vegetation phenology changes greatly with different climatic conditions.The preseason accumulated precipitation promoted the temperature sensitivity and precipitation sensitivity of vegetation.SOS is more sensitive to preseason temperature and precipitation in high temperature areas.However,the fitting effect of NDVI-SOS precipitation sensitivity and preseason mean temperature is better than that of SIF＿SOS precipitation sensitivity and preseason mean temperature.