Research On Seasonal Dynamics And Phenological Characteristics Of Urban Forests Based On Phenological Cameras

Urban forests are an important component of urban ecosystems.Due to the photosynthesis capacity of vegetation in the urban forests,they have a remarkable regulating effect on atmospheric CO₂.With global climate change,there are potential uncertainties in the carbon absorption capacity of urban forests,and these uncertainties are likely to be reflected in the phenological processes of urban forest.Therefore,the study of the growth state and phenology of the urban forests will benefit us in gaining further understanding of carbon and water fluxes in the ecosystem and may lead us to respond more scientifically to global climate change,which has vital significance for more rational forest management in the future.In this study,based on the phenological camera images and eddy covariance flux data from the observation research station which locates in urban forest,vegetation growing seasonal dynamics and phenology information are extracted,and the relationships between phenology and carbon flux,water flux are discussed.Firstly,images from the phenological camera are used to calculate multiple vegetation indices on community,population and individual scale,including GCC(Green Chromatic Coordinate),RCC(Red Chromatic Coordinate),and VIgreen(Green Vegetation Index).The seasonal dynamics of vegetation indices represents the changing status of vegetation growing pattern.Secondly,after comparing the results of different curve fitting and phenophases extraction methods,phenological dates of the urban forest are extracted using the combination of the methods of Klosterman and Gu,including UD(Upturn date),SD(Stabilization date),DD(Downturn date)and RD(Recession date).Thirdly,the intra-season variation dynamics of carbon and water fluxes are analyzed,as well as the correlation relationships between vegetation indices,carbon flux,and water flux.The main conclusions obtained are as follows:(1)GCC shows a consistent changing trend under all three study scales,showing an obvious unimodal growth trajectory,also shows the applicability to both the evergreen broad-leaved tree species Cinnamomum camphora and the deciduous tree species Bischofia polycarpa.Compared with Cinnamomum camphora,RCC have a better characterization of the changing growth states of Bischofia polycarpa.RCC is more sensitive to the changes in vegetation growth during winter and spring when compared to summer and autumn.The seasonal course of VIgreen has a similar trend as GCC,but it has weaker seasonal changes.(2)After comparing results from different methods of curve fitting and phenophases' extracting,the optimal combination have been selected,which is using the Klosterman method for growth curve fitting and applying the Gu method for extraction of phenophases.This combination suits both of the evergreen broad-leaved tree species Cinnamomum camphora and the deciduous tree species Bischofia polycarpa on all study scales.(3)In the study area,the phenophases results of Cinnamomum camphora are highly consistent on community-scale and population-scale.The results among the three populations of Cinnamomum camphora are basically consistent as well.Averagely,the growing season of the study area begins in late April(DOY 108)and stabilizes in early May(DOY 121),starts recession in early October(DOY 283),and ends in late October(DOY 293),with an average season length of 185 days.The phenological characteristics of Cinnamomum camphora and Bischofia polycarpa populations are significantly different.Bischofia polycarpa population has a longer season length(the mean gap is 46 days),enters the growing season earlier,and ends it later after reaching DD.Among 3 study scales,the LOS of Cinnamomum camphora at the individual scale is the longest,and DD appears the first while RD shows at the latest.The phenophases of the study area have a changing pattern: UD and SD results are similar on individual and other study scales,as well as between different Cinnamomum camphora individuals.On the other hand,DD,RD,and LOS results are high diversity in the same situation,and the degree of difference is:LOS>RD>DD.(4)The total carbon flux in the growing season has a negative value,which indicates the plantation forest ecosystem is a carbon sink.The maximum carbon uptake occurs at the end of June(DOY 180)(-3.3 ?mol·m-2·s-1).From July to September,the intensity of CO₂ uptake is gradually weakening.The CO₂ flux fluctuated the strongest during UD and SD,and the maximum gap is 2.7 ?mol·m-2·s-1.The water flux of the plantation ecosystem shows a conspicuous unimodal curve during the study period.From SD to the end of the study(DOY 273),the highest value of H2 O flux appears in late July(DOY 208)(3.69 mmol·m-2·s-1);the lowest value appears in mid-September(DOY 256)(-0.36 mmol·m-2·s-1).The fluctuation range of H2 O flux during UD and SD is the largest during the study period.There is a significant positive correlation between CO₂ flux and GCC during the growing season(P<0.01),and the correlation coefficient was 0.270.CO₂ flux had a certain correlation with VIgreen and RCC,and the correlation coefficients are 0.178 and-0.081,respectively.There is a certain correlation between the H2 O flux and the three vegetation indices,and the correlation coefficients are GCC: 0.136,VIgreen: 0.131,and VIgreen:-0.080,respectively.