Temporal And Spatial Dynamics Of Vegetation Cover And Its Responses To Climatic Factors In The Loess Plateau

Vegetation is a sensitive indicator of terrestrial ecosystem,terrestrial surface energy exchange,and biogeochemical cycles.This paper studied the temporal-spatial vegetation dynamics and its response to climate changes in the central Loess Plateau in the following four aspects.Firstly,we analyzed the trend variation of vegetation growth and climatic factors;secondly,we explored interactive effects between climatic factors on vegetation growth.Thirdly,we examined the dynamics and transaction of land use and land cover in Yan'an area over the past 20 years and predicted the changes of land use and land cover of Yan'an in 2025 using CA-Markov model.Finally,the time-lag effect between precipitation and temperature on vegetation growth was probed at different time-lag scales and site.The main findings are as follows,(1)The vegetation growth showed a strong significant linear growth trend for all seasons(the growing season,spring,summer,autumn and winter)in Yan'an area from2000 to 2015(at a rate of 0.008 for each year in growing season).The precipitation in Yan'an area showed a strong significant upward trend in the spring,with a growth rate of 1.44 mm/yr,and for all seasons except winter precipitation showed an increasing trend.Both the growing season and summer temperature showed a significant downward trend,with a rate of-0.062~oC/yr and-0.073~oC/yr,respectively.Yan'an area had no warming trend in most seasons from 2000 to 2015 and temperature showed a significant decreasing tendency in the growing season and summer.However,there is an upward trend in autumn and winter,which indicated that there was a warm winter tendency in Yan'an from 2000 to 2015.For growing season,the vegetation growth mainly depended on the precipitation at the early stage(April),and then precipitation and temperature played a similar role in influencing the vegetation growth(May).However,TEM started to dominate the vegetation growth at the middle stage(June,July),and then there was no major dominating factor at the late stage(August-October).(2)Correlation analysis between vegetation and climatic factors at the pixel scale showed that in spring vegetation growth and precipitation has a significant positive correlation and these areas are mainly distributed in the western,eastern and central regions of the study area,however,in summer vegetation growth and temperature in most areas had a negative correlation with a significant area of 37.5%.In growing season temperature and precipitation had a strong positive main effect on EVI respectively and had a significant negative interaction ef fect on EVI in growing season.The results of Simple Slope analysis showed that the slope of the simple regression decreased with the increase of precipitation in growing season;when the values of temperature were lower than 20~oC,precipitation played an obvious positive effect of temperature on vegetation growth and the strength of effect decreased as the temperature increased in growing season.The results of the Johnson-Neyman analysis showed that when the values of precipitation was lower than 151.28(observed values was 8.28 to 209.30),the simple slope of temperature on EVI prediction was significantly not equal to zero in growing season.(3)From 1995 to 2005,forest had the largest proportion of changing areas among all land types.The areas of unused land increased sharply,and the land use dynamics degree of building land was 6.54%which indicated that from a large amount of land areas had been used as building land during the period 2005-2015 and at the same time the dynamic degree of cultivated land was-1.45%which indicated that cultivated land was continuously decreasing,which may be due to the policy of returning farmland to forests and grasslands implemented in the Loess Plateau.We predicted the land use and land cover in Yan'an in 2025 using CA-Markov model,and the results showed that the dynamics degree of building land and unused land was the highest among other land types respectively(25.76%and 156.65%).(4)If considering the time-lag effect of precipitation and temperature on vegetation growth,the average interpretation of the prediction model increased by6.57%.Vegetation growth presented an obvious time-lag effect of 3-5 months for the precipitation.The low-density grassland and cultivated land had the shortest time-lag effect for the precipitation,and the woodland and high-density grassland had the longest time-lag effect for the precipitation.Vegetation growth showed an obvious time-lag effect of 1 and 5 month for the precipitation.The low-density grassland had the shortest time-lag effect for the temperature,and the high-low grassland had the longest time-lag effect for the temperature.For the grassland,the average month of the time-lag effect both for the precipitation and temperature increased as its coverage increased.The contribution rate distribution of the optimal combination model of precipitation and temperature showed an obvious periodicity and intermittent aggregation.The optimal combination models with larger model contribution rate were more prone to that the gap of time-lag month of precipitation and temperature was very small or very large.Compared with temperature,the time-lag effect of vegetation growth on precipitation presented the characteristics that the average time-lag month was longer and the time-lag strength was weaker.(5)As shown in the results of the monthly time-lag effect analysis in the growing season,time-lag effect between vegetation and cumulative precipitation showed a large variation in different months of the growing season.In most months of the growing season(except September),the proportion of positive and negative correlation was almost equivalent between vegetation and precipitation when considering lag effect.The average time-lag month of vegetation growth on precipitation showed a decreasing-slow rising-decreasing trend from April to October.The average time-lag month of vegetation growth on temperature exhibited an obvious increasing-decreasing-increasing tendency from April to October.The average cumulative length of precipitation showed a more stable variation over the period of growing season.The average lag time of vegetation growth for different vegetation types on temperature varied greatly in different months of the growing season.The variation of average lag time of vegetation growth on precipitation and the variation of the cumulative precipitation length was small among different vegetation types.