【作者】 刘云；

【导师】 李升峰；

【作者基本信息】 南京大学 ， 环境科学， 2018， 硕士

【摘要】 本文利用2001～2012年的GLASS LAI数据并结合遥感气候数据,对全球三大热带雨林地区12年的植被生长及其对气候变化的响应进行研究,采用点面相结合(“大边界,小样地”)、相关性分析与过程分析相结合的研究方法,深入分析热带雨林地区植被生长状况及其对气候变化的响应机理。主要研究结论如下:(1)2001～2012年全球热带雨林研究区植被生长(LAI)总体上呈现缓慢下降趋势。其中全球热带雨林LAI变化以波动趋势不明显区域为主(面积占比54.7%),波动明显区域为辅;LAI波动明显区域中虽上升趋势区域面积占比略多于下降趋势区域,但由于极显著下降趋势区域多于极显著上升趋势区域,从而导致雨林地区LAI总体上呈现缓慢下降趋势。在空间分布上,12年间LAI波动趋势不明显区域与上升区域主要分布在热带雨林核心区,而LAI下降区域主要分布在亚马逊雨林东南边缘,非洲雨林南部及北部边缘,亚洲雨林马来半岛以北等雨林边缘区。(2)2001～2012年间全球热带雨林植被生长(LAI)与年均温的相关关系总体以正相关和弱相关为主(面积占比合计95.4%),与年降水量的关系总体以弱相关为主(面积占比约80%)。(3)2001～2012年间全球热带雨林植被LAI与月均温的关系以正相关和弱相关为主,负相关区域面积占比较低(面积占比约7.2%)且集中分布在亚马逊雨林东南边缘与非洲雨林北部边缘等雨林边缘区;LAI与月降水量的关系以弱相关与负相关为主,年内植被生长与降水量的月相关性和年降水量的空间分布特点存在一定对应关系,即:总体上,高降水量区对应弱相关,中等降水量区对应负相关,低降水量区对应正相关。(4)雨林植被生长与气温、降水月变化的生态响应研究表明:雨林地区存在9种不同响应情景。并基于样地分布位置核心与否、LAI月均值是否大于4等划分依据筛选出典型样地11个,主要反映5种生态响应情景类型。分别是:①LAI与温度正相关、与降水负相关(符号简写为T+P-);②LAI与温度弱相关、与降水负相关(TOP-);③LAI与温度弱相关、与降水弱相关(TOPO);④LAI与温度弱相关、与降水正相关(TOP+);⑤LAI与温度正相关、与降水负相关(T+PO)。其中美洲地区有T+P-、TOPO以及TOP-等3种情景,非洲地区有T+PO与TOP+等2种情景,亚洲地区只选择TOPO这1种情景。(5)2001～2012年间热带雨林典型样地植被生长(LAI)同样总体上呈现平缓波动下降趋势。其中美洲雨林地区典型样地LAI值较高且年际波动较小,非洲与亚洲雨林地区典型样地LAI值相对较低且年际波动相对较大。LAI年内变化可分为双峰波动与单峰小波动两种情景,其中非洲雨林典型样地主要表现出双峰波动,美洲与亚洲雨林典型样地主要表现出单峰小波动。年内LAI波动幅度大小与月均温、月降水量的相关性大小成正对应关系,相关性越大,年内LAI波动幅度越大。(6)对2001～2012年间雨林典型样地LAI与气温降水的生态响应研究表明:年降水量、低月降水量以及年内温度降水波动方向是影响LAI与温度、降水量月相关性的主要生态指标。表现为:T+P-与TOP-型响应情景主要分布在年降水量适中(2200～2500mm)、年内降水波动较大(39～359mm)、且存在2～3个月相对干季(39～90mm)的美洲雨林地区,其中年内温度与降水波动基本反向的样地主要表现为T+P-型响应情景,年内温度与降水波动基本同向的样地主要表现为TOP-型响应情景;T+PO与TOP+型响应情景主要分布在年降水量较低(1500～1900mm)、年内降水波动相对较小(53～238mm)、温度与降水波动不完全同向的非洲雨林地区。其中年内无明显干季的样地主要表现为T+PO型响应情景,年内存在2～3个月相对干季(53～100mm)的样地主要表现为TOP+型响应情景;TOPO型响应情景主要分布在年降水量高(2700～4600mm)、年内降水波动较大(100～600mm)、温度与降水变化不完全同向的美洲与亚洲雨林地区。(7)对2001～2012年间雨林典型样地多年LAI年内显著变化期与温度、降水相关性的进一步研究表明:上述T+P-与TOP-型2种响应情景中,多年LAI年内变化显著期的LAI与同期温度和累计降水量总体均呈负相关;T+PO与TOP+型等2种响应情景中多年LAI年内变化显著期的LAI与温度呈弱正相关,前期降水对植被生长存在滞后与累积效应,即:存在连续3个月降水充沛(累计降水量400～600mm)时,LAI与降水呈负相关;连续3个月降水不够充沛(累计降水量在330mm左右)时,LAI与降水呈正相关;TOPO型响应情景中LAI年内变化显著期与同期温度和累计降水量相关性较复杂,但总体上较弱。(8)月变化响应情景与年内变化显著期结果的关系总共可分为4类:一是与温度、降水相关关系一致,对应TOP-型响应情景;二是与温度相关关系不一致,与降水相关关系一致,对应T+P-型响应情景;三是与温度、降水相关关系均不一致,对应T+PO型;四是与温度、降水相关关系无明显规律,主要对应TOP+与TOPO型响应情景。更多还原

【Abstract】 The paper studied the growth of vegetation and its response to climate change in the three tropical rainforest regions in the world based on the GLASS LAI data and remote sensing climatic data from 2001 to 2012.Using the combination of points and faces("big boundary,small plots"),correlation analysis and process analysis,the paper analyzed the vegetation growth of tropical rain forests and its response mechanism to climate change deeply.The main findings are as follows:(1)From 2001 to 2012,the vegetation growth(LAI)in the global tropical rainforest region showed a slow decline.Among them,LAI changes in the global tropical rain forest are dominated by regions with less fluctuating trends(54.7%of the area),supplemented by regions with significant fluctuations;Although the area of the upward trend is slightly more than the area of the downward trend in the regions where the LAI fluctuates markedly,the area with extremely significant downward trend is more than the area with extremely significant upward trend.As a result,the LAI in the rain forest area generally shows a slow downward trend.The LAI fluctuation of the unclear and upward trends in the 12 years is mainly distributed in the tropical rain forest core area,while the LAI decline area is mainly distributed in the edge of the rain forest,such as the southeast edge of the Amazon rainforest,the southern and northern edges of the African rainforest,and the rainforest north of the Asian rainforest.(2)The correlation between global tropical rainforest vegetation growth(LAI)and average annual temperature from 2001 to 2012 was mainly positive and weak(95.4%of total area),and the overall relationship between LAI and annual precipitation was weakly related(about 80%of the area).(3)From 2001 to 2012,the correlation between LAI and monthly average temperature of tropical rain forest vegetation was mainly positive and weak,and the area of negative correlation was relatively low(about 7.2%of area)and concentrated on the edge of the rain forest,such as the southeast edge of the Amazon rainforest and the northern edge of the rain forest in the African rainforest;The relationship between LAI and monthly precipitation is mainly weak and negative.The monthly correlation between vegetation growth and precipitation has a certain corresponding relationship with the spatial distribution of annual precipitation,namely:generally,high rainfall areas correspond to the weak correlation areas.the medium rainfall areas correspond to the negative correlation areas,and the low precipitation areas correspond to the positive correlation areas.(4)The study on the ecological response of rain forest vegetation growth to the month changes of temperature and precipitation shows that there are 9 different response scenarios in rain forest areas.Based on the distribution center of plot location and the average monthly value of LAI greater than 4,etc.,a total of 11 typical plots were selected,which mainly reflected 5 ecological response scenarios.They are:①LAI is positively correlated with temperature and negatively correlated with precipitation(symbol abbreviated as T+P-);② LAI is weakly correlated with temperature and negatively correlated with precipitation(TOP-);③LAI is weakly correlated with temperature and weakly related with precipitation(TOPO);④LAI is weakly correlated with temperature and positively correlated with precipitation(TOP+);⑤LAI is positively correlated with temperature and negatively correlated with precipitation(T+PO).There are three scenarios such as T+P-,TOPO,and TOP-in the Americas region,and two scenarios such as T+PO and TOP+ in the Africa region,and only TOPO in the Asian region.(5)From 2001 to 2012,the vegetation growth(LAI)of typical tropical rain forests showed a generally gentle fluctuation.The LAI values in the typical plots of the Americas are higher and the inter-annual variability are smaller.The LAI values of typical plots in the African and Asian rainforest regions are relatively lower and the inter-annual variability are larger.The changes of LAI during the year can be divided into two kinds of scenarios:double-peak fluctuations and single-peak fluctuations.The typical sample plots of African rain forests mainly exhibit double-peak fluctuations,and the typical sample plots of the American and Asian rain forests mainly show single-peak small fluctuations.There is a positive correlation between the magnitude of LAI fluctuations and the average monthly temperature with monthly precipitation during the year.The greater the correlation,the greater the LAI fluctuation during the year.(6)The study on the ecological response of LAI and temperature precipitation in typical rainforest plots during 2001-2012 shows that the annual precipitation,low monthly precipitation,and fluctuation direction of temperature and precipitation during the year are the main ecological indicators that affect the monthly correlation of LAI with temperature and precipitation.The performance is:T+P-and TOP-type response scenarios are mainly distributed in the America’s rain forest region where the annual precipitation is moderate(2200-2500mm),the annual precipitation fluctuation is relatively large(39-359mm),and exists 2-3 months relative dry season(3 9-90mm).The plots whose temperature and precipitation fluctuations were basically reversed during the year were mainly T+P-type responses.The plots with basically the same temperature and precipitation fluctuations during the year were mainly TOP-type responses.The T+PO and TOP+ response scenarios are mainly distributed in the African rain forest region where the annual precipitation is relatively low(1500-1900mm),the annual precipitation fluctuation is relatively small(53-238mm),and the temperature and precipitation fluctuations are not exactly in the same direction.The plots with no obvious dry season during the year were mainly T+PO-type response scenarios,and the plots with 2-3 months relative dry season(53-100 mm)were mainly represented as TOP+ response scenarios.TOPO-type response scenarios are mainly distributed in the American and Asian rain forest regions where annual precipitation is high(2700-4600mm),precipitation fluctuation is large(100-600mm),and temperature and precipitation are not in the same direction.(7)Further studies on the correlation between temperature,precipitation,and significant changes in the years of LAI in typical sites of rain forests from 2001 to 2012 show that:In the above T+P-and T0P-type response scenarios,the LAI during the significant LAI changes over the years is negatively correlated with the temperature and cumulative precipitation over the same period.In the two response scenarios of T+P0 and T0P+,the LAI during the significant changes of LAI during the year was weakly positively correlated with temperature,and there was a lagging and cumulative effect of previous precipitation on vegetation growth when there is sufficient rainfall for three consecutive months(accumulated precipitation is 400-600mm),LAI is negatively correlated with precipitation.When there is not enough precipitation for three consecutive months(the cumulative precipitation is about 330mm),LAI is positively correlated with precipitation.The correlation between the significant period of LAI changes during the year of the T0P0-type response scenario and the temperature and cumulative precipitation over the same period is more complex,but overall weaker.(8)The relationship between the monthly change response scenario and the significant phase change result during the year can be divided into 4 categories.The first is consistent with the relationship between temperature and precipitation,corresponding to T0P-type response scenarios;The second is that the temperature-related relationship is inconsistent,and the precipitation-related relationship is consistent,corresponding to the T+P-type response scenario;The third is that the correlations with temperature and precipitation are inconsistent,corresponding to T+P0 type;The fourth is that there is no obvious correlation between temperature and precipitation,which mainly corresponds to the T0P+ and T0P0 response scenarios.更多还原