【作者】 任正超；

【导师】 柳小妮； 朱华忠；

【作者基本信息】 甘肃农业大学 ， 草业科学， 2010， 硕士

【摘要】 植被作为地表碳循环重要组分,其净初级生产力(NPP)不仅直接反映自然环境条件下植被群落生产能力,表征陆地生态系统质量状况;而且也是判定生态系统碳源、碳汇以及调节生态过程的主要因子。近20多年来,随温室效应等气候与环境问题加剧,联合国气候变化框架公约(UNFCCC)外交谈判中,对碳循环需要提供更充分的科学依据。因此,了解我国及周边地区陆地生态系统碳收支时空格局及其变化趋势,具有重要科学与政治意义。以TM和MODIS NDVI为遥感数据源,辅以地面气象观测和其它本底数据,综合利用遥感技术(RS)、地理信息系统技术(GIS)和全球定位系统技术(GPS),并应用改进后CASA模型,模拟估算俄罗斯布里亚特共和国地区2000-2008年的植被NPP年际、季节和月份变化动态。同时,以实测NPP数据验证模型适用性和精度,分析其对气候因子的响应方式和反馈机制,初步揭示影响NPP变化的气候驱动机制。1 CASA模型的改进CASA模型分为光合有效辐射、光能利用率和土壤含水量3个子模型。由于CASA模型土壤含水量子模型较复杂,数据获取有一定难度;所以对其先进行简化和优化,将原模型中表示土壤水分的蒸发潜力,即相对干燥率(RDR)和可能蒸散量(PET),通过生物温度来计算。通过NPP实测数据、MODIS NPP数据产品和其它模型的模拟结果对比分析,发现改进后的CASA模型,其模拟值分布于趋势线附近,均值接近,分别为323.69 gC·m-2和355.68 gC·m-2,平均相对误差4.94%。实测值和模拟值间的相关系数R为0.88(P<0.01),说明,改进后的CASA模型模拟精度较高,可运用于布里亚特共和国地区植被NPP估算。2 NDVI和EVI对比分析植被NDVI最大和最小值分别出现在伊沃尔金斯基和塔尔布加泰斯基地区,而EVI最大和最小值则出现在塔尔布加泰斯基和伊沃尔金斯基地区。不同地区最大与最小NDVI均值的植被类型不同在伊沃尔金斯基和穆哈尔什比尔斯基地区,森林植被最大,稀树草原最小;在吉丁斯基地区,草原与森林混合植被最大,高山植被最小;在塔尔布加泰斯基地区,森林最大,草原与森林混合植被最小,而恰赫金斯基地区则与之相反;在比丘尔斯基地区,森林>稀树草原;在色楞津斯基地区沼泽、草地最大,高山植被最小。不同地区最大与最小EVI均值的植被类型不同在伊沃尔金斯基地区,草原与森林混合植被最大,森林最小;在吉丁斯基地区,沼泽和草地最大,草原与森林混合植被最小;在塔尔布加泰斯基地区,草原与森林混合植被最大,稀树草原最小;在恰赫金斯基地区,森林植被最大,草原与森林混合植被最小;在比丘尔斯基地区,稀树草原>森林;在穆哈尔什比尔斯基地区,稀树草原最大,森林最小;在色楞津斯基地区,高山植被最大,沼泽和草地最小。整个南部地区森林植被的NDVI和EVI值最大,沼泽和草地最小。NDVI的最大值、最小值、均值和标准差变化范围均明显>EVI。NDVI能客观地区分不同类型植被,利于植被类型遥感解译和定量分析。3植被NPP时空分布格局年际变化2000-2008年,布里亚特共和国植被NPP总体为波动中呈现上升趋势,均值为544.29gC·m-2·a-1,总量为1.91E+14gC·a-1,平均增幅为0.39gC·m-2·a-1。2001年和2003年为植被NPP低值期,2003年达最小值,为345.94gC·m-2·a-1。2003年以后,植被NPP呈上升趋势,至2008年达最大值,为668.76gC·m-2·a-1;且2003-2004年、2007-2008年增长幅度大,其余年份增幅平缓。月际和季节变化2000-2008年,布里亚特共和国植被NPP的月变化为,1-3月,单位面积内植被NPP约为0,最小值出现在2月,为0.002gC·m-2·month-1;自4月开始,其植被NPP急速增长,至7月达峰值,为131.13gC·m-2·month-1;随后急骤下降,11-12月降至0左右。生长季(4-10月)的NPP均值总量为537.37gC·m-2。春(3-5月)、夏、秋和冬季的NPP总量分别为81.83 gC·m-2、365.73 gC·m-2、94.16 gC·m-2和0.73gC·m-2,分别占全年的15.08%、67.42%、17.36%和0.14%。区域变化23个辖区中,乌兰乌德市、奥金斯基地区的植被NPP值在年际和月际水平上都较低;而伊沃尔金斯基、普里贝加尔斯基、扎卡缅斯基地区和比丘尔斯基地区则较高。经纬向变化无论年际还是月际水平上,在经度上,植被NPP均表现为双峰分布格局,总体表现为随经度递增而增大规律;在纬度上,植被NPP表现为单峰分布格局,总体表现为随纬度递增而减小规律。空间变化2000-2008年,布里亚特共和国植被NPP增加的区域主要分布在北部及西部地区。由西南向东北,其NPP表现出增加、变化平缓和增加的变化趋势。绝大部分地区的植被NPP变化不显著(p>0.05),占植被总面积的88.95%;而变化显著地区(p<0.05)仅占总面积的11.05%。75.05%的植被NPP呈增加趋势,其中显著(p<0.05)和极显著(p<0.01)增加面积占总面积的9.77%;24.95%的植被NPP呈降低趋势,其中极显著(p<0.01)降低的面积仅占总面积的0.23%。植被类型变化年际水平上,不同植被类型NPP在2000-2001年和2002-2003年为下降趋势,而2001-2002年和2003-2008年为上升趋势;与2000-2008年所有植被NPP的年际变化规律一致。月际水平上,不同植被类型NPP积累均集中于生长季(4-10月),11-3月不同植被类型NPP都保持在0左右。4-7月和8-10月分别为植被NPP积累增长期和递减期,且递增和递减速率较大。4植被NPP与气候因子相互关系分析植被NPP与气候因子相关性通过对该区植被NPP与主要气候因子的简单相关性和偏相关性分析,得知,年际水平上,植被NPP与主要气候因子均无呈显著相关性(p>0.05);但月份水平上,其相关性均呈极显著水平(p<0.01)。温度和降水量对植被NPP的空间响应布里亚特共和国地区西部、南部小面积地区和贝加尔湖沿岸地区以及北部大面积地区的植被NPP均与温度呈显著正相关(p<0.05),西部和北部大面积地区均与降水量呈显著正相关(p<0.05),而中部地区与降水量呈显著或极显著负相关(p<0.05或p<0.01)。说明,西部和北部地区的植被生长主要受温度和降水共同影响,南部和贝加尔湖沿岸地区的则主要受温度影响。在综合运用遥感数据、气象数据、数学模型的基础上,对2000-2008年俄罗斯布里亚特共和国地区植被NPP进行时空变化模拟,并与气候因子进行了相关性分析。同处西伯利亚冷高压气候循环系统的布里亚特共和国地区植被NPP与中国北方地区的植被NPP在时空分布格局上有着许多相似之处。本研究改进的CASA模型可以运用于中国北方地区的植被NPP模拟估算以及本研究成果对于中国北方地区植被NPP的模型估算和生态跨境研究具有重要的借鉴意义。更多还原

【Abstract】 Vegetation Net Primary Productivity (NPP) is a key component of the terrestrial carbon cycle. As the direct reflection of plant community productivity for a certain natural environment, it is the basis of matter and energy cycle of terrestrial ecosystem. As highlighted during the international negotiation process for the United Nations Framework Convention on Climate Change (UNFCCC), a better grasp upon the controls and distribution of NPP is pivotal for sustainable human use of the biosphere. Based on Remote Sensing, Geographic Information System and Gobal Positioning System, This paper comprehensively used remote sensing data, ground meteorological data, other additional data and improved Carnegie Ames Stanford Approach (CASA) model to estimate vegetation NPP in Buryatiya Republic, Russia from 2000 to 2008. After the comparison and validation with observed data and other NPP product data, the NPP time-series of Buryatiya terrestrial vegetation from 2000 to 2008 was built. Spatio-temporal variations and potential trend of NPP were analyzed in these 9 years, and the relationship between NPP and global climatic change was comprehensively studied. From these researches, some basic conclusions were drawn as follows:1. Improvement of CASA modelCASA model contains three submodels of Photosynthetically Active Radiation, Light Use Efficiency and Soil Water Content, but the parameters of Soil Water Content model are complex. So, there is some difficulty to obtain the reaserch data. This paper simplifys the estimation model via inputing Bio-temperature to model in order to calculate potential evaporation of soil water. With the comparison and validation with observed data and other NPP product data, the result shows: Average value of observed data is 323.69gC·m-2 and estimation data is 355.68gC·m-2. The difference between them is small and the average relative error is 4.94%. Correlation coefficient between observed data and estimation data is 0.88(p<0.01), which prove the improved CASA model can be used to estimate the vegetation NPP in Buryatiya Republic. Precise of improved CASA model is high.2. Comparison and analysis between NDVI and EVIMaximal average value of vegetation NDVI presents in Ivolginskii region, but minimal average value of NDVI in Tarbagataiskii region. Maximal average value of vegetation EVI presents in Tarbagataiskii region, but minimal average value of NDVI in Ivolginskii region. It is very interesting that the opposite phenomenon is found.Vegetation with maximal average NDVI value in Ivolginskii region is forest, but steppe has the minimal NDVI value in this region. In Dzhidinskii region, mixed vegetation of grassland and forest has the maximal average NDVI value, but high mountain vegetation with the minimal NDVI value. In Tarbagataiskii region, vegetation with maximal average NDVI value is forest, and mixed vegetation of grassland and forest has the minimal average NDVI value. In Kyahtinskii region, the phenomenon is opposite to Tarbagataiskii region. Forest has the bigger average NDVI value than steppe in Bichurskii region. In Muhorshibirskii region, forest has the maximal average NDVI value, and steppe with the minimal average NDVI value. Mixed vegetation of swampe and meadow has the maximal average NDVI value, and high mountain vegetation has the minimal average NDVI value in Selenginskii region.Vegetation with maximal average EVI value in Ivolginskii region is mixed vegetation of forest and grassland, but forest has the minimal EVI value in this region. In Dzhidinskii region, mixed vegetation of swampe and meadow has the maximal average EVI value, but mixed vegetation of forest and grassland with the minimal EVI value. In Tarbagataiskii region, vegetation with maximal average EVI value is mixed vegetation of forest and grassland, and steppe has the minimal average EVI value. In Kyahtinskii region, forest has the maximal average EVI value, and mixed vegetation of forest and grassland has the minimal average EVI value. Steppe has the bigger average EVI value than forest in Bichurskii region. In Muhorshibirskii region, steppe has the maximal average EVI value, and forest with the minimal average EVI value. Mixed vegetation of swampe and meadow has the minimal average EVI value, and high mountain vegetation has the maximal average EVI value in Selenginskii region.Forest has the maximal average value of NDVI and EVI, and mixed vegetation of swampe and meadow has the minimal average value of NDVI and EVI in total southern region of Buryatiya. Fluctuation range of maximal value, minimal value, average value and standard deviation of NDVI is obvious bigger than EVI, which can show that NDVI has the superior capacity to distinguish vegetation types on remote sensings objectively and is propitious to remote sensing interpretation and quantitative analysis in future.3. Spatio-temporal distribution pattern of vegetation NPP(1) Annual variation of vegetation NPP: Average value of vegetation NPP in Buryatiya Republic from 2000 to 2008 is 544.29gC·m-2·a-1, and total NPP is 1.91E+14gC·a-1. The trend of vegetation NPP in Buryatiya Republic from 2000 to 2008 is increasing among fluctuation as a whole. The value of vegetation NPP locates lowest point in 2003 with 345.94gC·m-2·a-1, but locates wave crest in 2008 with 668.76gC·m-2·a-1. The increase range is great from 2003 to 2004, 2007 to 2008, but it increase gently in other years. The average increase range of vegetation NPP is 0.39gC·m-2·a-1 in this area from 2000 to 2008.(2) Monthly variation of vegetation NPP: The trend of monthly variation shows: vegetation NPP has the smallest value with 0.002gC·m-2·month-1 per unit area in February, but it has the biggest value with 131.13gC·m-2·month-1 per unit area in July. From January to March, vegetation NPP locates zero per unit area, but it increases rapidly in April and reaches highest point in July. With the climate change, vegetation NPP decreases sharpely from Augest to October, and locates zero per unit area in December.(3) Seasonal variation of vegetation NPP: Vegetation NPP in Spring(March to May), Summer(June to Augest), autumn(September to November) and Winter(December to February next year) are 81.83 gC·m-2, 365.73 gC·m-2, 94.16 gC·m-2 and 0.73 gC·m-2, with proportion of 15.08%, 67.42%, 17.36% and 0.14% of total NPP in all year per unit area. (4) Regional variation of vegetation NPP: Among all the twenty-three regions of Buryatiya Republic, vegetation NPP has lower value in Ulan-Ude city, Okinskiy region on both yearly and monthly level, but it has higher value in Ivolginskii region, Severobaykalsk region, Zakamensk p. region and Bichurskii region.(5) Longitudinal and latitudinal variation of vegetation NPP: Both yearly and monthly level, vegetation NPP in Buryatiya Republic shows“double-humped”distribution pattern on longitudinal scale, and“Singlet”distribution pattern on latitudinal scale. It presents increasing rule fllowing raised longitude and decreasing rule with increased latitude.(6) Spatial variation of vegetation NPP: The regions of increased vegetation NPP locate in northern and western areas in Buryatiya Republic from 2000 to 2008. From southwest to northeast, vegetation represents the trend of increased distinctively, vary indistinctively and increased distinctively. Vegetation NPP changes indistinctively (p>0.05) in most area with proportion of 88.95%, only in 11.05% area, the vegetation NPP changes distinctively (p<0.05). Vegetation NPP shows increased trend with proportion of 75.05%, thereinto, the proportion of increased distinctively (p<0.05) and increased very distinctively (p<0.01) is 9.77%. Vegetation NPP shows decreased trend with proportion of 24.95%, thereinto, the proportion of decreased very distinctively (p<0.01) is 0.23%.(7) NPP Variation of different vegetation: On yearly level, different vegetations NPP present decreased trend from 2000 to 2001, 2002 to 2003, but increased trend from 2001 to 2002, 2003 to 2008. It is similar with the annual variation rule of vegetation NPP from 2000 to 2008. On monthly level, NPP accumulation of different vegetations arise in growing season, but growth arrest from November to March next year. Vegetation NPP accumulation increasing period is April to July, and decreasing period is Augest to October. The rates of increasing and decreasing are great.4. Correlation research between vegetation NPP and climatic factors(1) Correlation analysis between vegetation NPP and climatic factors: with the simple correlation and partial correlation analysis between vegetation NPP and climatic factors in research area, result shows that: on yearly level, there is non-significant correlation (p>0.05) between vegetation NPP and climatic, but it is opposite to (p<0.01) on monthly level.(2) Spatial response of temperature and precipitation to vegetation NPP: In small area of western, southern Buryatiya, baikal shore and large area of northern Buryatiya, vegetation NPP and temperature have significant positive correlation (p<0.05), but vegetation NPP and precipitation have significant positive correlation (p<0.05) in large area of western and northern Buryatiya and significant negative correlation (p<0.05) even very significant negative correlation (p<0.01). It shows that temperature in company with precipitation impact on vegetation NPP in western and northern Buryatiya, but it is mainly controlled by temperature in southern Buryatiya and baikal shore.Based on Remote Sensing data, meteorological data and mathematical model, temporal-spatial variation simulation of vegetation NPP and its correlation with climatic factors were applied. Vegetation between Buryatiya Republic and Northern China has a lot of similarities, because they locate in cold and high-pressure climate circulation system in Siberia. The improved CASA model can be used in vegetation NPP estimation of Northern China and research results have reference significance for ecological cross-border study.更多还原