【作者】 郝翠；

【导师】 李洪远；

【作者基本信息】 南开大学 ， 环境科学， 2012， 博士

【摘要】 2009年12月7日举行的哥本哈根气候大会，将碳减排的研究提到政治高度。虽然目前针对碳减排的最有效的措施是掌握碳捕获与封存技术，将气态形式的碳封存到地下。但是这是一项长期而艰巨的革新，要在全国甚至全球推广需要很长的过程。在这个过程中，我们也需要采取其他的减缓措施，改善由土地利用和管理变化造成的二氧化碳排放就是一个在现阶段来说十分有效的同时经济可行的举措。本文从土壤学、生态学和地理信息系统理论的角度，对滨海新区土地利用现状进行分析及预测，并利用从气候变化角度对由于土地利用变化引起的土壤有机碳累积及释放动态进行分析研究，从点模拟到空间模拟进行了全面的分析验证。首先在数据可获范围内，同时在GIS和RS技术支持下，研究滨海新区自1979年至2010年的土地利用变化，根据滨海新区的土地利用途径将土地类型分为林地、草地、农田、城市绿地、建设用地、水体、裸地等几类；在此基础上根据区域特点在研究土地利用演化规律时采用元胞自动机与Logit多元回归拟合的方法，从空间、时间维度上分析土地利用的动态演化，预测未来滨海新区至2050年的土地利用模式，验证模型采用历史数据及2020和2050土地利用规划数据进行。结果表明：滨海新区土地利用自1979至2010发生了巨大的变化，根据遥感解译数据汇总分析，农田的面积变化在1979-1990年代呈急速上升趋势，在2000-2010年间基本处于恒定状态，在2010年后急速减少，明显看出滨海新区的城市化速度；裸地的变化呈逐年下降趋势，表现出该区域对土地的利用效率逐渐加大；滩涂的变化趋势缓慢，但是一直处于减少的总趋势；林地的总面积较小，在滨海新区内未表现明显增加。模拟结果显示未来滨海新区土地利用的城市化进程仍然较快。为研究土地利用与土壤有机碳之间的关系，本研究于2010年在滨海新区采集了表层0-30cm内的土壤样品，共248个样方，分别测试其土壤有机碳、土壤pH、土壤质地、土壤含盐量、土壤容重等指标，并对各个数据指标的正态性、分散程度、峰度和偏度进行了分析，对非正态性的数据变换后，对各个因子之间的相互关系及土壤有机碳与其他指标间的回归关系作了研究。结果表明：土壤有机碳和土壤容重、土壤粘粒、土壤全盐、土壤pH均呈极显著相关（p=0.01）；而土壤环境之间，土壤全盐与土壤粘粒呈极显著相关，土壤粘粒含量中，盐离子占比例较大；而土壤全盐与土壤pH相关性不大，仅为-0.010，说明土壤粘粒含量中的碱性离子含量不大；此外，土壤容重与土壤粘粒含量和土壤pH呈显著相关关系(p=0.05)，土壤容重的主要来源为土壤粘粒和碱性离子；其他因子之间无明显相关性。土壤环境因子之间的相关关系进一步说明土壤容重、土壤粘粒、土壤全盐、土壤pH之间互相影响，并共同对土壤有机碳的积累产生影响。根据土壤样品的分析测试结果，利用ROTHC及ROTHC-salinity模型原理，将土壤有机碳分为DPM、RPM、HUM、BIO、IOM五部分，每部分的降解速率不同，根据经验降解公式对土地利用变化下的土壤有机碳动态进行模拟，并利用统计学拟合优度LOFIT、误差均值M、相关系数r等参数对模拟数据和未参与模拟的实验数据进行检验，结果表明：滨海新区的土地利用转化方式有草地转林地、草地转农田、草地转城市绿地，草地转建设用地，农田转草地，农田转建设用地，裸地转农田几种类型，ROTHC和ROTHC-salinity的模拟结果都在统计可信范围之内，但是从参数比较上看，ROTHC-salinity模型显然更适合模拟滨海新区的土壤有机碳动态变化，因此在以后的空间分析研究中采用ROTHC-salinity的模拟结果。结合天津市土壤质地图、土地利用分类图及预测图、土壤有机碳模拟数据，建立空间模型，从7848个空间网格内获取每个网格点的数据，将每个点作为一个点模型运行ROTHC-salinity，在1979至2050的时间维度上预测滨海新区土地利用变化导致的土壤有机碳及温室气体排放演化动态。结果表明：1979-2050年的土壤碳排放增加了30-50t/hm~2，平均每年土壤碳排放增加0.42-0.70t/hm~2。更多还原

【Abstract】 The Copenhagen Convention held on December7th in2009mentioned to reducecarbon emission, which raised it to political height. Thought the most effective mea-sures in the present are carbon capture and carbon sequestration technology, with atmo-spheric carbon converted into solid carbon in soil, this is a long and hard innovation,needing a long process to promote in national and global scope. During this process,other reduction measures should be taken to improve carbon loss into the air due to landuse and management, which can be seen as effective and economically viable measuresat present.This paper conducted the research of land use in Binhai New Area from theaspect of soil science, ecological science and geographic information system theory,and an-alyzed the organic soil carbon accumulation and release dynamic because ofland use change from the sight of climate change, then use verification model frompoint sim-ulation to spatial simulation. First of all, within data availability and thesupport of GIS and RS, we did the research of land use change of Binhai New Areafrom1980to2010, according to land use classification standard, land use in BinhaiNew Area can be classified into forest, grass, arable land, urban green land, urbanconstruction land, water body and bared land. Second, the land use evolutionregulation was analyzed by the combination of celluar automata model and multiplelogit regression model according to the regional characteristics from spatial and timedimensions, and then the land use dynamic till2050was predicted and verified by2020and2050land use planning figures. The results showed that land use in BinhaiNew Area has changed a lot from1980to2010, according to the analysis of remotesensing data, the area of arable land has showed a sharp rise from1980to1990, andthen showed a relatively smooth state from2000to2010, and then showed a sharpdecrease after2010; the change of bared land kept decreasing each year, showing anincreased land use efficiency; the change of beach has kept a slow decrease in this timeperiod; the total area of forest is very small in Binhai New Area, and showed littlechange in these years. The whole modeling results showed that the land use change to urban lands in the future will be continued.To study the relationship between land use and soil organic carbon, we collected248soil samples in2010to measure the soil organic carbon in0-30cm soil surface,soil pH, soil texure, soil salt percent and soil bulk density, and analyzed the data nor-mality, dispersion, kurtosis and skewness of each data set. After data transformationof the non-normality data sets, the relationship between each factor was studied andthe regression relationship between soil organic carbon and other factors wasanalyzed. The results showed that the relationship between soil organic carbon, soilbulk density, percent soil clay, percent soil salt and soil pH were significantly high(p=0.01); while due to the strong correlation between soil clay and soil salt, we canconclude that there are high percent salt ions in soil clay. The correlation betweensoil salt and soil pH showed a weak value-0.01, meaning that alkaline ions of the soilbulk density were not high. The correlation between soil bulk density and soil clay,soil pH showed a high value, meaning that the high bulk density was due to alkalineions and percent clay. The correlations between soil environmental factors showedthe impact of these factors and the impact to soil organic carbon accumulation.According to the results of soil samples and the theory of ROTHC andROTHC-salinity, the soil organic carbon is split into5parts, DPM, RPM, HUM, BIOand IOM, each of which has a different soil decomposition rate. The soil organiccarbon dynamic under different land use changes were modeled based on empiricaldecomposition for-mula, and the model results were verified by statistical indicatorsLOFIT, M and r. The results showed that there were several land use changes inBinhai New Area, namely the conversion from grass to forest, grass to arable, grass tourban green, grass to urban land, arable to grass, arable to urban land, bare to arable.Both results of ROTHC and ROTHC-salinity models can be seen as true statistically,however, the results of Rothc-salinity are more compatible to simulate the soil organiccarbon dynamics in Binhai New Area.Combined with soil texure figures, land use classification figures and the modeleddata sets of soil organic carbon, the spatial model to run the whole area was built topredict the soil organic carbon dynamics and green house gas emissions from1979to2050, with running ROTHC-salinity model7848times. The results showed that between1979and2050the soil carbon emission will reach to30-50t/hm~2, at a rate of0.42-0.70t/hm~2each year.更多还原