EGI OpenIR  > 研究系统  > 空间对地观测与系统模拟研究室
Thesis Advisor李龙辉
Degree Grantor中国科学院大学
Place of Conferral新疆乌鲁木齐
Degree Discipline理学硕士
Keyword长白山 数据同化 叶面积指数 碳通量 蒸散发 Changbai Mountains data assimilation leaf area index carbon flux evapotranspiration
AbstractThe forest of Northeast China is located in the middle and high latitudes,accounting for more than 30% of the total forest area in China. Broad-leaved Koreanpine forest is the zonal forest vegetation in the eastern part of Northeast China, andthe trend of climate changed warming and dry will cause the eco-climatic suitabilityof broad-leaved Korean pine forest to decline significantly. The broad-leaved Koreanpine forest in Changbai Mountain has continuous and long-term observation databased on eddy covariance technique, which provides data support for the validation ofthe model. Model simulation and site observation provide a way to study the carbonand water fluxes in the broad-leaved Korean pine forest ecosystem.The Biome-BGC model is a widely used biogeochemical cycle model tosimulate the carbon, nitrogen and water storage of vegetation, litter and soil indifferent scales. Data assimilation provides an effective way to integrate the modelsimulation and remote sensing observation, through the integration of remote sensingdata in the run of the model, adjusting the model trajectory to reduce model error andimprove simulation accuracy. Ensemble Kalman filter (EnKF) is a sequentialAssimilation Algorithm for estimating the prediction error covariance using a MonteCarlo ensemble prediction method. This paper uses the EnKF assimilated MODIS leafarea index (LAI) into the Biome-BGC model in growing season to simulate the waterand carbon fluxes in the broad-leaved Korean pine forest in Changbai Mountains. Atthe same time, the simulated snow sublimation and the parameters of the calculationmethod of soil temperature are improved, which can effectively reduce the error of theecological respiration in winter. Based on the measured data of eddy covariance, theeffects of data assimilation and model improvement on the simulation precision ofwater carbon flux in the broad-leaved Korean pine forest in Changbai Mountain wereanalyzed. The results are as follows:(1) As compared with the original model simulated without data assimilation, theimproved Biome-BGC model with the assimilation of the MODIS LAI makes the correlation coefficient between the simulated values and the observed values of thegross ecosystem primary productivity (GPP) increased by 0.06, and reduced thecentered root-mean-square error (RMSE) by 0.48 gC•m-2•d-1, ecosystem respiration(RE) correlation coefficient increased by 0.02, centered root-mean-square error isreduced by 0.20 gC•m-2•d-1; the correlation coefficient of net ecosystem exchange ofcarbon (NEE) is increased by 0.35, centered root-mean-square error decreased by0.50gC•m-2•d-1.(2) The correlation coefficient between the simulated values and the observedvalues of evapotranspiration (ET) increased by 0.05, and the data assimilation andmodel improvement have no significant effect on the central root mean square error ofthe simulated and observed ET.(3) The LAI of the Biome-BGC model seriously deviates from the normal LAIvariation, the trend of MODIS LAI is similar to that of real LAI, which lead to theLAI of assimilation is close to MODIS LAI.The data assimilation based on EnKF algorithm improves the accuracy of thecarbon flux simulation in the broad-leaved Korean pine forest in Changbai Mountains,and has an important influence on the more accurate estimation of the carbon flux inthe regional scale.
Other Abstract东北森林地处我国中高纬度地区,占全国森林总面积超过 30%。阔叶红松林是我国东北东部的地带性顶级森林植被。暖干的气候变化趋势会导致阔叶红松林生态适应性显著下降。长白山阔叶红松林有着连续的、长期的基于涡度相关技术的观测资料,为模型的验证提供了数据支持。模型模拟与站点观测为我们提供了研究阔叶红松林生态系统水碳通量的途径。Biome-BGC 模型是模拟不同尺度植被、凋落物、土壤中碳、氮、水的储量和通量的一个被广泛应用的生物地球化学循环模型。数据同化为模型与遥感观测结合提供了一条有效的途径,通过在模型运行过程中融入遥感观测数据,调整模型运行轨迹从而降低模型误差,提高模拟精度。集合卡尔曼滤波算法(EnKF)是一种用蒙特卡罗的集合预报方法估计预报误差协方差的顺序同化算法。本文利用 EnKF 同化生长季中分辨率成像光谱仪(MODIS)叶面积指数(LAI)与Biome-BGC 模型模拟的 LAI 模拟长白山阔叶红松林的水碳通量。同时,通过改进模拟的雪面升华与土壤温度计算方法的参数,旨在降低冬季生态呼吸的模拟误差。结合涡度相关实测数据,分析数据同化与模型改进对模型模拟长白山阔叶红松林水碳通量模拟精度的影响。结果表明:(1)相对于原始模型,数据同化与模型改进后使得生态系统总初级生产力(GPP)的模拟值与观测值之间的相关系数提高 0.06,中心化均方根误差(RMSE)降低 0.48 gC•m-2•d-1;生态系统呼吸(RE)的相关系数提高 0.02,中心化均方根误差降低 0.20 gC•m-2•d-1;净生态系统碳交换量(NEE)相关系数提高 0.35,中心化均方根误差降低 0.50gC•m-2•d-1。(2)模型改进后模拟与观测的蒸散发(ET)的相关系数提高了 0.05,数据同化与模型改进对模拟与观测的 ET 的中心化均方根误差没有显著影响。(3)Biome-BGC 模型模拟的 LAI 与真实 LAI 的变化趋势不同,而 MODISLAI 变化趋势与真实 LAI 变化趋势相近,致使同化值趋近于 MODIS LAI。基于 EnKF 算法的数据同化提高了长白山阔叶红松林碳通量模拟精度,对于更加精确地估算区域碳通量有着重要的影响。
Subject Area地图学与地理信息系统
Document Type学位论文
First Author Affilication中国科学院新疆生态与地理研究所
Recommended Citation
GB/T 7714
郑磊. 长白山阔叶红松林生态系统水碳通量模拟[D]. 新疆乌鲁木齐. 中国科学院大学,2017.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[郑磊]'s Articles
Baidu academic
Similar articles in Baidu academic
[郑磊]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[郑磊]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.