遥感GPP模型在高寒草甸的应用比较

随着遥感数据时空分辨率的提高,大范围实时监测总初级生产力GPP(Gross Primary Productivity)的变化成为可能。本研究收集了黑河流域阿柔冻融观测站的气象观测资料和MODIS数据,驱动VPM、TG、VI和EC-LUE4个模型估算了该站点的GPP,并应用涡动相关观测的GPP验证了模拟结果,并比较了这4个模型的模拟精度。结果表明:阿柔站2009年的涡动相关观测的GPP、NEE(Net Ecosystem Exchange)和ER(Ecosystem Respiration)分别为:804.2gC/m2/yr、129.6gC/m2/yr和673.6gC/m2/yr。该站点光合作用固定的碳有83.8%通过生态系统的呼吸作用释放到大气中。基于遥感的GPP模型能够很好地模拟高寒草甸的GPP,全年的判定系数在0.94以上,生长季的判定系数大于0.84。     

陆地总初级生产力遥感估算精度分析

准确估算陆地总初级生产力GPP(Gross Primary Productivity)数值对碳循环过程模拟有重要影响。本文介绍了多种基于植被指数以及基于光能利用率的遥感GPP算法,综述了不同算法在其研究区域的估算精度;并分析了MODIS/GPP以及BESS/GPP两种遥感GPP产品在不同植被类型的估算精度。通过对比全球碳通量站网络GPP数据表明,MODIS/GPP产品在全球估算结果具显著相关性(R2=0.59)及中等标准误差(RMSE=2.86 g C/m2/day),估算精度较高的植被类型有落叶阔叶林,草地等;估算精度较低类型包括常绿阔叶林,稀树草原等。本文对GPP产品中存在的不确定性进行分析,通过综述前人研究中发现的遥感估算GPP方法中存在的问题,指出可能的提高卫星遥感GPP产品估算精度的方法及发展趋势。     

叶绿素吸收的光合有效辐射比率的遥感估算模型研究

光合有效辐射比率（FPAR）是陆地生态系统碳循环研究的重要参数。根据FPAR的物理意义，本文提出了有效FPAR的概念：即叶绿素吸收的FPAR（FPARchl），利用模拟数据建立了有效FPAR信息模型。探讨了基于叶绿素含量估算有效FPAR的可行性，为进一步采用有效FPAR来估算NPP／GPP提供思路和途径。利用实测数据对有效FPAR模型进行验证表明：（1）叶片吸收的FPAR（FPARleaf）与有效FPAR（FPARchl）之间的差异较大。FPARchl通常不足FPARleaf的50％，两者之间虽存在一定的相关性，而这种相关性是非线性的；（2）叶绿素含量与FPARleaf的相关性较高，而与FPARchl的相关性更高，基于叶绿素含量可以以较高的精度估算FPARleaf和FPARchl经过验证，估算的平均相对误差分别为2．90％和6．6％；（3）应用模拟数据建立的FPARleaf及有效FPAR模型均在叶绿素含量大于20μg／cm。时估算的精度较高，而低于20μg／cm^2时估算的精度较低，本文中叶绿素含量大于20μg／cm^2的60个样本的FPARleaf及FPARchl的估算的平均相对误差，分别为1．79％和5．07％。     

机器学习算法在森林地上生物量估算中的应用

森林地上生物量是森林生产力的重要评价指标,对其进行高效监测对维持全球碳平衡和保护生态系统具有重要意义。本文首先基于冠层高度模型数据,通过分水岭分割算法得到单木冠幅边界;然后在单木冠幅范围内提取23个LiDAR变量,结合佩诺布斯科特试验森林的87组实测数据,利用随机森林和支持向量机建立森林地上生物量估算模型;最后对样地模型估算的结果进行了比较,讨论了预测结果及其精度。结果表明:本文选用的随机森林模型和支持向量机模型在估算森林地上生物量的应用中获得了较高的精度;并且,随机森林模型在基于机载雷达数据估测森林地上生物量中的估算精度更高,模型泛化能力更强,制图精度也更好,具有更好的适用性。     

基于IDL语言的植被指数UNVI软件插件

植被指数是地球陆表植被覆盖度和植被活力的指示因子,对环境监测、植被理化参量估算等应用研究有重要的意义。基于植被的反射光谱特征,通过遥感数据波段的组合,可以计算得到遥感植被指数。传统的植被指数如NDVI、EVI等仅利用有限波段信息的线性或非线性组合构建而成,没有充分利用遥感传感器所提供的多波段遥感信息,通用归一化植被指数UNVI(Universal Normalized Vegetation Index)充分利用了遥感传感器提供的多波段植被光谱信息,因此在反演植被叶绿素、生物量等植被理化参量上较其他传统植被指数更具优势。为方便UNVI指数的计算,本文基于IDL语言开发了UNVI软件插件,可直接作为ENVI商业遥感软件进行调用,并可满足多个传感器的UNVI计算需求。为了验证UNVI的应用效果,以植被总初级生产力GPP(Gross Primary Productivity)估算为例,比较了不同植被指数估算GPP的效果,结果表明:基于UNVI估算的GPP与通量站点获得的GPP具有较高的相关性(相关系数R2为0.79),验证了UNVI在GPP估算方面的优势。本文提供的UNVI软件插件可为遥感研究和应用人员提供便捷的计算工具。     

基于遥感的区域尺度森林地上生物量估算研究

森林是陆地生态系统最大的碳库,精确估算森林生物量是陆地碳循环研究的关键。首先从机载LiDAR数据中提取高度和密度统计量,采用逐步回归模型进行典型样区生物量估算;然后利用机载LiDAR数据估算的生物量作为样本数据,与多光谱遥感数据Landsat8 OLI的波段反射率及植被指数建立回归模型,实现区域尺度森林地上生物量估算。实验结果显示,机载LiDAR数据估算的鼎湖山样区生物量与地面实测生物量的相关性R2达0.81,生物量RMSE为40.85 t/ha,说明机载LiDAR点云数据的高度和密度统计量与生物量存在较高的相关性。以机载LiDAR数据估算的生物量为样本数据,结合多光谱遥感数据Landsat8 OLI估算粤西北地区的森林地上生物量,精度验证结果为:R2为0.58,RMSE为36.9 t/ha;针叶林、阔叶林和针阔叶混交林等3种不同森林类型生物量的估算结果为:R2分别为0.51(n=251)、0.58(n=235)和0.56(n=241),生物量RMSE分别为24.1 t/ha、31.3 t/ha和29.9 t/ha,估算精度相差不大。总体上看,利用遥感数据可以开展区域尺度的森林地上生物量估算,为森林固碳监测提供有力的参考数据。     

随机森林回归模型的悬浮泥沙浓度遥感估算

三峡工程蓄水以来,清水下泄,坝下游河段发生了长时间、长距离的沿程冲刷,河流悬浮泥沙浓度发生改变,给沿岸生态系统带来了不利影响。随机森林算法灵活、稳健,已被广泛应用于各类生态环境变量的回归预测分析,但其在水体悬浮泥沙浓度估算方面的能力尚未得到充分认识。基于泥沙站点监测数据和MODIS卫星遥感反射率数据,通过构建随机森林非参数回归预测模型,对三峡工程坝下游宜昌至城陵矶河段在建坝前后14年间(2002年—2015年)各月的悬浮泥沙浓度进行遥感估算。研究表明:(1)基于随机森林的悬浮泥沙浓度估算模型表现较好,模型预测值与实测值间相关性好、预测精度高,优于其他模型(线性回归、支持向量机、人工神经网络模型)。(2)在参与模型构建的MODIS波段变量中,红波段被认为是最重要的预测变量,但不能单独使用它进行预测,悬浮泥沙遥感预测需要多变量共同参与。(3)将悬浮泥沙数据按季节分类所构建的随机森林模型,其平均误差为0.46 mg/L,平均相对均方根误差为12.33%,估算效果最优,能够满足较高精度下悬浮泥沙浓度估算的需求。综上,可以考虑以季节为划分依据,用随机森林回归模型估算悬浮泥沙浓度,并用于后期坝下游河道悬浮泥沙浓度时空反演。     

基于遥感的某市城市圈森林碳储量估算与空间分布特征

森林生态系统碳储量占有整个陆地生态系统碳储量约50%,利用遥感数据进行森林碳储量估算对加快实现“碳达峰”和“碳中和”具有重要意义。本研究利用Landsat 8 OLI遥感影像和DEM数据提取植被指数和地形因子,转换净初级生产力数据为生物量数据,并利用多元逐步回归分析法建立武汉城市圈森林植被碳储量遥感估算模型。根据统计数据和估算模型得出,武汉城市圈碳储量空间分布表现为东北部和南部山脉区域的碳储量和碳密度较高,而中东部武汉市和黄石市中心区域相对较低,且植被碳密度主要集中在中海拔地区。     

三峡库区森林叶面积指数多模型遥感估算

叶面积指数(leaf area index,LAI)是定量研究森林生态系统能量交换的一个重要结构参数。本文利用野外观测LAI,以及Landsat TM计算的7种常用植被指数和5个自定义植被指数,通过筛选建立了不同森林类型的LAI估算模型,其中,针叶林采用多元逐步回归模型,阔叶林与混交林采用主成分分析模型,最终通过多个模型估算三峡库区区域尺度森林LAI。利用样地实测LAI数据进行精度验证,针叶林、阔叶林和混交林的均方根误差分别为0. 829 4,1. 111 5和1. 790 9,判定系数R2均达到了0. 77以上。研究结果将为森林生态系统和碳循环研究提供基础数据。     

基于森林模型参数先验知识估算高分辨率叶面积指数

目前，估算高分辨率叶面积指数LAI（Leaf Area Index）的常用方法是采用大量地面测量数据和遥感数据建立统计模型，再用统计模型估算LAI。然而，与农田地面测量实验相比，森林地面测量实验获取的观测数据更加有限，这使得基于统计模型的森林高分辨率LAI的估算精度低，难以满足应用需求。为此，本文提出一种基于森林模型参数先验知识、使用森林研究区少量的LAI地面测量数据和归一化植被指数NDVI数据估算森林高分辨率LAI的方法。首先，获取全球20个森林实验区的LAI地面测量数据和NDVI数据，建立LAI-NDVI统计模型并提取森林模型参数的先验知识。然后，以一个新的森林站点Concepción作为研究区，将该研究区的数据分为建模数据和验证数据两个部分。使用研究区有限的建模数据对森林模型参数先验知识进行本地化校正得到优化模型，优化模型用于估算森林高分辨率LAI，使用验证数据评价LAI的估算精度。同时，选取了Camerons站点、Gnangara站点、Hirsikangas站点评价本文方法的LAI估算精度。使用地面测量LAI验证基于森林模型参数先验知识估算高分辨率LAI的结果精度，经验证4个森林站点的均方根误差分别为0.6680，0.4449，0.2863，0.5755。研究结果表明：在仅有少量观测数据时，采用本方法能有效地提高森林高分辨率LAI的估算精度。因此，本方法可为森林高分辨率LAI的遥感估算提供参考。     

Large-scale estimates of gross primary production on the Qinghai-Tibet plateau based on remote sensing data

Vegetation gross primary production (GPP) is an important variable for the carbon cycle on the Qinghai-Tibetan Plateau (QTP). Based on the measurements from 12 eddy covariance flux sites, we validated a light use efficiency model (i.e. EC-LUE) to evaluate the spatial-temporal patterns of GPP and the effect of environmental variables on QTP. In general, EC-LUE model performed well in predicting GPP at different time scale over QTP. Annual GPP over the entire QTP ranged from 575 to 703 Tg C, and showed a significantly increasing trend from 1982 to 2013. However, there were large spatial heterogeneities in long-term trends of GPP. Throughout the entire QTP, air temperature increase had a greater influence than solar radiation and precipitation (PREC) changes on productivity. Moreover, our results highlight the large uncertainties of previous GPP estimates due to insufficient parameterization and validations. When compared with GPP estimates of the EC-LUE model, most Coupled Model Intercomparison Project (CMIP5) GPP products overestimate the magnitude and increasing trends of regional GPP, which potentially impact the feedback of ecosystems to regional climate changes.     

The potential of the greenness and radiation (GR) model to interpret 8-day gross primary production of vegetation

Remote sensing of vegetation gross primary production (GPP) is an important step to analyze terrestrial carbon (C) cycles in response to changing climate. The availability of global networks of C flux measurements provides a valuable opportunity to develop remote sensing based GPP algorithms and test their performances across diverse regions and plant functional types (PFTs). Using 70 global C flux measurements including 24 non-forest (NF), 17 deciduous forest (DF) and 29 evergreen forest (EF), we present the evaluation of an upscaled remote sensing based greenness and radiation (GR) model for GPP estimation. This model is developed using enhanced vegetation index (EVI) and land surface temperature (LST) from the Moderate Resolution Imaging Spectroradiometer (MODIS) and global course resolution radiation data from the National Center for Environmental Prediction (NCEP). Model calibration was achieved using statistical parameters of both EVI and LST fitted for different PFTs. Our results indicate that compared to the standard MODIS GPP product, the calibrated GR model improved the GPP accuracy by reducing the root mean square errors (RMSE) by 16%, 30% and 11% for the NF, DF and EF sites, respectively. The standard MODIS and GR model intercomparisons at individual sites for GPP estimation also showed that GR model performs better in terms of model accuracy and stability. This evaluation demonstrates the potential use of the GR model in capturing short-term GPP variations in areas lacking ground measurements for most of vegetated ecosystems globally.     

Diverse responses of vegetation production to interannual summer drought in North America

Droughts are projected to occur more frequently with future climate change of rising temperature and low precipitation. However, its impact on regional and global vegetation production is not well understood, which in turn contributes to uncertainties to model carbon sequestration under drought scenarios. Using long-term continuous eddy covariance measurements (168 site-year), we present an analysis of the influences of interannual summer drought on vegetation production across 29 sites representing diverse ecoregions and plant functional types in North America. Results showed that interannual summer drought, which was evaluated by the increase in summer temperature or decrease in soil moisture, would cause reductions of both summer gross primary production (GPP) and net ecosystem production (NEP) in non-forest sites (e.g., grasslands and crops). On the contrary, forest ecosystems presented a very different pattern. For evergreen forests, lower summer soil moisture decreased both GPP and NEP; however, higher summer temperature only reduced NEP with no apparent impacts on GPP. Furthermore, summer drought did not show evident impacts on either summer GPP or NEP in deciduous forests, suggesting a better potential of deciduous forests in resisting summer drought and accumulating carbon from atmosphere. These observations imply diverse responses of vegetation production to interannual summer drought and such features would be useful to improve the strengths and weaknesses of ecosystem models to better comprehend the impacts of summer drought with future climate change.     

Developing a gross primary production model for coniferous forests of northeastern USA from MODIS data

Accurate estimation of ecosystem carbon fluxes is crucial for understanding the feedbacks between the terrestrial biosphere and the atmosphere and for making climate-policy decisions. A statistical model is developed to estimate the gross primary production (GPP) of coniferous forests of northeastern USA using remotely sensed (RS) radiation (land surface temperature and near-infra red albedo) and ecosystem variables (enhanced vegetation index and global vegetation moisture index) acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. This GPP model (called R-GPP-Coni), based only on remotely sensed data, was first calibrated with GPP estimates derived from the eddy covariance flux tower of the Howland forest main tower site and then successfully transferred and validated at three other coniferous sites: the Howland forest west tower site, Duke pine forest and North Carolina loblolly pine site, which demonstrate its transferability to other coniferous ecoregions of northeastern USA. The proposed model captured the seasonal dynamics of the observed 8-day GPP successfully by explaining 84–94% of the observed variations with a root mean squared error (RMSE) ranging from 1.10 to 1.64 g C/m²/day over the 4 study sites and outperformed the primary RS-based GPP algorithm of MODIS.     

森林地上生物量遥感反演方法综述

森林地上生物量反演对理解和监测生态系统及评估人类生产生活的影响有着重要作用,日益发展的遥感技术使全球及大区域的生物量估算成为可能。近年来,不同的遥感技术和反演方法被广泛用于估算森林生物量。本文首先总结了现有的全球及区域生物量产品及其不确定性,然后综述了3类方法在森林地上生物量遥感反演中的应用,即基于单源数据的参数化方法、基于多源数据的非参数化方法和基于机理模型的反演方法,阐述了各类反演方法的特点、优势及局限性。最后从机理模型研究、多源遥感数据协同、生物量季节变化研究和遥感数据源不断丰富4个方面对今后的生物量遥感反演研究进行了展望。     

Impact of aerosols on terrestrial gross primary productivity in North China using an improved boreal ecosystem productivity simulator with satellite-based aerosol optical depth

ABSTRACT

Atmospheric aerosols can alter the direct and diffuse components of global solar radiation, which further influences terrestrial gross primary productivity (GPP) via photosynthesis. To investigate the impact of aerosols on GPP, GPP is modeled using the Boreal Ecosystem Productivity Simulator (BEPS) under two aerosol scenarios (S1& S2) over cropland and grassland ecosystems in the highly polluted North China. In S1, the aerosol-effect is not considered and an original empirical method is used when estimating direct and diffuse solar radiation in BEPS. In S2, BEPS is improved by a new empirical method which incorporates the impact of aerosols using the remote sensing-based aerosol optical depth (AOD). Results suggest that aerosols can reduce GPP of the sunlit leaves by decreasing direct solar radiation, but increase GPP of the shaded leaves by increasing diffuse solar radiation. The impact of aerosols on GPP is more significant over the cropland ecosystem (p < 0.05) with a more complex canopy structure during the peak period of the growing season. Furthermore, an AOD value of 0.3–0.6 with a diffuse fraction (the fraction of diffuse solar radiation in global solar radiation) around 30-40% can largely increase total GPP over the cropland ecosystem. The study improves the accuracy of GPP modeling using BEPS by highlighting the aerosol-effect on GPP via solar radiation over highly polluted regions.

Abbreviations: gross primary productivity (GPP); aerosol optical depth (AOD); boreal ecosystem productivity simulator (BEPS)     

利用双变量同化与PCA-Copula法的冬小麦单产估测

为了进一步提高冬小麦产量估测的精度,基于集合卡尔曼滤波算法和粒子滤波（particle filter, PF）算法,对CERES–Wheat模型模拟的冬小麦主要生育期条件植被温度指数（vegetation temperature condition index,VTCI）、叶面积指数（leaf area index, LAI）和中分辨率成像光谱仪（moderate-resolution imaging spectroradiometer, MODIS）数据反演的VTCI、LAI进行同化,利用主成分分析与Copula函数结合的方法构建单变量和双变量的综合长势监测指标,建立冬小麦单产估测模型,并通过对比分析选择最优模型,对2017—2020年关中平原的冬小麦单产进行估测。结果表明,单点尺度的同化VTCI、同化LAI均能综合反映MODIS观测值和模型模拟值的变化特征,且PF算法具有更好的同化效果;区域尺度下利用PF算法得到的同化VTCI和LAI所构建的双变量估产模型精度最高,与未同化VTCI和LAI构建的估产模型精度相比,研究区各县（区）的冬小麦估测单产与实际单产的均方根误差降低了56.25 kg/hm2,平均相对误差降低了1.51%,表明该模型能有效提高产量估测的精度,应用该模型进行大范围的冬小麦产量估测具有较好的适用性。