中国草原产草量遥感监测

根据MODIS遥感数据和同期地面调查数据,对我国2005年草原产草量进行了系统估算。用MODIS数据计算全国草原的归一化植被指数NDVI,针对6个不同类型草原区建立了NDVI和地面样方的产草量之间的关系模型,用这些模型推算全国草原产草量分布。结论如下：（1）2005年我国草原有3个牧草高产中心,分别位于东北呼伦贝尔草原、锡林浩特草原和大兴安岭西麓;青海东部、四川西北部和甘肃中南部以及新疆西北部;（2）2005年全国草原干草总产量达到29421．39万t,平均单产达到829．67kg／hm^2干草;（3）2005年干草产量位列前7位的省区依次是内蒙古、青海、新疆、四川、西藏、黑龙江和甘肃;例如,内蒙古因草原面积大而成为我国第一大草原牧草生产省,2005年有6037．08万t干草;（4）总产草量位于前5位的草地类型依次为高寒草甸类、温性草原类、低地草甸类、温性草甸草原类与山地草甸类;（5）2005年8月份全国草原产草量与2004年同期相比总体持平;各草原大省的变化情况分别为：青海、甘肃2省区的草产量略有增加,青海增加了9．02％、甘肃增加了3．63％;内蒙古减少约3％;西藏、新疆和四川3省区基本与2004年同期持平。研究结果对我国草原监理、草原畜牧业发展和草地生态系统研究具有较大的参考价值。     

中国草原生态系统退化的价值损失量遥感估算

在传统的草原生态系统退化价值估算基础上,利用遥感技术对草原生态系统退化的价值损失进行估算,利用2003和2005年的MODIS数据获得草原生态系统退化的价值损失量空间分布图.结果表明:2003-2005年价值损失量共为66.603亿美元;单位面积价值损失量在0～1 000美元·km-2的面积约占草原总面积的1/4,价值损失量1 000～3 000美元·km-2的面积约占草原总面积的1/3;内蒙古、新疆、西藏、青海、甘肃、云南和四川等西部7省的价值损失总量为52.219亿美元,占总价值损失量的78.41%,其中,内蒙古的价值损失量最大,占总价值损失量的25.89%.     

锡林郭勒生物圈保护区草原生态系统服务间接价值动态评估

草地是我国面积最大的陆地生态系统类型,对于维持我国北方的生态安全具有重要价值。北方草原的大面积退化已引起人们的极大关注。以内蒙古锡林郭勒生物圈保护区为研究区域．利用2个年份的遥感数据,结合地面样方调查,在GIS的支持下,绘制出草原退化空间分布图。选取草原生态系统在控制土壤侵蚀、减少土壤肥力流失、营养物质循环、CO2储存和O2释放方面的功能指标,对草原生态系统服务间接价值进行了动态评估。1985年和1999年草原退化面积分别为7190．9和7689．3km^2,占研究区总面积的67．19％和71．86％。2个年份草原生态系统间接服务总价值分别为28．24和26．75亿元．下降了5．28％,单位面积间接服务价值分别为2990和2830元。下降了5．35％。人类的过度放牧是导致草原大面积退化,生态系统功能降低和服务价值下降的主要原因。     

内蒙古典型草原地带退化草原的恢复动态

对退化草原自然恢复演替过程的监测,是认识其恢复过程、机理、探讨恢复的限制因子和进一步制定快速、低投入生态恢复措施的基础。通过对内蒙古典型草原地带退化草原［星毛委陵菜（Potentila acaulis）、冷蒿（Artemisia frigida）和寸草苔（Carex duriuscula）等为主］封育恢复动态 8年的监测表明：在退化草原群落自然恢复过程中,植物高度上升;植被盖度和生物量先增加,尔后基本稳定,或略有下降的趋势。群落的物种丰富度增加微弱,这可能与该区草原植物和野生动物协同进化有关,同时与数千年的家畜放牧史有关。即植物物种对放牧有较大的耐性,停止放牧,物种的丰富度也增加较少;群落结构的变化主要表现在不同种群优势度的消长上。群落均匀度指数的动态过程分析表明,恢复8年的草原群落尚未达到天然草原状态,尚需进一步的监测研究。     

内蒙古羊草草原群落主要植物的热值动态

对内蒙古羊草草原主要植物种群的干重热值动态研究表明,热值随植物种类,植物部位,取样时植物所处物候期及气候条件的不同而变化,羊草草原主要植物种群地上部分热值的变动范围在15703－18141J／g之间,其中灌木小叶锦鸡儿（Caragang microphylla)的热值最高,禾草类植物的热值多数较高,而大多数杂类草的热值相对较低,主要植物种群地下部分热值的分布范围为15051－16410J/g。其中根茎型草地下部分热值较高。不同种类植物地下部分热值差异并不与地上部分一致,根茎型禾草地上、地下部分热值差异较小,而须根型植物差异较在,不同种群的植物地上部分热值随植物候期的不同而波动,其变化规律是与植物种群本身的生物学特性相联系的,不同植物种群热值的年际小 规律有所不同,羊草（Leymus chinensis)、大针茅（Stipa grandis)和洽草（Koeloria cristata)的年际热值波动相关显著。但与生长季降水量和生长季累积日照时数之间无明显相关性,在某种程度上,植物热值的 种内变化反映了植物生长状况的差异。     

内蒙古典型草原生态系统定位研究最新进展

中国科学院内蒙古草原生态系统定位研究站最近几年在草原生态系统多功能、草原生态系统碳循环与温室气体、气候变化对植物群落结构影响长期观测、草原生态系统退化与恢复过程、草原生态系统硫生物地球化学循环、放牧生态系统管理等领域取得较大进展。研究结果指出,农垦活动使有机碳损失34％,草原植被——土壤系统是甲烷的弱汇;20多年来,这一地区气候变化有变暖趋势,冬季增温明显,春季干旱进一步加剧,影响初级生产力生产;系统提出草原退化是草原生态系统退化,并对退化阶段进行了划分;在内蒙古冷蒿小禾草退化草原施用硫肥对提高初级生产力与羊毛性能方面有明显效果。     

荒漠草原牧草净生产力的遥感估算及其可行性研究

应用2005年5~10月份EOS/MODIS资料,采用光能利用率估算方法,对中国西北荒漠绿洲区甘肃省安西县荒漠草原区的月净第一生产力、年净第一生产力进行估算,并结合地面观测,分析了不同牧草类型NPP的时间变化和空间分布特征。结果表明,安西县荒漠草原区2005年牧草生长期内NPP的变化基本是6、7月份最大,其它月份较小,5~6月份是牧草叶面积变化比较大的时期,也是NPP变化比较显著的时期;荒漠牧草对光的利用能力和对CO2同化能力有很大差异,白刺、红砂、黑果枸杞是荒漠草场净生产力相对比较大的牧草类型;荒漠牧草区年净第一生产力的空间分布有很大的差异,禁牧区牧草的年NPP相对非禁牧区的NPP较大,这一结论与地面观测结果相一致,NPP估算值平均误差为12.6%。     

荒漠草原牧草生物量的遥感估算及空间分布

应用2005年5-10月份EOS/MODIS资料,采用光能利用率算法,估算了中国西北荒漠绿洲区域NPP,并结合地上生物量观测资料,建立了基于NPP的地上生物量估算模型,同时分析了生物量和NPP时空分布特征。结果表明,2005年牧草生长期内NPP的变化基本是6、7月份最大,其它月份较小,5-6月份是牧草叶面积变化比较大的时期,也是NPP变化比较显著的时期;生物量模型估算得到的44个样本的生物量误差基本都小于20%,说明模型能够反映牧草的实际生长状况;不同荒漠牧草对光的利用能力和对CO2同化能力有很大差异,骆驼刺、红砂、黑果枸杞是荒漠牧草生物量较大的牧草类型;牧草类型、牧草盖度以及平均高度是造成不同牧草产量相差较大的重要原因。     

基于MODIS NPP数据的青海湖流域产草量与载畜量估算研究

快速评价区域草地生产力状况, 是制定牧区草畜平衡策略, 保障草地生态系统健康可持续发展的前提基础。基于MODIS NPP产品数据, 对青海湖流域草地生产力进行了估算和评价。结果表明: (1)牛羊可食性最高的高寒草甸、高寒草原和温性草原, 其单位面积产草量估算值的均方根误差RMSE为26.15 g·m-2, 表明该方法可以快速估算区域尺度的产草量。(2)全年干草产量为145.42万t, 其中能被牛羊直接采食利用的牧草共计59.18万t, 理论载畜量为81.07万羊单位; (3)影响单位面积干草产量的主要气候因子, 海拔3500 m以下地区, 是活动积温且与单位面积干草产量呈负相关关系; 3500 m以上地区, 是温度且与单位面积产草量呈正相关关系。研究结果可为高寒地区合理规划牧业生产活动、生态补偿等提供科学依据和决策支持, 也可为全国其他牧区的草原产草量的估算提供参考。     

低空无人机遥感在草原监测评价中的应用进展

低空域无人机遥感技术具有高时效性、高分辨率、低成本、易操控等优势,作为地面与高空遥感（航天与航空遥感）间测量尺度空缺的有益补充,低空无人机遥感扩展了样地样方空间尺度,提高了中、小尺度遥感观测信息的精细化程度,实现了草原生境信息的快速采集、处理与分析应用,是草原"星-空-地"一体化监测的重要组成。针对草原监测评价,总结了国内外低空无人机遥感在草原基况调查（草原草层高度监测、草原植被覆盖度监测与草原地上生物量估算）、草原动态监测（草原植被长势监测、草原产草量估测与草畜平衡监测）和草原应急管理（草原火灾、雪灾与生物灾害监测）中的应用。结合大数据、人工智能、云计算与物联网等新型技术,分析了低空无人机遥感在草原生态监测领域存在的不足和未来的发展方向,以期为低空无人机遥感关于草原监测评价与智慧草原的后续研究提供参考。     

Remote sensing monitoring upon the grass production in China

Using MODIS remote sensing data and ground truth data, a thorough investigation was conducted to monitor the productivity of grasslands in China for the year 2005. The Normalized Difference Vegetation Index (NDVI) was first computed from the MODIS data. Then the data from the NDVI images were used to correlate with the grass yield data from the ground sampling campaigns. Six regional models were accordingly established from the correlation for estimation of grass production in the six main types of steppe in China. The main results from the estimation could be summarized as follows: (1) High grass productivity in 2005 was obtained in the following 3 regions: the grassland covering Hulunbuir, Xilinhaote, and the western Daxing'anling, the region including the eastern Qinghai, the northwestern Sichuan and the mid-southern Gansu, and the northwestern Xinjiang region. (2) Total hay output from the grasslands in China amounted to 294213.86 thousand tons in 2005 with an average yield of 829.67 kg/hm2. (3) The following 7 provinces were the largest grass producers in China: Inner Mongolia, Qinghai, Xinjiang, Sichuan, Tibet, Heilongjiang and Gansu. For example, Inner Mongolia produced 60370.82 thousand tons of hay in 2005, and hence became the No. 1 grass producer of China. (4) Among the steppe types, the following 5 had the largest grass production: Alpine meadow, Temperate steppe, Low-land meadow, Temperate meadow steppe and Montane meadow with total production accounting for 62.2% in China. (5) Grass production of the entire China in August 2005 remained at the same level as that in August 2004. However, the situations of major grassland provinces were different: grass production in both Qinghai and Gansu in 2005 increased to 9.02% and 3.63%, respectively, when compared with that in 2004. The grass production in Inner Mongolia decreased by 3%, while the production in Tibet, Xinjiang and Sichuan remains unchanged when compared with that in 2004. These results were very important for grassland administration, pasture grazing and grassland ecosystem studies in China.     

Remote sensing monitoring upon the grass production in China

Using MODIS remote sensing data and ground truth data, a thorough investigation was conducted to monitor the productivity of grasslands in China for the year 2005. The Normalized Difference Vegetation Index (NDVI) was first computed from the MODIS data. Then the data from the NDVI images were used to correlate with the grass yield data from the ground sampling campaigns. Six regional models were accordingly established from the correlation for estimation of grass production in the six main types of steppe in China. The main results from the estimation could be summarized as follows: (1) High grass productivity in 2005 was obtained in the following 3 regions: the grassland covering Hulunbuir, Xilinhaote, and the western Daxing'anling, the region including the eastern Qinghai, the northwestern Sichuan and the mid-southern Gansu, and the northwestern Xinjiang region. (2) Total hay output from the grasslands in China amounted to 294213.86 thousand tons in 2005 with an average yield of 829.67 kg/hm2. (3) The following 7 provinces were the largest grass producers in China: Inner Mongolia, Qinghai, Xinjiang, Sichuan, Tibet, Heilongjiang and Gansu. For example, Inner Mongolia produced 60370.82 thousand tons of hay in 2005, and hence became the No. 1 grass producer of China. (4) Among the steppe types, the following 5 had the largest grass production: Alpine meadow, Temperate steppe, Low-land meadow, Temperate meadow steppe and Montane meadow with total production accounting for 62.2% in China. (5) Grass production of the entire China in August 2005 remained at the same level as that in August 2004. However, the situations of major grassland provinces were different: grass production in both Qinghai and Gansu in 2005 increased to 9.02% and 3.63%, respectively, when compared with that in 2004. The grass production in Inner Mongolia decreased by 3%, while the production in Tibet, Xinjiang and Sichuan remains unchanged when compared with that in 2004. These results were very important for grassland administration, pasture grazing and grassland ecosystem studies in China.     

陕北榆林地区土地沙漠化动态遥感研究

土地沙漠化是人类面临的一个重要环境问题,在人为过度的经济活动干拢和生态脆弱的自然环境变化的影响下,正在继续蔓延,日益扩大,已引起国际社会和有识之士的高度重视。掌握其动态变化的规律,监测其发展趋势乃成为当前国际沙漠化研究的一个重要内容。地处陕西北部长城沿线的榆林地区,生态上较脆弱,气候上较敏感。笔者利用1984年和1994年两期ＴＭ图像进行判读勾绘,编制土地沙漠化类型图,在ＧＩＳ的支持下,建立空间数据库,套合叠加产生一种新的图件土地沙漠化动态图,获取动态变化资料数据,掌握了该区十年来土地沙漠化的动态变化趋势。1　区域环…     

运用遥感估算中国陆地植被净第一性生产力

净第一性生产力 (NPP)研究方法很多 ,运用NOAA_AVHRR的可见光波段、近红外波段和热红外波段来提取和反演地面参数 ,进而准确估算陆地植被净第一性生产力 ,是一种全新的研究手段。利用遥感数据进行生物量和净第一性生产力的估算 ,主要是采用光能利用率模型 ,即通过NPP与植物吸收的光合有效辐射 (APAR)和植物将所吸收的光合有效辐射转化为有机物的转化率 (ε)的关系来实现的。用数学公式可表达为 :NPP =(FPAR×PAR)×[ε ×σT×σE×σS× (1-Ym)× (1-Yg) ]。在遥感和地理信息系统技术的支持下 ,以 1990年每旬的 8km分辨率的NOAA_AVHRR 1～ 5通道的影像为数据源 ,对中国每旬的陆地植被净第一性生产力进行估算 ,然后累加得出全年的NPP值。估算结果 :1990年中国陆地植被NPP总量为 6 .13× 10 9tC·a-1,NPP最高值为 1812 .9gC/m2 。根据计算的结果 ,对中国大陆植被NPP的分布规律进行了分析。遥感模型能够以面代点 ,比较真实地反映陆地植被NPP的时空分布状况 ,与中国植被分布的地理规律性相符 ,这是其他统计模型所无法比拟的。     

Mapping Tension: Remote Sensing and the Production of a Statewide Land Cover Map

This article examines the process by which remotely sensed land cover maps work to both simplify and complicate landscapes. The central argument is not merely that the construction of land cover maps is complex, but that the points of complexity often arise through the process of trying to simplify. In other words, the forces of complexity are intimately connected to the forces of simplicity and vice versa. This article takes as a case study the production of WISCLAND, (Wisconsin Initiative for Statewide Cooperation on Landscape Analysis and Data), a statewide land cover map of Wisconsin derived from remote sensing data and GIS (Geographic Information Systems) technologies and proceeds by analysis of mapping methodology, practice, and representation. In addition to the development of a more nuanced critique of the use of land cover maps, it facilitates the possibility for a constructive dialogue between remote sensing practitioners and the critical GIS community.     

内蒙古典型草原土壤固碳潜力及其不确定性的估算

以内蒙古锡林郭勒盟典型草原围封和放牧对照的实测数据为基础,比较了不同植物群落类型的碳储量和固碳量,并估算了重度退化草地完全恢复后的土壤固碳潜力.结果表明: 在局地尺度上,以羊草、大针茅和克氏针茅为优势种的3种植被类型的年均固碳量均随着围封年限的增加而减少.羊草和大针茅草原群落的碳储量显著高于克氏针茅草原群落,而在过牧导致的退化过程中,后者流失了更多的土壤有机碳,其固碳潜力高于前两者.在区域尺度上,内蒙古典型草原土壤表层0～20 cm的固碳潜力在-0.03×10⁴～3.71×10⁴ kg C·a^-1范围内变化,总固碳潜力为12.1×10⁸kg C·a^-1.不确定性分析发现,忽略石砾含量不会造成土壤固碳潜力估算值的较大偏差,而气象数据空间插值造成的误差为±4.7×10⁹ kg C·a^-1.假设未来生长季降水的平均变化值为-3.2 mm·(10 a)^-1时,土壤固碳潜力将减少1.07×10⁸ kg C·(10 a)^-1.     

Estimation of Net Primary Productivity of Terrestrial Vegetation in China by Remote Sensing (in English)

Among the many approaches for studying the net primary productivity ( NPP ), a new method by using remote sensing was introduced in this paper. With spectral information source (the visible band, near infrared band and thermal infrared band) of NOAA-AVHRR, we can get the relative index and parameters, which can be used for estimating NPP of terrestrial vegetation. By means of remote sensing, the estimation of biomass and NPP is mainly based on the models of light energy utilization. In other words, the biomass and NPP can be calculated from the relation among NPP , absorbed photosynthetical active radiation (APAR) and the rate (ε) of transformation of APAR to organic matter, thus：NPP=(FPAR×PAR)×［ε*×σT×σE×σS×(1-Ym)×（1-Yg）］ . Based upon remote sensing (RS) and geographic information system (GIS), the NPP of terrestrial vegetation in China in every ten days was calculated, and the annual NPP was integrated. The result showed that the total NPP of terrestrial vegetation in China was 6.13×109 t C·a-1in 1990 and the maximum NPP was 1 812.9 g C/m. According to this result, the spatio-temporal distribution of NPP was analyzed. Comparing to the statistical models, the RS model, using area object other than point one, can better reflect the distribution of NPP , and match the geographic distribution of vegetation in China.