Knowledge Resource Center for Ecological Environment in Arid Area
| Earth observation of seasonal vegetation changes and controls in South Asia | |
| 其他题名 | Earth observation of seasonal vegetation changes and controls in South Asia |
| SANGEETA SARMAH | |
| 出版年 | 2018 |
| 学位类型 | 博士 |
| 导师 | 贾根锁 |
| 学位授予单位 | 中国科学院大学 |
| 中文摘要 | 植被是地表系统重要的组成部分，和大气系统关联密切。气候因子的季节变化和年变化，以及人类活动影响着植被动态。对全球植被变化的研究表明，伴随着农业活动的扩张，尤其灌溉农业，南亚是植被绿度变化最显著的区域之一。但是，南亚区域季风气候对植被绿度的季节趋势和波动的影响如何，我们的理解还存在较多不确定性。本研究中，我们着眼于分析植被动态的季节变化，而非年变化，以揭示长期植被动态的季节影响。使用基于遥感的归一化植被指数（NDVI）研究灌溉农业、雨养农业和自然植被1982-2013年夏季风时期（6-9月）和冬季风时期（12-4月）的时空动态。本研究基于超高分辨率辐射仪（AVHRR）第三版NDVI数据集（NDVI3g）和CRU降水、温度数据集，分析了NDVI季节变化、气候因子（降水、温度）的变化和土地利用/土地覆盖变化（LUCC），用最小二乘法分析了季节尺度NDVI和气候波动的时空趋势，用皮尔森相关分析法研究了NDVI和气候因子间的相关关系。为研究冬季风时期和夏季风时期植被对降水变化的延迟响应，对NDVI和降水间做延迟相关分析。前人研究显示NDVI对温度不存在延迟响应，因此本研究没有分析这一延迟响应相关性。为揭示近十年南亚土地利用/土地覆盖变化现象，和对长期NDVI趋势的影响，研究分析了2001-2013年南亚的土地利用变化。本研究揭示南亚植被季节动态，有助于提升对南亚区域植被生产力变化的具体理解。研究发现，不同土地利用/土地覆盖类型在两季间有不同的变化趋势。在南亚区域，冬季风时期贡献了更多的植被年变化。这是本研究的重要结论。由于冬季风时期灌溉农田生产力的提高，南亚植被有显著变绿趋势。同时期，自然植被有显著变黄趋势。冬季风时期的最高温度显著升高，但是，南亚的降水变化趋势不显著。LUCC主要发生在喜马拉雅山麓、印度东部和西高止山脉，这些区域土地覆盖类型从森林变为农田或城市，或退化为开放灌木和裸地。研究显示，气候变化和LUCC共同影响自然植被区域NDVI的长期变化趋势，尤其在山麓区域。在其他区域，人类活动（农业推进）有更重要的影响。半干旱区夏季风时期的降水波动，和半湿润地区冬季风时期的降水波动对NDVI有显著影响。同时，夏季风时期，南亚西部干旱区植被动态和气温波动呈现负相关关系；冬季风时期，南亚南部和东部植被动态和气温波动呈负相关。这些结果表明，NDVI与气候波动之间关系的空间分布有明显的季节变化。目前，先进的耕作技术有利于该区域农田生产力的提升。但是，农作物产量受到气候变化的威胁。本研究利用了AVHRR NDVI数据集研究南亚的植被动态特征。但前人研究表明，不同植被指数数据集在同一区域可能会展现不同的趋势特征。本研究基于不同遥感数据集分析南亚区域植被变化趋势和波动特征，将有助于准确了解该地区的植被变化趋势。目前在南亚区域相似的研究还比较少。本研究基于中等分辨率成像光谱仪（MODIS）NDVI和EVI，分析了南亚区域在两季间，次区域尺度植被时空变化趋势的一致性。研究假设MODIS数据集具有较高的质量。利用Mann-Kendall趋势估计方法，分析了2001-2013年三种数据集在不同季节（冬季风时期和夏季风时期）的时空趋势，以确定三种数据集反映的季节动态的一致性。在8个生态区，计算所有趋势的相关性，以了解三种数据集反映的趋势在不同生态区有怎样的不同。通过对不同植被指数（VI）数据集反映的植被时空变化趋势一致性的评价，加深我们对这一复杂地形区域植被变化异质性的了解。研究结果表明，NDVI3g反映的植被时空变化趋势与MODIS NDVI和EVI都很相似，以农田为主的半干旱区有变绿趋势。所有数据集都反映出，热带和亚热带区域有显著变黄趋势，尤其在冬季风时期。NDVI3g和NODIS NDVI以及NDIV3g和MODIS EVI间的相关性在冬季风时期比夏季风时期强虽然三种数据集都能说明南亚区域的植被变化趋势，但在夏季风时期，不同数据集在热带和亚热带密集植被（湿润区）和复杂地形区域反映出一些差异。NDVI 3g和MODIS EVI的趋势更为一致，尤其在冬季风时期。这一结果表明，高质量的内部订正方法使NDVI3g数据集质量有所提高，使其能够更可靠地反映南亚区域长期植被变化趋势。但是，在夏季风时期，数据订正给三种不同的数据集都引入误差。研究人员在利用遥感时间序列数据研究植被季节动态时应更加谨慎。应在各个次区域进行广泛的实地测量和验证，以进一步校准遥感数据集，量化南亚的植被动态变化。生态系统生产力的气候敏感性对生态系统结构和功能有较大影响。植被动态及其与气候变化的相互作用在不同生态地理区存在显著差异，因此研究气候对群落尺度的影响可以提高对生态系统-气候相互作用的认识。由于大尺度研究使用的植被指数数据集空间分辨率较粗糙，数据集序列时间不一致，因此可能遗漏一些区域细节现象，因此较小空间尺度的植被动态异质性研究是非常重要的，因为这可能对全球或区域尺度植被动态的研究带来较大影响。因此，应从恰当处理的遥感时间序列数据中得到植被动态，研究次区域尺度降水、温度等气候因子的影响。目前，基于1km尺度MODIS EVI时间序列数据集分析植被对气候变化敏感性的研究还较少。本研究基于MODIS EVI 13年（2001-2013年）时间序列数据，研究南亚区域冬季风时期和夏季风时期温度和降水对植被季节波动的影响。使用TRMM数据集和CRU数据集，用以说明南亚的气候波动。在分析南亚不同生态区季节EVI变化和气候波动相关性时，采用Pearson相关分析方法。结果表明，在不同地理区域和不同季节，生态系统对不同气候因子敏感性不同。在冬季风时期和夏季风时期，降水都是促进南亚植被生长的关键因素，但在冬季风季节，气温影响更显著。在冬季风时期，多数生态系统植被活动对气候变化的敏感性高于夏季风时期，表明冬季风时期的重要性。在夏季风时期，半干旱区EVI和TRMM波动有更显著的相关性，而在冬季风时期，半湿润区植被波动和降水有显著的相关性。这一结果与NDVI3g和CRU降水间相关性的分析相似。热带和亚热带落叶阔叶林、山地草原、灌木对冬季风时期的温度变化有更显著的敏感性。干旱灌木区是受干旱等极端气候变化影响的重要生态区，尤其在夏季风时期。由于这些区域分布着大量农田，因此极端气候可能威胁到未来南亚粮食安全。本研究反映了植被生长和气候波动相关性的空间和季节异质性，但是我们没有揭示出对气候最敏感的生态区域，因为季节性植被和气候波动的关系在不同生态区有很大不同，需要进一步具体研究南亚区域其他引起植被变化的相关因子。本研究有助于加深对印度次大陆植被季节动态的认识，但也存在一定局限性。由于降水和温度是植被季节变化的主要影响因子，因此本研究重点关注降水和温度。但是其他影响因子的研究，如物候变化，二氧化碳施肥效应，氮沉降等，将有助于进一步解释植被季节变化的其他方面。此外，本研究没有涉及对最佳数据集进行评估，以确定南亚植被变化趋势。因此，未来不同次区域更高分辨率数据集和广泛的实地观测资料将有助于南亚植被动态其他方面的研究。 |
| 英文摘要 | Vegetation Condition is an important component of land surface system and is intimately linked with the climate system. Previous studies have demonstrated that vegetation dynamics is mostly driven by seasonal and inter annual variations of climatic parameters along with other anthropogenic activities. Global studies on vegetation trend analysis reflect that the South Asia (SA) is the one of the most remarkable regions for changing vegetation greenness, accompanying its major expansion of agricultural activities, especially irrigated farming. The influence of the monsoon climate on the seasonal trends and anomalies of vegetation greenness is not well understood in this area. Herein, we analyzed the seasonal rather than annual dynamics of the vegetation in order to illustrate the significance of the seasonal influences on long term vegetation patterns. The satellite-based Normalized Difference Vegetation Index (NDVI) was used to investigate various spatiotemporal patterns in vegetation activity in summer monsoon (June- September) and winter monsoon (December-April) seasons and among irrigated croplands, rainfed croplands, and natural vegetation areas during 1982-2013. Seasonal NDVI variations with climatic factors (precipitation and temperature) and land use and land cover changes (LUCC) also have been explored in this study. For this purpose Advanced Very High Resolution Radiometer (AVHRR) version 3 NDVI (NDVI3g) and Climate Research Unit (CRU) datasets for precipitation and temperature were used. Here, spatiotemporal trend analysis of seasonal NDVI and climatic anomalies were done using least square linear regression method. Pearson’s correlation analysis were done among the NDVI and climatic parameters to evaluate the relationship between them. Lagged correlation of NDVI and precipitation were also computed to investigate the delayed response of vegetation to rainfall variations during both monsoon seasons, however it was not done in case of temperature as literatures confirmed zero lag between NDVI and temperature. Land use change analysis was also carried out since 2001 to 2013 to distinguish the LUCC over SA during latest decade and the impact of this on the long term NDVI trend. The results of this study revealed that the seasonal dynamics of vegetation could enhance the detailed understanding of the vegetation productivity over the region. We found distinctive vegetation trends between two monsoon seasons and among the major land use/cover classes. Our study exposed that the winter monsoons contributed greater variability to the overall vegetation dynamics of SA which could be marked as an important finding of this research. This region experienced major greening due to the increased productivity over irrigated croplands during the winter monsoon season; in the meantime, browning trends were prominent over natural vegetation areas during the same season. Maximum temperatures showed a significant increasing trend during the WM season; however, the precipitation trend was not discernible over SA. The LUCC analysis demonstrated noticeable changes in the Himalayan foothills, the eastern part of central India and Western Ghats where the conversions occurred mostly from forest to either croplands or urban areas. Some also degraded to open shrub lands or bare lands. This study indicated that both the climate variability and land use/cover change activities had an integrated effects on the long term NDVI trends in natural vegetation areas, especially in the hilly regions, while anthropogenic activities (agricultural advancements) played a crucial role in the rest of the area. Precipitation anomaly showed significant impact on NDVI activity mostly in semi-arid regions during summer monsoon and sub-humid regions in winter monsoon season. Likewise, during summer monsoon, the western dry areas of SA displayed a negative correlation with temperature anomaly while in winter monsoon season the negative correlation largely shifted towards southern and eastern part of the region. These results revealed that there was a distinct seasonal variation in spatial distribution of relationship between NDVI and climate anomalies over SA. This study demonstrated that until now, advanced cultivation techniques have proven to be beneficial for the region in terms of the productivity of croplands. However, the crop productivity is at risk under recent climate change scenarios.The characterization of vegetation dynamics over SA have been mostly conducted using satellite time series of AVHRR NDVI, however, literature said that different vegetation index datasets may show mixed trend patterns over the same region. Studies investigating the characterization of the vegetation trends and variabilities derived from various satellite datasets are very rare over South Asia which would be helpful to understand the diverse vegetation trend patterns over this region. This research analyzed the consistency of the vegetation spatiotemporal trends from NDVI3g with Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI and Enhanced Vegetation Index (EVI) at sub-regional scale during monsoon seasons in SA, assuming MODIS products are of higher quality. The seasonal (summer monsoon and winter monsoon) spatiotemporal trends from all the three datasets were analysed using Mann-Kendall trend estimation method during the overlapping period of 2001 to 2013. Vegetation trends were computed during summer monsoon and winter monsoon seasons to characterize the seasonal patterns of the consistency of the three datasets. Correlations among all the trends were also computed over eight ecoregions of SA to see how the consistency among them differ in various ecoregions. This evaluation of the consistency of vegetation spatiotemporal trends derived from various vegetation index (VI) datasets enhanced our understanding of the heterogeneous trends of vegetation growth over this vast complicated terrain. Results indicated that the spatiotemporal vegetation trends derived from NDVI3g were quite analogous to both MODIS NDVI and EVI demonstrating greening over semi-arid regions where croplands were the dominant land use category. All the datasets displayed browning over tropical and subtropical forest areas, especially during winter monsoon season. Correlations among NDVI3g and MODIS NDVI as well as NDVI3g and EVI were better during winter monsoon season than summer monsoon season. Although all the three datasets were compatible to illustrate the vegetation trends over the region, some discrepancies occurred in tropical and subtropical densely vegetated (humid) and complex topographic areas specifically during summer monsoon season. Noticeably, NDVI3g trends showed better agreement with MODIS EVI trends particularly during winter monsoon season. This result indicated that high quality inter calibration methods have improved NDVI3g datasets making it reliable to characterize the long term vegetation trends over South Asia. However, calibration errors still could introduce biases in all the three datasets during summer monsoon season. This exhibited that researchers should be careful while evaluating the seasonal vegetation dynamics over SA using satellite time series data. Extensive field measurements and validations across various sub-regions are required to further calibrate the satellite derived datasets to quantify the vegetation dynamics of South Asia. Climate sensitivity of ecosystem productivity has a strong impact on both the structuring and functioning of ecosystems. Vegetation dynamics and their interactions with climate change varies significantly across various eco-geographical regions, and thus exploring the effect of climate on vegetation dynamics at the biome scale could improve the understanding of the ecosystem-climate interaction. Studies of the small-scale heterogeneity of vegetation dynamics are important as these may have a big impact on the global or regional scale vegetation dynamics because the big picture studies sometime fail to depict some of the local detail due to the coarse spatial resolution of vegetation index data used and differences in observation period. Therefore, vegetation dynamics derived from correctly processed satellite time-series data should be used as an indicator of vegetation response to climatic growth constraints such as rainfall and temperature at a sub-regional scale. Moreover, studies investigating the climate sensitivity of vegetation changes using moderate resolution (1000 m) MODIS EVI time-series data over SA are very rare. In this study, we investigated the seasonal vegetation dynamics in response to climatic anomalies (precipitation and temperature) across South Asia using the 13-years (2001–2013) of MODIS EVI time-series data during summer monsoon and winter monsoon season. Tropical Precipitation Measuring Mission (TRMM) data for precipitation and CRU dataset for temperatures were used to illustrate the climate anomalies over the region. Pearson’s correlation method was applied on seasonal EVI and climatic anomalies to quantify the relationship among them over various ecoregions of SA. The results showed that all the ecoregions had variations of sensitivities to different climatic factors during both the monsoon seasons over this diverse geographical territory. Precipitation was the key factor facilitating the vegetation growth in this region during both the season, however, temperature seemed to have a significant impact during winter monsoon season. Vegetation activity in most of the ecoregions showed more sensitivity towards climatic variations during winter monsoon season than the summer monsoon season which indicated the significance of the winter monsoon season in the vegetation dynamics of this region. The EVI and TRMM anomaly showed a significant correlation over semi-arid ecoregions during summer monsoon season, however, vegetation anomalies over sub-humid ecoregions displayed significant association with rainfall during winter monsoon season. This result was similar as described by the correlation analysis between NDVI3g and CRU precipitation over SA. Tropical and subtropical deciduous broadleaf forest and montane grassland and shrub land seemed to have discernible sensitivity towards temperature variations during winter monsoon season. Xeric shrub land was the significant ecoregion vulnerable to extreme climatic variations like drought, especially during summer monsoon season as the large portion of this ecoregion is mostly dominated by croplands and this might risk the future food security over SA. This study reflected on the spatial and seasonal heterogeneity of the relationship between vegetation growth and climate anomalies, however we could not reveal the most climate sensitive ecoregion because the relationship between the seasonal vegetation and climate anomalies were so diverse across different ecoregions which required to be further studied in detail with respect to other possible key factors causing the vegetation change over South Asia.This study was helpful to advance the understanding of the seasonal vegetation dynamics of Indian subcontinent, however, there were some limitations. In this research we mainly focused on precipitation and temperature as the key drivers of vegetation changes over SA, while identification of the impacts of the other factors, such as phenological changes, CO2 fertilization, nitrogen deposition, etc., could further disclose various other aspects of seasonal vegetation change patterns found in this study. Also, evaluation of the best suitable dataset to quantify vegetation trends over SA was beyond the scope of this study. Therefore, future validations with fine resolution datasets and extensive field measurements across various sub-regions would be beneficial to find out some more exciting aspects of vegetation dynamics over South Asia. |
| 中文关键词 | 植被趋势 ; 气候变化 ; NDVI ; EVI ; 南亚 |
| 英文关键词 | vegetation trend climate change NDVI EVI South Asia |
| 语种 | 中文 |
| 国家 | 中国 |
| 来源学科分类 | 大气物理学与大气环境 |
| 来源机构 | 中国科学院大气物理研究所 |
| 资源类型 | 学位论文 |
| 条目标识符 | http://119.78.100.177/qdio/handle/2XILL650/288066 |
| 推荐引用方式 GB/T 7714 | SANGEETA SARMAH. Earth observation of seasonal vegetation changes and controls in South Asia[D]. 中国科学院大学,2018. |
| 条目包含的文件 | 条目无相关文件。 | |||||
| 个性服务 |
| 推荐该条目 |
| 保存到收藏夹 |
| 导出为Endnote文件 |
| 谷歌学术 |
| 谷歌学术中相似的文章 |
| [SANGEETA SARMAH]的文章 |
| 百度学术 |
| 百度学术中相似的文章 |
| [SANGEETA SARMAH]的文章 |
| 必应学术 |
| 必应学术中相似的文章 |
| [SANGEETA SARMAH]的文章 |
| 相关权益政策 |
| 暂无数据 |
| 收藏/分享 |
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
