气象卫星遥感在干旱监测中的应用

本文根据ＮＯＡＡ极轨气象卫星ＡＶＨＲＲ资料计算生成植被指数相对距平图用于旱情监测，并在江苏省１９９４年特大干旱中进行了应用服务     

基于决策树的城市黑臭水体遥感分级

水体黑臭程度遥感监测是了解城市水质现状和综合评价城市水环境治理效果的重要手段.以南京、常州、无锡和扬州为研究区,共采集171个样点,同步测量水质参数和光学参数,分析黑臭水体与一般水体的水色和光学特征,构建决策树模型进行重度黑臭水体、轻度黑臭水体和非黑臭水体（记为一般水体）识别.结果表明：①根据色度可将水体分为1~6类水体,其中,类型1~4为黑臭水体,分别为灰黑色、深灰色、灰色和浅灰色水体,类型5和类型6水体为一般水体,分别为绿色系和黄色系水体;②类型1水体的非色素颗粒物和有色可溶性有机物含量高,但色素颗粒物的吸收并不占主导,类型2和5水体的吸收以色素颗粒物吸收占主导,类型3、4和6水体的吸收以非色素颗粒物吸收占主导;③根据六类水体的反射光谱差异用黑臭水体差值指数（difference of black-odorous water index,DBWI）、三波段面积水体指数（green-red-nir area water index,G-R-NIR AWI）、绿光波段反射率和归一化黑臭水体指数（normalized difference black-odorous water index,NDBWI）构建的水体分类识别决策树,能够有效识别出重、轻度黑臭水体和一般水体;④将决策树模型应用于2019年4月9日扬州的PlanetScope影像上,并利用10个同步过境点进行验证,整体识别精度达到80.00%,K值达到0.67.通过水色分类后的城市水体分级模型方法,可推广应用于类似的水体,为黑臭水体监管提供技术方法.     

京津冀地区一次污染过程的星载地基激光雷达联合观测分析

针对2018年3月9—15日京津冀地区的一次空气重污染过程,进行了基于地基颗粒物激光雷达组网的星载-地基联合观测分析。颗粒物激光雷达观测到污染前期为局地污染累积过程,中期有明显的污染物区域传输过程,北京受太行山沿线城市污染输送影响较大。风廓线激光雷达观测结果表明:此次污染过程近地面主要为偏南风且风力较弱,冷空气到来时风向转为较强东北风,导致污染消散。微波辐射计观测到保定在污染过程中出现持续6 d的逆温层,同时在污染过程中近地面相对湿度较高,逆温层被打破后污染开始消散。在污染过程的各个阶段中,污染团的空间分布与变化特征均被很好地反映出来,可见地天联合观测对污染物的累积与输送研究有较大的意义,能对京津冀及周边地区的大气污染联防联控提供有力支持。     

秸秆焚烧影响南京空气质量的成因探讨

综合利用卫星遥感的火点和云覆盖信息,结合气团后向轨迹分析,探讨了由秸秆焚烧造成的空气污染物的区域尺度输送和本地源对城市空气质量的影响.结果表明,在一定气象条件下,污染物可以发生区域尺度的输送,上风火点与下风城市的污染有明显的相关,将空气污染分为局地型(如,2006年5月31日、2009年11月8日)、区域型(如,2008年10月28日),以及局地区域相结合型(如,2006年6月14日、2007年6月5日、2008年6月2日)3种.应用本文的方法,在有云时,可以通过部分火点和气团后向轨迹分析推测污染物源地.空气污染气象条件分析表明,秸秆焚烧若伴随高空(500hPa)有槽(或位于槽前),低空存在弱切变,气流由周边向中心辐合;同时,若在均压场控制下,等压线稀疏,风速较小或静风,污染物则易积聚而不易输送;逆温层的形成将污染物禁锢在混合层以下,不利于垂直扩散;再加上较大的相对湿度,有利于霾的形成,造成严重空气污染.     

基于卫星遥感的广东地区对流层二氧化碳时空变化特征

利用瓦里关全球本底站和番禺气象局地面观测的CO2浓度资料对SCIAMACHY反演得到的对流层CO2产品进行验证.结果显示:SCIAMACHY产品能较好地反映对流层CO2的分布状况,在珠三角地区反演和观测的残差为1.29×10-6,相关系数为0.69,可用于分析区域对流层CO2的时空分布特征.利用2003~2009年SCIAMACHY观测资料分析研究显示:广东地区对流层CO2柱浓度最高值出现在春季,最低值出现在夏季,浓度年均值和年增长率分别为384.84′10-6和1.53′10-6/a,大于全球和我国同期的观测结果;粤东、粤西、粤北和珠三角地区的浓度均在春、冬季显著高于夏季、秋季 ,相同季节内各区域之间的差异不显著;粤西地区CO2柱浓度的年增长率最高,为1.82′10-6/a,珠三角和粤东地区的年增长率相当,分别为1.65,1.64′10-6/a,粤北地区的年增长率最低,为1.61′10-6/a.     

基于机器学习和卫星遥感的PM2.5/10 空间连续分布反演方法研究

在过去的几十年里,快速的经济发展以及工业化、城镇化进程加速使得中国的资源环境承担的压力不断加大。作为影响空气质量的首要污染物,PM_2.5和PM₁₀（记为PM_2.5/10）直接影响着广大人民群众的身体健康。因此,针对PM_2.5/10浓度进行遥感反演研究,对环境监测和控制改善全国空气环境质量具有重要的意义。近些年来,随着对近地面PM_2.5/10浓度研究的不断深入,基于遥感影像数据进行PM_2.5/10浓度的反演方法也日益增多。本文利用Google Earth engine（GEE）平台获取了海量的Landsat 8 OLI遥感影像数据,并结合气象信息、空间特征等参数,采用机器学习中常用的多层映射反向传播神经网络构建了波段反射率与PM_2.5/10浓度之间的反演模型,以获得PM_2.5/10在研究区域的连续分布。为了提高基础PM_2.5/10反演模型的反演精度,还从影响因素和前溯时间两个维度出发,探寻了模型的最优化输入参数组合,并最终实现了对PM_2.5/10浓度的精准反演。以北京市地区为例,模型的PM_2.5和PM₁₀的反演精度R²分别达到0.814和0.796,均方根误差RMSE分别为19.21 μg?m^?3和28.31 μg?m^?3。鉴于该反演结果具有较高的准确性和可靠性,本文所建立的方法模型为研究PM_2.5/10在空间上的连续分布特征提供了新的思路和方法,具有较为重要的科研意义与广泛的应用价值。     

Applications of Very High‐Resolution Imagery in the Study and Conservation of Large Predators in the Southern Ocean

The Southern Ocean is one of the most rapidly changing ecosystems on the planet due to the effects of climate change and commercial fishing for ecologically important krill and fish. Because sea ice loss is expected to be accompanied by declines in krill and fish predators, decoupling the effects of climate and anthropogenic changes on these predator populations is crucial for ecosystem‐based management of the Southern Ocean. We reviewed research published from 2007 to 2014 that incorporated very high‐resolution satellite imagery to assess distribution, abundance, and effects of climate and other anthropogenic changes on populations of predators in polar regions. Very high‐resolution imagery has been used to study 7 species of polar animals in 13 papers, many of which provide methods through which further research can be conducted. Use of very high‐resolution imagery in the Southern Ocean can provide a broader understanding of climate and anthropogenic forces on populations and inform management and conservation recommendations. We recommend that conservation biologists continue to integrate high‐resolution remote sensing into broad‐scale biodiversity and population studies in remote areas, where it can provide much needed detail. Aplicaciones de Imágenes de Muy Alta Resolución en el Estudio y Conservación de Grandes Depredadores en el Océano Antártico     

卫星遥感技术在管理长江流域水资源方面的应用研究

详细讨论了卫星遥感技术在长江流域水资源管理方面的应用问题，提出了两种实用的方法：①卫星遥感技术在灌溉水资源管理方面的应用；②长江流域地下水、地表水资源估算。进一步分析了两种方法的特点，介绍了卫星遥感技术的优点及用途，建议在长江流域水资源管理中加以利用。     

Using Satellite Imagery to Assess Large-Scale Habitat Characteristics of Adirondack Park, New York, USA

Satellite imagery is a useful tool for large-scale habitat analysis; however, its limitations need to be tested. We tested these limitations by varying the methods of a habitat evaluation for white-tailed deer (Odocoileus virginianus) in the Adirondack Park, New York, USA, utilizing harvest data to create and validate the assessment models. We used two classified images, one with a large minimum mapping unit but high accuracy and one with no minimum mapping unit but slightly lower accuracy, to test the sensitivity of the evaluation to these differences. We tested the utility of two methods of assessment, habitat suitability index modeling, and pattern recognition modeling. We varied the scale at which the models were applied by using five separate sizes of analysis windows. Results showed that the presence of a large minimum mapping unit eliminates important details of the habitat. Window size is relatively unimportant if the data are averaged to a large resolution (i.e., township), but if the data are used at the smaller resolution, then the window size is an important consideration. In the Adirondacks, the proportion of hardwood and softwood in an area is most important to the spatial dynamics of deer populations. The low occurrence of open area in all parts of the park either limits the effect of this cover type on the population or limits our ability to detect the effect. The arrangement and interspersion of cover types were not significant to deer populations.     

Effectiveness of Africa's tropical protected areas for maintaining forest cover

The effectiveness of parks for forest conservation is widely debated in Africa, where increasing human pressure, insufficient funding, and lack of management capacity frequently place significant demands on forests. Tropical forests house a substantial portion of the world's remaining biodiversity and are heavily affected by anthropogenic activity. We analyzed park effectiveness at the individual (224 parks) and national (23 countries) level across Africa by comparing the extent of forest loss (as a proxy for deforestation) inside parks to matched unprotected control sites. Although significant geographical variation existed among parks, the majority of African parks had significantly less forest loss within their boundaries (e.g., Mahale Park had 34 times less forest loss within its boundary) than control sites. Accessibility was a significant driver of forest loss. Relatively inaccessible areas had a higher probability (odds ratio >1, p < 0.001) of forest loss but only in ineffective parks, and relatively accessible areas had a higher probability of forest loss but only in effective parks. Smaller parks less effectively prevented forest loss inside park boundaries than larger parks (T = ?2.32, p < 0.05), and older parks less effectively prevented forest loss inside park boundaries than younger parks (F_2,154 = ?4.11, p < 0.001). Our analyses, the first individual and national assessment of park effectiveness across Africa, demonstrated the complexity of factors (such as geographical variation, accessibility, and park size and age) influencing the ability of a park to curb forest loss within its boundaries.