大西洋副极地海域西边界变化与翻转环流的关系特征

正大西洋经向翻转环流(Atlantic meridional overturing circulation,AMOC)是全球环流系统的重要组分,其基本结构可简单地刻画为上下两层.AMOC上层分支将海洋表层高温、高盐的海水向极地方向输送,密度相对较低的海水在北大西洋副极地海域的强烈海气交互作用下,下沉形成北大西洋深层水,继而在大西洋中深层向赤道方向流动,构成AMOC下层分支.然而海洋中的实际情况要复杂得多.传统观点基于理论和气候模式,认为位于北大西洋副极地的拉布拉多海是驱动AMOC变异的关键海域;     

青藏高原对北大西洋深水形成影响机制的季节差异

利用耦合地球系统模式(CESM1.0), 通过对比有青藏高原的控制试验和无青藏高原的敏感性试验, 定性地分析青藏高原对不同季节北大西洋深水(NADW)形成的影响机制。研究结果表明, 青藏高原对NADW形成的影响机制因季节而异。移除青藏高原后, NADW形成将会减弱, 冷季(北半球10月—次年3月) NADW形成的显著减弱是由海洋表面净热通量显著增加引起的, 而暖季(4—9月)由于海冰大量融化引起淡水通量显著增加, 导致NADW形成的显著减弱。低分辨率耦合模式中, 青藏高原对NADW形成的这种季节性影响更加显著。不同季节青藏高原对NADW形成的影响机制主要区别在于, 青藏高原地形的存在使得冷季NADW形成区域的海洋表面失热增加, 在暖季造成该区域来自北部海冰的淡水输入减少。     

青藏高原对非洲北部降水影响的模拟研究

利用耦合模式CESM1.0, 研究青藏高原地形对非洲北部降水的影响。敏感性试验结果表明, 去掉青藏高原地形后, 首先, 大气环流迅速做出调整, 出现自热带大西洋向东北方向至北非的水汽输送异常和自印度洋向西至北非的水汽输送异常, 造成北非大气水汽含量增加和水汽辐合增强, 降水增多。然后, 当海洋环流调整到准平衡态时, 北大西洋海表温度降低, 南大西洋海表温度升高, 地表大气温度也发生相应的变化。在南北温度梯度的影响下, 原本由热带大西洋向北非的水汽输送发生转向, 导致北非的水汽含量减少和水汽辐合减弱, 使得降水比前一阶段减少。即便如此, 在没有青藏高原的试验中, 当海洋环流调整到平衡态时, 北非大部分区域水汽辐合仍然强于有青藏高原的真实地形试验, 区域平均降水也增多。结果表明, 青藏高原的隆升可能在一定程度上加剧了北非的干旱化。     

大西洋热盐环流减弱对热带太平洋气候平均态及年际变率的影响

利用一个完全耦合的海气模式, 通过对比分析两组试验中海表温度、盐度、风应力等气候态变化特征以及ENSO强度和频率的变化, 研究热带太平洋气候平均态及年际变率对热盐环流减弱的响应。在北大西洋高纬地区注入1 Sv淡水后, 大西洋经向翻转流(AMOC)减弱约90%, 这直接导致向北的经向热量输送减少, 使北大西洋有明显降温, 南大西洋略有升温。这些变化会经过大气和海洋的远程传播以及局地海气反馈作用, 影响热带太平洋气候平均态: 赤道东西太平洋的SST都略有增温, 但纬向温度梯度和纬向风应力并没有太大变化, 赤道太平洋温跃层的深度和倾斜度也基本保持不变。相应地, ENSO强度和频率也没有明显变化。由此得出结论: 热盐环流减弱会引起全球气候平均态的变化, 但对热带太平洋的年际变率没有太大影响。     

一种基于金字塔策略的人脸超分辨率方法

高分辨率预期图像与低分辨率输入图像分别位于高斯金字塔的第0层和第i层.超分辨率过程据此被分成i-1个子过程,逐层进行,即将第k层低分辨率图像的超分辨率估计图像(位于第k-1层)作为第k-2层高分辨率预期图像的输入,直至产生第0层高分辨率估计图像为止.在每个超分辨率子过程中,采用基于patch的搜索策略.文中对这种基于金字塔策略的人脸超分辨率方法进行了实验,结果表明该方法能够达到很好的超分辨率效果.     

不受DEM空间分辨率影响的地形指数计算

为了减小数字高程模型(DEM)分辨率对地形指数计算的影响,开发了一种不受DEM空间分辨率影响的地形指数计算方法.选取汉江上游褒河流域作为研究区域,通过引入分辨率因子考虑DEM空间分辨率对单宽上坡集水面积的影响,应用分形方法处理DEM空间尺度对地形坡度的影响,将该算法应用到褒河流域DEM数据的计算中.结果表明:此法可以成功地由低分辨率的DEM数据获得与目标高分辨率精度相仿的地形指数分布,能在一定程度上消除DEM空间分辨率对地形指数计算的影响,因而对无资料地区的水文预报有着十分重要的现实意义.     

基于半耦合字典的快速图像超分辨率重建算法

为了提高图像超分辨率重建的效率与质量,考虑到高、低分辨率稀疏表示系数的不同,改进了锚定邻域回归算法,并结合半耦合字典学习算法提出了一种快速图像超分辨率重建算法.首先采用半耦合字典学习算法得到高分辨率字典、低分辨率字典及映射矩阵;再采用岭回归算法求解低分辨率稀疏表示系数,并根据高分辨率稀疏表示系数与低分辨率稀疏表示系数之间的映射关系,得到高分辨率稀疏表示系数;然后,根据输入图像块特征寻找字典中与其最相关的字典原子,计算该字典原子所对应的投影矩阵,进行超分辨率重建.仿真结果表明:提出的算法不仅在重建速度上表现更快,重建图像的质量也得到提高,在客观指标和主观效果上均取得更好的效果.     

对流解析区域气候模式对青藏高原降水模拟能力的研究

以中国科学院区域气候-环境重点实验室研制的区域环境集成系统模式(RIEMS 2.0)为基础,采用中国科学院资源与环境数据中心提供的植被类型数据和北京师范大学提供的中国土壤质地数据,以及美国地质调查局提供的月植被覆盖度资料,进行模式本地化,从而建成了青藏高原对流解析区域气候模式．利用该模式对青藏高原进行了2001—2018年连续积分模拟,重点考察了区域气候模式在水平分辨率为9 km条件下对青藏高原降水模拟能力,结果表明:1）模式能够较好地模拟年、雨季降水的空间分布特征以及不同区域降水年变化,同时,模式模拟降水较观测偏多,偏差为13.01%~39.95%;区域气候模式模拟青藏高原降水较国际耦合模式“比较计划第六阶段（CMIP6）”45个全球模式模拟试验结果的年降水空间分布和强度有明显提高,并且更加接近观测值．2）模式能够较好地模拟出年降水时间和4个不同等级降水事件空间分布,特别是5~10、10~20、>20 mm这3个不同等级降水时间接近观测值．3）模式能够较好地模拟出青藏高原不同区域候平均降水随时间演变,降水强度除半干旱藏南地区较观测明显偏多外,对其他地区模式模拟的降水都非常接近观测值,同时与观测值之间相关系数为0.901~0.981,都通过99%置信度检验,与观测值之间的均方根误差为0.37~0.99 mm·d?1,其中对于极度干旱的柴达木地区候平均降水也能够较好地模拟出来,相关系数达到0.919;对青藏高原西南的南羌塘地区模拟最好,相关系数达到0.981．4）该研究表明采用青藏高原对流解析区域气候模式进行动力降尺度后,解决青藏高原等地区缺乏长时间序列高时空分辨率的气象数据集的瓶颈问题,为青藏高原气候和环境未来变化、生态安全屏障建设等提供坚实可靠的科学数据基础．      

基于学习的人脸图像超分辨率重构算法

提出了一种新的基于学习的人脸图像超分辨率重构算法,利用高分辨率图像和低分辨率图像的拓扑结构相似性,将现有的低分辨率人脸图像在低分辨率人脸图像字典中展开,在保持系数不变的同时将字典换为高分辨率人脸图像字典,最终得到待重构的高分辨率人脸图像.在系数估计时,使用主成分分析的方法,同时加入了最小全变分作为约束,算法充分利用了不同人脸图像之间的相似性和人脸图像本身的内部相关性.实验结果表明,结果既保持了对原有图像的忠实性,又比较适合人眼观察.     

基于位置约束的参考图引导的遥感影像超分辨率网络

在遥感对地观测领域中,往往需要对生态或地质环境进行长时间跨度的监测.然而,仅通过卫星拍摄的高分辨率影像来实现在长时间跨度范围内对同一位置区域的覆盖是较为困难的.因此,之前的研究多采用单图像超分辨率算法来对低分辨率影像进行重建.但单张低分辨率影像中的信息是有限的,所以该类算法的重建结果往往较为平滑和模糊.实际上,对于同一位置区域而言,通常可以获取到来自不同卫星在不同时间段拍摄的多幅分辨率不同的遥感影像.其中的高分辨率影像可以作为低分辨输入在重建过程中的参考影像.因此,提出了一种基于位置约束的参考图引导的遥感影像超分辨率网络.该网络首先通过一个基于位置编码的纹理转换模块来实现参考图像与低分辨率图像之间的相似特征块匹配.同时引入位置约束,即根据低分辨率输入中邻近元素所对应的参考相似块的聚合度来对转换后的参考图像进行编码,从而提高匹配的准确度.同时,还提出了一种基于通道注意力机制的多尺度特征自适应融合模块,该模块有效提高了网络在进行跨尺度特征融合时的特征表达能力,进一步提高了网络的重建性能.实验结果表明,相较于其他新近的超分辨率方法,该网络在4×和8×超分辨率任务上均表现出了更为优越的重建性能...     

Absence of deep-water formation in the Labrador Sea during the last interglacial period

The two main constituent water masses of the deep North Atlantic Ocean-North Atlantic Deep Water at the bottom and Labrador Sea Water at an intermediate level-are currently formed in the Nordic seas and the Labrador Sea, respectively. The rate of formation of these two water masses tightly governs the strength of the global ocean circulation and the associated heat transport across the North Atlantic Ocean. Numerical simulations have suggested a possible shut-down of Labrador Sea Water formation as a consequence of global warming. Here we use micropalaeontological data and stable isotope measurements in both planktonic and benthic foraminifera from deep Labrador Sea cores to investigate the density structure of the water column during the last interglacial period, which was thought to be about 2 degrees C warmer than present. Our results indicate that today's stratification between Labrador Sea Water and North Atlantic Deep Water never developed during the last interglacial period. Instead, a buoyant surface layer was present above a single water mass originating from the Nordic seas. Thus the present situation, with an active site of intermediate-water formation in the Labrador Sea, which settled some 7,000 years ago, has no analogue throughout the last climate cycle.     

Deep convection in the Irminger Sea forced by the Greenland tip jet

Open-ocean deep convection, one of the processes by which deep waters of the world's oceans are formed, is restricted to a small number of locations (for example, the Mediterranean and Labrador seas). Recently, the southwest Irminger Sea has been suggested as an additional location for open-ocean deep convection. The deep water formed in the Irminger Sea has the characteristic temperature and salinity of the water mass that fills the mid-depth North Atlantic Ocean, which had been believed to be formed entirely in the Labrador basin. Here we show that the most likely cause of the convection in the Irminger Sea is a low-level atmospheric jet known as the Greenland tip jet, which forms periodically in the lee of Cape Farewell, Greenland, and is associated with elevated heat flux and strong wind stress curl. Using a history of tip-jet events derived from meteorological land station data and a regional oceanic numerical model, we demonstrate that deep convection can occur in this region when the North Atlantic Oscillation Index is high, which is consistent with observations. This mechanism of convection in the Irminger Sea differs significantly from those known to operate in the Labrador and Mediterranean seas.     

Mid-depth recirculation observed in the interior Labrador and Irminger seas by direct velocity measurements

The Labrador Sea is one of the sites where convection exports surface water to the deep ocean in winter as part of the thermohaline circulation. Labrador Sea water is characteristically cold and fresh, and it can be traced at intermediate depths (500-2,000 m) across the North Atlantic Ocean, to the south and to the east of the Labrador Sea. Widespread observations of the ocean currents that lead to this distribution of Labrador Sea water have, however, been difficult and therefore scarce. We have used more than 200 subsurface floats to measure directly basin-wide horizontal velocities at various depths in the Labrador and Irminger seas. We observe unanticipated recirculations of the mid-depth (approximately 700 m) cyclonic boundary currents in both basins, leading to an anticyclonic flow in the interior of the Labrador basin. About 40% of the floats from the region of deep convection left the basin within one year and were rapidly transported in the anticyclonic flow to the Irminger basin, and also eastwards into the subpolar gyre. Surprisingly, the float tracks did not clearly depict the deep western boundary current, which is the expected main pathway of Labrador Sea water in the thermohaline circulation. Rather, the flow along the boundary near Flemish Cap is dominated by eddies that transport water offshore. Our detailed observations of the velocity structure with a high data coverage suggest that we may have to revise our picture of the formation and spreading of Labrador Sea water, and future studies with similar instrumentation will allow new insights on the intermediate depth ocean circulation.     

青藏高原对全球大气温度和水汽分布的影响

利用耦合的气候模式CESM, 定量研究青藏高原对全球大气温度和水汽分布的影响。通过对比采用真实地形的参考实验(Real)和去掉青藏高原的敏感性实验(NoTibet)发现, 去掉青藏高原会使北半球大气变冷、变干, 对南半球的影响不明显。北半球中高纬度从地表至平流层均有强烈降温, 地表的降温中心在北大西洋, 年平均降温幅度达5ºC, 高空的降温中心在100 hPa的平流层, 年平均降温幅度达2ºC。北大西洋和南亚地区湿度减少, 南大西洋和东非地区湿度增加。北半球变冷主要是海洋向北经向热量输送减少的结果, 一方面增强了北半球的经向温度梯度, 导致Hadley环流增强, 加强了中低纬地区向北的大气热量输送, 部分补偿了海洋向北减少的热量输送, 维持了北半球中低纬度的能量平衡; 另一方面, 使得北半球中高纬度蒸发作用减弱, 大气中水汽含量减少, 北半球变得寒冷干燥。初步的研究表明, 青藏高原对北半球气候有重大影响, 影响范围可达北半球高纬度地区。     

Evolution of the East Asian monsoon and its response to uplift of the Tibetan Plateau since 1.8 Ma recorded by major elements in sediments of the South China Sea

Evolution of the East Asian monsoon and its response to uplift of the Tibetan Plateau has been investigated in the study of global change. Core sediment samples drilled in the South China Sea during ODP Leg184 are the best materials for studying long-term variability of the East Asian monsoon. R-mode factor analysis of major elements in the fine grain-sized carbonate-free sediments (<4 μm) of the upper 185 mcd splice of ODP Site 1146 drilled during Leg184 in the South China Sea shows that Ti, TFe2O3, MgO, K2O, P, CaO, and Al2O3 are representative of a terrestrial factor. The variation in the terrestrial factor score is subject to chemical erosion in the source region and thus indicates the evolution of the East Asian summer monsoon. The terrestrial factor score has three stepwise decreases at ~1.3 Ma, ~0.9 Ma, and ~0.6 Ma, indicating the phased weakening of the East Asian summer monsoon is related to wholly stepwise, quick uplifts of the Tibetan Plateau since 1.8 Ma. The periodic fluctuation of the terrestrial factor score since ~0.6 Ma indicates that the glacial-interglacial cycles have been the main force driving the evolution of the East Asian monsoon. As in the case of Chinese loess, the long-term evolution of the East Asian monsoon recorded in sediments of the South China Sea reflects a coupled effect of the glacial-interglacial cycle and uplift of the Tibetan Plateau.     

Peat record reflecting Holocene climatic change in the Zoige Plateau and AMS radiocarbon dating

Through the use of reliable AMS dating of high resolution (15-30 years) peat and the establishment of monsoon climate proxies sequence, we have been able to recognize several cold, dry events in the Tibetan Plateau during the Holocene. The more obvious ones occurred around 12800, 11300, 10200, 9580, 8900, 6400, 4400, 3700, 2800 and 1500 cal. aBP. These events correlate well with both ice rafting events recorded in high latitude North Atlantic Ocean sediment cores and cooling events in the low latitude SST. Spectral analysis indicates high frequency climate variation on centennial-millennial time scale during the Holocene. This further reflects Holocene climate instability and the existence of centennial-millenium scale rhythm in mid latitude areas as well.     

Peat record reflecting Holocene climatic change in the Zoigê Plateau and AMS radiocarbon dating

Through the use of reliable AMS dating of high resolution (15–30 years) peat and the establishment of monsoon climate proxies sequence, we have been able to recognize several cold, dry events in the Tibetan Plateau during the Holocene. The more obvious ones occurred around 12800, 11300, 10200, 9580, 8900, 6400, 4400, 3700, 2800 and 1500 cal. aBP. These events correlate well with both ice rafting events recorded in high latitude North Atlantic Ocean sediment cores and cooling events in the low latitude SST. Spectral analysis indicates high frequency climate variation on centennial-millennial time scale during the Holocene. This further reflects Holocene climate instability and the existence of centennial-millenium scale rhythm in mid latitude areas as well.