Global solar radiation (GSR) is an essential physical quantity for agricultural management and designing infrastructures. Because GSR has often been modeled as a function of sunshine duration (SD) and day length for a...Global solar radiation (GSR) is an essential physical quantity for agricultural management and designing infrastructures. Because GSR has often been modeled as a function of sunshine duration (SD) and day length for a given set of locations and calendar days, analyzing interannual trends in GSR and SD is important to evaluate, predict or regulate the cycles of energy and water between geosphere and atmosphere. This study aimed to exemplify interannual trends in GSR and SD, which had been recorded from 2001 to 2022 in 40 meteorological stations in Japan, and validate the applicability of an SD-based model to the evaluation of GSR. Both the measured GSR and SD had increased in many of the stations in the study period with averaged rates of 0.252 [W·m−2·y−1] and 0.015 [h·d−1·y−1], respectively. The offset and the slope of the SD-based model were estimated by fitting the model to the measured data sets and were found to have been almost constant with the averages of 0.201[-] and 0.566[-], respectively, indicating that characteristics of the SD-GSR relation had not varied for the 22-year period and that the model and its parameter set can be stationarily applicable to the analyses and predictions of GSR in recent years. The stable trends in both parameters also implied that the upward trend in SD can be a main explanatory factor for that in the measured GSR. The upward trend in SD had coincided with the increase in the frequency of heavy-shortened rains, suggesting that the time period of each rainfall event had gradually decreased, which may be attributable to the obtained upward trend in SD. Further studies are required to clarify if there is some cause-effect relation between the changes in rainfall patterns and the standard level of solar radiation reaching the land surface.展开更多
对比仅包含多光谱信息、仅可实现二维土地覆盖分类的传统光学遥感数据,机载多光谱激光雷达(multispectral light detection and ranging,MS-LiDAR)的优势在于同时包含多光谱和空间信息、可实现三维土地覆盖分类,但现有的机载MS-LiDAR数...对比仅包含多光谱信息、仅可实现二维土地覆盖分类的传统光学遥感数据,机载多光谱激光雷达(multispectral light detection and ranging,MS-LiDAR)的优势在于同时包含多光谱和空间信息、可实现三维土地覆盖分类,但现有的机载MS-LiDAR数据的土地覆盖分类研究所需特征维度过高、算法复杂度高。因此,提出了一种整合空间相关性和归一化差分比率指数(Normalized Difference Ratio Index,NDRI)特征的逐步分类算法。该算法首先融合机载MS-LiDAR数据的多波段独立点云,获取兼具空间位置及其多光谱信息的单一点云数据;然后利用空间邻域增长下的地面滤波算法分离地面和非地面点;接着基于不同目标的激光反射特性差异设计将草地(树木)自地面(非地面)中分离的NDRI指数,并利用类间方差最大原则下的自适应最优NDRI指数实现地面和非地面点的精细分类;最后利用3D多数投票法优化分类结果。采用加拿大Optech Titan实测MS-LiDAR数据测试提出算法的有效性及可行性,实验结果表明:算法的平均总体精度和Kappa系数分别可达90.17%和0.861,可有效实现城区MS-LiDAR数据的三维土地覆盖分类;分步处理的方式更有利于针对具体的分离目标的特点设计简单且有效的规则,算法设计更简单、复杂度低;NDRI可为其他机器学习算法的显著性特征的设计和选择提供理论支撑。展开更多
利用2007—2016年国际卫星云气候计划(International Satellite Cloud Climatology Project,ISCCP)、云和地球辐射能量系统(Clouds and the Earth s Radiant Energy System,CERES)和中分辨率成像光谱仪(Moderate Resolution Imaging Spe...利用2007—2016年国际卫星云气候计划(International Satellite Cloud Climatology Project,ISCCP)、云和地球辐射能量系统(Clouds and the Earth s Radiant Energy System,CERES)和中分辨率成像光谱仪(Moderate Resolution Imaging Spectroradiometer,MODIS)卫星反演云产品,对比分析了不同数据反演的中国地区云系结构的宏微观特征,并采用复合评价指标定量评估了不同数据之间时间和空间上的一致性。结果表明:三套卫星数据都能较好地反演出中国地区总云量呈南高北低、东高西低、夏高冬低的分布特征,但通过比较时间技巧(Temporal Skill,S_(T))及空间技巧(Spatial Skill,S_(S))复合评价指标及其各项分量发现,与MODIS相比,CERES与ISCCP数据反演的总云量时间序列演变特征明显更为一致,且其评分均有南方优于北方,夏季优于冬季的特征;进一步分析不同高度云量的S_(T)评分发现,CERES和ISCCP两套数据在南方地区的总云量差异主要来自于低云量的绝对偏差,而北方地区的偏差则同时存在于低云和中云;对比分析MODIS和CERES反演的云滴有效半径发现,高云对应的冰相云一致性较高,而中低云相对应的液相云的偏差则有夏季高于冬季的规律。针对夏季液相和冰相云滴粒径及概率密度分析则表明,相比CERES数据,MODIS对夏季液水和冰水粒子的有效半径在不同地区均有不同程度的高估,液(冰)水谱宽则更宽(窄)。展开更多
文摘Global solar radiation (GSR) is an essential physical quantity for agricultural management and designing infrastructures. Because GSR has often been modeled as a function of sunshine duration (SD) and day length for a given set of locations and calendar days, analyzing interannual trends in GSR and SD is important to evaluate, predict or regulate the cycles of energy and water between geosphere and atmosphere. This study aimed to exemplify interannual trends in GSR and SD, which had been recorded from 2001 to 2022 in 40 meteorological stations in Japan, and validate the applicability of an SD-based model to the evaluation of GSR. Both the measured GSR and SD had increased in many of the stations in the study period with averaged rates of 0.252 [W·m−2·y−1] and 0.015 [h·d−1·y−1], respectively. The offset and the slope of the SD-based model were estimated by fitting the model to the measured data sets and were found to have been almost constant with the averages of 0.201[-] and 0.566[-], respectively, indicating that characteristics of the SD-GSR relation had not varied for the 22-year period and that the model and its parameter set can be stationarily applicable to the analyses and predictions of GSR in recent years. The stable trends in both parameters also implied that the upward trend in SD can be a main explanatory factor for that in the measured GSR. The upward trend in SD had coincided with the increase in the frequency of heavy-shortened rains, suggesting that the time period of each rainfall event had gradually decreased, which may be attributable to the obtained upward trend in SD. Further studies are required to clarify if there is some cause-effect relation between the changes in rainfall patterns and the standard level of solar radiation reaching the land surface.
文摘对比仅包含多光谱信息、仅可实现二维土地覆盖分类的传统光学遥感数据,机载多光谱激光雷达(multispectral light detection and ranging,MS-LiDAR)的优势在于同时包含多光谱和空间信息、可实现三维土地覆盖分类,但现有的机载MS-LiDAR数据的土地覆盖分类研究所需特征维度过高、算法复杂度高。因此,提出了一种整合空间相关性和归一化差分比率指数(Normalized Difference Ratio Index,NDRI)特征的逐步分类算法。该算法首先融合机载MS-LiDAR数据的多波段独立点云,获取兼具空间位置及其多光谱信息的单一点云数据;然后利用空间邻域增长下的地面滤波算法分离地面和非地面点;接着基于不同目标的激光反射特性差异设计将草地(树木)自地面(非地面)中分离的NDRI指数,并利用类间方差最大原则下的自适应最优NDRI指数实现地面和非地面点的精细分类;最后利用3D多数投票法优化分类结果。采用加拿大Optech Titan实测MS-LiDAR数据测试提出算法的有效性及可行性,实验结果表明:算法的平均总体精度和Kappa系数分别可达90.17%和0.861,可有效实现城区MS-LiDAR数据的三维土地覆盖分类;分步处理的方式更有利于针对具体的分离目标的特点设计简单且有效的规则,算法设计更简单、复杂度低;NDRI可为其他机器学习算法的显著性特征的设计和选择提供理论支撑。
文摘利用2007—2016年国际卫星云气候计划(International Satellite Cloud Climatology Project,ISCCP)、云和地球辐射能量系统(Clouds and the Earth s Radiant Energy System,CERES)和中分辨率成像光谱仪(Moderate Resolution Imaging Spectroradiometer,MODIS)卫星反演云产品,对比分析了不同数据反演的中国地区云系结构的宏微观特征,并采用复合评价指标定量评估了不同数据之间时间和空间上的一致性。结果表明:三套卫星数据都能较好地反演出中国地区总云量呈南高北低、东高西低、夏高冬低的分布特征,但通过比较时间技巧(Temporal Skill,S_(T))及空间技巧(Spatial Skill,S_(S))复合评价指标及其各项分量发现,与MODIS相比,CERES与ISCCP数据反演的总云量时间序列演变特征明显更为一致,且其评分均有南方优于北方,夏季优于冬季的特征;进一步分析不同高度云量的S_(T)评分发现,CERES和ISCCP两套数据在南方地区的总云量差异主要来自于低云量的绝对偏差,而北方地区的偏差则同时存在于低云和中云;对比分析MODIS和CERES反演的云滴有效半径发现,高云对应的冰相云一致性较高,而中低云相对应的液相云的偏差则有夏季高于冬季的规律。针对夏季液相和冰相云滴粒径及概率密度分析则表明,相比CERES数据,MODIS对夏季液水和冰水粒子的有效半径在不同地区均有不同程度的高估,液(冰)水谱宽则更宽(窄)。
