Detecting near-surface soil freeze-thaw cycles in high-altitude cold regions is important for understanding the Earth＇s surface system, but such studies are rare. In this study, we detected the spatial-temporal varia...Detecting near-surface soil freeze-thaw cycles in high-altitude cold regions is important for understanding the Earth＇s surface system, but such studies are rare. In this study, we detected the spatial-temporal variations in near-surface soil freeze-thaw cycles in the source region of the Yellow River（SRYR） during the period 2002–2011 based on data from the Advanced Microwave Scanning Radiometer for the Earth Observing System（AMSR-E）. Moreover, the trends of onset dates and durations of the soil freeze-thaw cycles under different stages were also analyzed. Results showed that the thresholds of daytime and nighttime brightness temperatures of the freeze-thaw algorithm for the SRYR were 257.59 and 261.28 K, respectively. At the spatial scale, the daily frozen surface（DFS） area and the daily surface freeze-thaw cycle surface（DFTS） area decreased by 0.08% and 0.25%, respectively, and the daily thawed surface（DTS） area increased by 0.36%. At the temporal scale, the dates of the onset of thawing and complete thawing advanced by 3.10（±1.4） and 2.46（±1.4） days, respectively; and the dates of the onset of freezing and complete freezing were delayed by 0.9（±1.4） and 1.6（±1.1） days, respectively. The duration of thawing increased by 0.72（±0.21） day/a and the duration of freezing decreased by 0.52（±0.26） day/a. In conclusion, increases in the annual minimum temperature and winter air temperature are the main factors for the advanced thawing and delayed freezing and for the increase in the duration of thawing and the decrease in the duration of freezing in the SRYR.展开更多
[目的]提高近红外光谱技术在线检测柚子糖度的精度。[方法]采用自主研发的柚子在线无损检测设备采集3种光照区域的柚子的漫透射光谱数据,在650~950 nm的波长范围内采用标准正交变量变换(SNV)、多元散射校正(MSC)、归一化(normalize)、S...[目的]提高近红外光谱技术在线检测柚子糖度的精度。[方法]采用自主研发的柚子在线无损检测设备采集3种光照区域的柚子的漫透射光谱数据,在650~950 nm的波长范围内采用标准正交变量变换(SNV)、多元散射校正(MSC)、归一化(normalize)、SG一阶求导(savitzky-golay first order derivative,SG-1st)对原始数据进行预处理,使用自适应性加权算法(CARS)筛选反映柚子糖度的光谱特征,建立了偏最小二乘回归(PLSR)模型。使用未参与到建模的30个柚子样本进行在线验证。[结果]光照区域C结合SNV-CARS-PLSR方法的建模效果最优。其预测集的决定系数为0.95,均方根误差为0.30°Brix。在线验证时决定系数为0.90,均方根误差为0.58°Brix。模型对于柚子糖度具有较强的在线检测能力。[结论]在光斑直径为70 mm且位于柚子赤道上方20 mm的光照区域C的条件下采集的柚子光谱数据所建立的预测模型能更有效地实现柚子糖度的在线预测。展开更多
基金supported by the National Science and Technology Support Plan of China (2015BAD07B02)
文摘Detecting near-surface soil freeze-thaw cycles in high-altitude cold regions is important for understanding the Earth＇s surface system, but such studies are rare. In this study, we detected the spatial-temporal variations in near-surface soil freeze-thaw cycles in the source region of the Yellow River（SRYR） during the period 2002–2011 based on data from the Advanced Microwave Scanning Radiometer for the Earth Observing System（AMSR-E）. Moreover, the trends of onset dates and durations of the soil freeze-thaw cycles under different stages were also analyzed. Results showed that the thresholds of daytime and nighttime brightness temperatures of the freeze-thaw algorithm for the SRYR were 257.59 and 261.28 K, respectively. At the spatial scale, the daily frozen surface（DFS） area and the daily surface freeze-thaw cycle surface（DFTS） area decreased by 0.08% and 0.25%, respectively, and the daily thawed surface（DTS） area increased by 0.36%. At the temporal scale, the dates of the onset of thawing and complete thawing advanced by 3.10（±1.4） and 2.46（±1.4） days, respectively; and the dates of the onset of freezing and complete freezing were delayed by 0.9（±1.4） and 1.6（±1.1） days, respectively. The duration of thawing increased by 0.72（±0.21） day/a and the duration of freezing decreased by 0.52（±0.26） day/a. In conclusion, increases in the annual minimum temperature and winter air temperature are the main factors for the advanced thawing and delayed freezing and for the increase in the duration of thawing and the decrease in the duration of freezing in the SRYR.
文摘[目的]提高近红外光谱技术在线检测柚子糖度的精度。[方法]采用自主研发的柚子在线无损检测设备采集3种光照区域的柚子的漫透射光谱数据,在650~950 nm的波长范围内采用标准正交变量变换(SNV)、多元散射校正(MSC)、归一化(normalize)、SG一阶求导(savitzky-golay first order derivative,SG-1st)对原始数据进行预处理,使用自适应性加权算法(CARS)筛选反映柚子糖度的光谱特征,建立了偏最小二乘回归(PLSR)模型。使用未参与到建模的30个柚子样本进行在线验证。[结果]光照区域C结合SNV-CARS-PLSR方法的建模效果最优。其预测集的决定系数为0.95,均方根误差为0.30°Brix。在线验证时决定系数为0.90,均方根误差为0.58°Brix。模型对于柚子糖度具有较强的在线检测能力。[结论]在光斑直径为70 mm且位于柚子赤道上方20 mm的光照区域C的条件下采集的柚子光谱数据所建立的预测模型能更有效地实现柚子糖度的在线预测。
