Relationship of underground water level and climate in Northwest China’s inland basins under the global climate change:Taking the Golmud River Catchment as an example

To identify the response of groundwater level variation to global climate change in Northwest China’s inland basins,the Golmud River Catchment was chosen as a case in this paper.Approaches of time series analysis and correlation analysis were adopted to investigate the variation of groundwater level influenced by global climate change from 1977 to 2017.Results show that the temperature in the Golmud River Catchment rose 0.57℃ every 10 years.It is highly positive correlated with global climate temperature,with a correlation coefficient,0.87.The frequency and intensity of extreme precipitation were both increased.Generally,groundwater levels increased from 1977 to 2017 in all phreatic and confined aquifers and the fluctuation became more violent.Most importantly,extreme precipitation led to the fact that groundwater level rises sharply,which induced city waterlogging.However,no direct evidence shows that normal precipitation triggered groundwater level rise,and the correlation coefficients between precipitation data from Golmud meteorological station located in the Gobi Desert and groundwater level data of five observation wells are 0.13,0.02,−0.11,0.04,and−0.03,respectively.This phenomenon could be explained as that the main recharge source of groundwater is river leakage in the alluvial-pluvial Gobi plain because of the high total head of river water and goodness hydraulic conductivity of the vadose zone.Data analysis shows that glacier melting aggravated because of local temperature increased.As a result,runoff caused groundwater levels to ascend from 1977 to 2017.Correlation coefficients of two groundwater wells observation data and runoff of Golmud River are 0.80 and 0.68.The research results will contribute to handling the negative effects of climate change on groundwater for Northwestern China.

三维水平集智能分割算法在梅尼埃病患者内耳道磁共振图像分割中的应用效果

目的探讨三维水平集智能分割(IS3DLS)算法对梅尼埃病患者内耳道磁共振成像(MRI)图像的分割效果。方法选择68例梅尼埃病患者,采用MRI系统进行内耳检查,对内耳MRI图像分别采取专家手工分割算法、区域生长水平集分割算法、IS3DLS算法进行图像分割处理。①手工分割算法:由专家直接在MRI原始图像上画出感兴趣区(ROI)的边缘,分割目标。②区域生长水平集分割方法:在内耳ROI内设置种子点,调至最优参数,获得最佳区域生长分割结果,去除无关的组织,获得预分割结果作为初始轮廓掩膜,水平集演化分割内耳。③任选28组内耳图像数据,由耳科专家手动分割获得三维内耳体素模型,用3D-Slicer软件获得形状训练样本。另取6组内耳图像数据作为测试样本,建立内耳统计形状模型,计算平均形状模型。对图像进行加权,采用配准技术自动定位和获取内耳ROI,对内耳ROI和平均形状模型进行配准,得到初始轮廓。采用IS3DLS对初始轮廓和特征图像进行分割。以专家手工分割算法为标准,采用马修斯相关系数(MCC)、戴斯相似系数(DSC)、假阳性率(FPR)和假阴性率(FNR)计算IS3DLS算法、区域生长水平集分割算法与专家手工分割算法的相似性和准确性。以真位置(TP)和假位置(FP)进行IS3DLS算法、区域生长水平集分割算法两种分割算法的误差比较,记录IS3DLS和区域生长集分割算法的运行时间并比较。结果IS3DLS算法的MCC、DSC、FPR、FNR分别为0.9599、0.9594、0.0325、0.0365,与专家手工分割算法相当(P<0.05);与专家手工分割算法相比,IS3DLS算法的MCC高于区域生长集分割算法,提示其与专家手工分割算法的结果更接近;IS3DLS算法的FPR、FNR均低于区域生长集分割算法;IS3DLS算法和区域生长集分割算法均具有较高的TP值;IS3DLS算法的FP值、HD值均小于区域生长集分割算法(P<0.05);IS3DLS的运行时间明显少于区域生长集分割(P<0.01)。结论IS3DLS算法对梅尼埃病患者的内耳道MRI图像有较好的分割效果,处理时间更短,有助于梅尼埃病的诊断和后续治疗。