中文摘要：

以河北省农林科学院衡水市旱作节水农业实验站（旱作所）H5观测资料为例，利用连续小波变换（CWT）对气压、理论固体潮和浅层地下水位进行了多时间尺度分析，结合小波方差分析确定了气压、固体潮和地下水位等水文时间序列的主要周期（或频率）成分和能量分布；再运用连续小波逆变换在主要周期尺度上重构了气压、固体潮和地下水位时间序列，用最小二乘法求出了水位对不同频率气压和固体潮的敏感程度（即响应系数），依此消除气压和固体潮对观测水位的影响；最后通过极差和标准差评判改正效果。结果表明：旱作所H5观测孔水位受气压影响较大，气压系数为0．799-0．881cm／HPa；对固体潮的响应较小，固体潮系数均接近于0cm／μGal；响应系数的频率效应不明显；与采用一般线性回归方法和快速Mallat离散小波变换算法相比，CWT方法所得气压系数和固体潮系数的地球物理意义更明确；3种计算方法得到的水位改正结果显示，CWTT法与Mallat算法接近．优于一般线性回归方法.

英文摘要：

Based on the hourly data at the Dry-Land Farming Institute （DFI）, using the algorithms of CWT, wavelet variance and LSM, multi-time scale features of barometric pressure, solid tide and shallow groundwater level were analyzed. Barometric coefficients and Earth-tide coefficients were calculated at several main time-scales and were used to correct the water level. Then we used the statistical properties of range and standard deviation to evaluate the correction effect. The results show that the range of baromet ric coefficients varies from 0. 799 to 0. 881 cm/HPa, and the Earth-tidal coefficients are all close to 0 cm/ μGal. It is concluded that the effect of barometric pressure on water level is more significant than that of Earth-tides, and the response coefficients are affected rarely by frequency. Compared with either Mallat or linear regression algorithms, the geophysical significance of both barometric and Earth-tide coefficients cal culated by the CWT algorithm are more obvious. The statistical results of water level corrections show that the CWT algorithm is feasible and effective to eliminate the responses of barometric pressure and Earth tide from water level fluctuation.