Abstract: There exist noise signals independent of strata properties in most well logs in field oil production. The wavelet transform technique applied for removal treatment of the noise signals is presented based on the feature. Using his techmique mainly including noise-removal meth-ods of dimensional wavelet and wavelet package. a part of noise signal values could be reslored to in-situ origina! ones by density-porosity logs and neutron porosity logs processing. The results show hat the noise curves have been inhibited aler the processing of wavelet lransform. With higher qualily of logs,it allows lo make comprehensive inlerpretation together with rest of well logs, provide more proper prediction of related numbers and locations of gas layers, and grasp reliable bases for natural gas exploitation.

Key words: gas zone, noise, density, neutron, porosity, prediction