Drilling operations in polar regions and mountainous areas are complicated by nature of the extreme environment． Yet conventional rotary drilling technologies can be used to drill ice for scientific samples and oth- er research． Due to such reasons as power consumption and weight complications，it is hard to apply a conven- tional rotary drilling rig for glacial exploration． Use of small，relatively lightweight，portable engine- powered drilling systems in which the drill lifting from the borehole is carried by the winch． It is reasonable enough for near- surface shallow ice- drilling down to 50 m． Such systems can be used for near- surface ablation- stakes in- stallation，also temperature measurements at the bottom of active strata layer，revealing of anthropogenic pollu- tion，etc． The specified used in this research is an auger ice drill powered by a gasoline engine． At this stage， it is crucial to choose effective drilling parameters such as weight on bit (WOB) and drill bit rotation rate． Sen- sors equipped on the rig have measured the main parameters of the drilling process，such as drill speed，WOB， drill rotation speed，torque and temperature． This paper addresses research on drilling parameters of engine powered auger ice drill and supplies some recommendations for optimization of any ice- core drilling process．

Porosity is one of the most important properties of oil and gas reservoirs． The porosity data that come from well log are only available at well points． It is necessary to use other method to estimate reservoir porosity． Seismic data contain abundant lithological information． Because there are inherent correlations between reservoir property and seismic data，it is possible to estimate reservoir porosity by using seismic data and attributes． Probabilistic neural network is a powerful tool to extract mathematical relation between two data sets． It has been used to extract the mathematical relation between porosity and seismic attributes． Firstly， a seismic imped- ance volume is calculated by seismic inversion． Secondly， several appropriate seismic attributes are extracted by using multi- regression analysis． Then a probabilistic neural network model is trained to obtain a mathematical relation between porosity and seismic attributes． Finally，this trained probabilistic neural network model is im- plemented to calculate a porosity data volume． This methodology could be utilized to find advantageous areas at the early stage of exploration． It is also helpful for the establishment of a reservoir model at the stage of reservoir development．

Petroleum geologists have paid great attentions to the volcanic reservoirs of Songliao Basin in NE Chi- na． There are plenty of subvolcanic rocks in the Songliao Basin accompanying the Early Cretaceous Yingcheng Formation． The logging data show the good reservoir potential of these intrusive rocks but the distribution char- acteristics and formation mechanisms of these reservoirs are not clearly understood． Based on the previous stud- ies by using coring，cuts and logging data of Yingtai rift depression， the reservoirs' characteristics of intrusive rocks are presented． There are two types of intrusive rocks namely the syenodiorite- porphyrite and diabase which occur as laccolith and/or sill，both having the characteristics of low gamma and high density with little primary porosity and permeability． The prevalent reservoir porosity is the secondary porosity，such as spongy/cavernous pore，tectonic fracture． The laboratory data of porosity of diabase can reach 6.7%，but the permeability is less than 0． 6 ×10^-3 μm ² ，median pressure is high，indicating that the pore throat of this kind reservoir is small． The maximum logging porosity is about 12%． The change of porosity does not correlate to the buried depth． It is the major significant differences in the distributive characteristics compared to the normal sedimentary rock reservoirs． Most of intrusive rocks underwent alteration diagenesis whilst some were subjected to precipitation diagenesis． The spongy and cavernous pore can be formed during the alteration processes of plagioclase to illite and pyroxene to chlorite． The secondary porosity is greatly correlated with the alteration intensity of matrix， plagioclase and pyroxene． There are pyroxenes and more plagioclases in diabase，which cause the higher alteration intensity than the syenodiorite- porphyrites in the same acid fluid． So the porosity of diabase is higher than that of syenodiorite- porphyrites． The top or/and bottom part of intrusive rocks develop the higher porosity． Because those parts are easy to contact formation fluid，and the shrink fractures give the more surface for reaction be- tween fluid and rock． The porosity of intrusive rocks is same to the volcanic rocks in Yingtai rift depression and Xujiaweizi rift depression which bear the prolific gas． It suggests good reservoir potential． Intrusive rocks are hosted by the dark mudstone which indicates semi- deep and deep lake facies belt．

When using borehole sensors and microseimic events to image，spatial aliasing often occurred be- cause of the lack of sensors underground and the distance between the sensors which were too large． To solve the aliasing problem，data reconstruction is often needed． Curvelet transform sparsity constrained inversion was widely used in the seismic data reconstruction field for its anisotropic，multiscale and local basis． However，for the downhole case，because the number of sampling point is much larger than the number of the sensors，the advantage of the curvelet basis can't perform very well． To mitigate the problem，the method that joints spline and curvlet- based compressive sensing was proposed． First，we applied the spline interpolation to the first arri- vals that to be interpolated． And the events are moved to a certain direction，such as horizontal，which can be represented by the curvelet basis sparsely． Under the spasity condition，curvelet- based compressive sensing was applied for the data，and directional filter was also used to mute the near vertical noises． After that，the events were shifted to the spline line to finish the interpolation workflow． The method was applied to a synthetic mod- el，and better result was presented than using curvelet transform interpolation directly． We applied the method to a real dataset，a microseismic downhole observation field data in Nanyang，using Kirchhoff migration method to image the microseimic event． Compared with the origin data，artifacts were suppressed on a certain degree．

BP (Back Propagation) neural network and PSO (Particle Swarm Optimization) are two main heu- ristic optimization methods， and are usually used as nonlinear inversion methods in geophysics． The authors ap- plied BP neural network and BP neural network optimized with PSO into the inversion of 3D density interface re- spectively，and a comparison was drawn to demonstrate the inversion results． To start with，a synthetic density interface model was created and we used the proceeding inversion methods to test their effectiveness． And then two methods were applied into the inversion of the depth of Moho interface． According to the results，it is clear to find that the application effect of PSO- BP is better than that of BP network． The BP network structures used in both synthetic and field data are consistent in order to obtain preferable inversion results． The applications in synthetic and field tests demonstrate that PSO- BP is a fast and effective method in the inversion of 3D density interface and the optimization effect is evident compared with BP neural network merely，and thus，this method has practical value．

Based on the accumulated data for the gold deposits in the central Jilin Province in recent years and our understanding of the gold metallogenic province，the Haigou gold metallogenic province is delineated and the denudation degree of gold deposits in this province is discussed． The potential and the ore- searching direc- tion of the province are also considered． The Haigou gold metallogenic province occurs as an independent prov- ince with low denudation degree and high ore- producing potential． Ｒegional fault belts and small basic intru- sions are two ore- constrains and could serve as the ore- searching indictors in the province．

The unconventional oil and gas resources presented in oil shales have meant these potential sources of hydrocarbons，which has become a research focus． China contains abundant oil shale resources，ranking fourth in the world，with ca． 7 254.48 ×10 ⁸ t within 24 provinces，including 48 basins and 81 oil shale deposits． A- bout 48% of the total oil shale resources are concentrated in the eastern resource region，with a further 22% in the central resource region． 65% of the total quantity of oil shale resources is present at depths of 0- 500 m， with 17% of the total resources being defined as high- quality oil shales yielding more than 10% oil by weight． Chinese oil shale resources are generally hosted by Mesozoic sediments that account for 78% of the total resources． In terms of the geographical distribution of these resources，some 45% are located in plain regions， and different oil shale basins have various characteristics． The oil shale resources in China represent a highly prospective future source of hydrocarbons． These resources having potential use not only in power generation and oil refining but also in agriculture，metal and chemical productions，and environmental protection．

The transformation of basic functions is one of the most commonly used techniques for seismic denois- ing，which employs sparse representation of seismic data in the transform domain． The choice of transform base functions has an influence on denoising results． We propose a learning- type overcomplete dictionary based on the K- singular value decomposition (K- SVD) algorithm． To construct the dictionary and use it for random seis- mic noise attenuation，we replace fixed transform base functions with an overcomplete redundancy function library． Owing to the adaptability to data characteristics，the learning- type dictionary describes essential data characteristics much better than conventional denoising methods． The sparsest representation of signals is ob- tained by the learning and training of seismic data． By comparing the same seismic data obtained using the learning- type overcomplete dictionary based on K- SVD and the data obtained using other denoising methods， we find that the learning- type overcomplete dictionary based on the K- SVD algorithm represents the seismic data more sparsely，effectively suppressing the random noise and improving the signal- to- noise ratio．