The load/unload response ratio (LURR) method has been applied for many years in seismicity forecasting and has achieved good results.Over time,it has become the main analysis method in the field of seismomagnetism and is often used in short-term earthquake analysis or prediction.In this study,we used LURR to extract anomalies from digital geomagnetic data recorded at 23 geomagnetic stations during 2008 to 2014 in North China and investigated the correspondence between earthquake occurrence and the anomalies.First,we carefully processed the geomagnetic data to eliminate any distortions caused by missing data and magnetic storms.For example,we examined a few instances of electromagnetic interference around the stations and chose the wavelet method to process the resulting noise.The research region,North China,has numerous evenly distributed geomagnetic stations,and as our research object,we selected 24 earthquakes stronger than M_S4 that occurred in this region.The results show that there were ten groups of LURR anomalies in the seven-year study period in North China,of which seven were thought to be related to earthquakes and the other three were not.In addition,16 of the selected 24 earthquakes had precursory anomalies.We found the maximum anomaly intensity to be proportional to the anomaly area in each group of anomalies.Therefore,the larger the anomaly area,the greater the maximum intensity P(Z) is likely to be.Moreover,we found that seismic energy is associated with the number of stations recording anomalies in one group.Therefore,the greater the number of stations recording anomalies in one group,the larger is the earthquake radiation energy.Lastly,in the groups of anomalies related to earthquakes,we found a relationship between each anomaly,mainly reflected in the position continuity and inheritance.For example,an anomaly was more likely to occur near the position of the preceding anomaly.We also found a relationship between the epicenter of the earthquake and the anomaly location.