This paper presents a method to integrate remote sensing (RS) data processing, generation of isoseismal lines and human-computer interaction modules into an improved GIS-based disaster reduction system. Using the RS data processing module, a statistical sample gray value and the RS-intensity at each fi eld survey point in the region are calculated from the probabilistic relationship between the RS-variable and earthquake intensity, and stored in the GIS-based system database. Then, isoseismal lines are generated by a trend surface model from RS-intensity. They are further improved via modifi cation of the isoseismal lines based on the empirical attenuation relationship calculated by using the RS-variable in the human-computer interaction module. The fi eld survey shows that the proposed method gives a good generation of isoseismic lines. As a result, the accuracy of the damage and loss evaluation and the effi ciency of the emergency decision making capability are improved. The paper presents a method to integrate remote sensing (RS) data processing, generation of isosemal lines and human-computer interaction modules into an improved GIS-based disaster reduction system. Using the RS data processing module, a statistical sample gray value and the RS -intensity at each fi eld survey point in the region are calculated from the probabilistic relationship between the RS-variable and earthquake intensity, and stored in the GIS-based system database. Then, isoseismal lines are generated by a trend surface model from RS- They are further improved via modifi cation of the isoseismal lines based on the empirical attenuation relationship calculated by using the RS-variable in the human-computer interaction module. The fi eld survey shows that the proposed method gives a good generation of isoseismic lines As a result, the accuracy of the damage and loss evaluation and the effi ciency of the emergency decision making capability are improved.