2018年2月6日，台灣發生6. 4級地震。花蓮成為地震中傷亡最嚴重的城市之一。地震的發生是由於歐亞板塊和菲律賓海板塊之間的聚合運動。 2月6日的地震本身是2月3日開始的幾起地震之一，而2月6日的地震是最強烈的地震。由當地照片顯示市區道路開裂，建築物被損毀並倒塌。道路開裂顯現花蓮市下方有一些新形成的斷層。大地電磁法是一種適合用於探勘較深地下構造的地球物理探勘法。然而，主要的破裂集中在人口集中的花蓮市區，因此收取資料時容易受到人為噪訊干擾而難以取得較好的數據，基於此背景，我們先部署了11個MT站、2個剖面橫跨米崙斷層帶，目的是辨識測站附近人為活動所產生的極低頻噪訊對數據的影響，以利於後續數據分析，並取得較好資料來解釋花蓮地震後的地下構造。本研究收集了2018年7月3日至10日的數據。然後使用Python代碼編寫的程序及一維和二維反演進行分析。目前為止已獲得了各測站的視電阻率-週期圖，部分訊號包含噪訊，因此我們將其辨識、過濾並進行平滑以獲得更好的結果。我們分析各測站中一小時的數據，觀察到未過濾數據和已過濾數據的差異，最後總結出影響數據的主因是受到非極化電極設置及人為活動。;In February 6, 2018, a 6.4 magnitude earthquake hit Taiwan. Hualien became one of the cities that had most damage and casualties by the earthquake. The earthquake occurred because of the converged move between Eurasian Plate and Philippine Sea Plate. There were several photographs that showed cracked roads and some buildings were damaged and collapsed to the ground. Cracked roads suggested there were some newly-formed fault underneath Hualien County. Magnetotelluric (MT) method is one of the geophysics method that can describe the subsurface after the earthquake. However, it will be hard to get a good data due to Hualien County’s dense and crowded area. Based on that hypothesis, we conducted an exploration at Hualien using magnetotelluric method to identify the noise that influenced the data and give better interpretation of Hualien subsurface condition after the earthquake. We deployed 11 MT stations grouped in 2 profiles across the Milun Fault. We deployed those stations overnight in order to get the data with assumed very low noise generated by the electromagnetic wave of people’s activities around the station point. The data acquisition was conducted on July 3-10, 2018. The gathered data from Hualien then were analyzed and processed using several programs written in Python codes and using 1D and 2D inversion. From the data processing so far, we obtained the results of apparent resistivity-versus-period graph for each station. Some part of the signal contained noise, so we identified it, filtered it, and then smoothened it for better result. We analyzed one hour of data for each station, and we can observe the difference of unfiltered and filtered data. We obtained the conclusion that the data was mainly influenced by the noise originated from electrical installations and human activities around the stations.
