Electrical resistivity imaging data for hydrogeological and geological hazard investigations in Taiwan

This data article presents electrical resistivity imaging (ERI) data and inverted models with the objectives of hydrogeological characterization, land subsidence studies, and geological structural detections in Taiwan. The ERI data for hydrogeological studies includes 5 ERI profiles from Changhua, 33 from Yunlin, 36 from Yilan, 23 from Taichung, 23 from Chiayi and Tainan, and 23 from Taipei basins. In addition, time-lapse ERI profiles are presented for 10 ERI from Yilan, 10 ERI from Pingtung, 11 ERI from Taichung, and 31 ERI from Minzu basins. Moreover, 10 ERI data were used to detect the Rusui Fault, 12 for the Qishan Fault, 13 for the Yuli Fault, and 25 for the Shanyi Fault. This data article contains 265 ERI profiles with a total survey length of 59,905 m. A single ERI profile contains hundreds to thousands of subsurface apparent resistivity data points. The data was collected between 2010 and 2022 from different regions of Taiwan. The main findings from the ERI data consisted here were reported by Lin et al. [1] for the Pingtung basin, Chang et al. [2] for the Minzu basin, and Jordi et al. [3] for the Taichung basin in order to estimate hydraulic parameters and characterize the aquifer systems. The ERI data presented here can be used for a variety of hydrogeological, geological, engineering, and environmental applications, and it can be further interpreted using machine learning and statistical methods. Therefore, the ERI data will helps in various subsurface applications, academic research, and educational purposes.


a b s t r a c t
This data article presents electrical resistivity imaging (ERI) data and inverted models with the objectives of hydrogeological characterization, land subsidence studies, and geological structural detections in Taiwan. The ERI data for hydrogeological studies includes 5 ERI profiles from Changhua, 33 from Yunlin, 36 from Yilan, 23 from Taichung, 23 from Chiayi and Tainan, and 23 from Taipei basins. In addition, timelapse ERI profiles are presented for 10 ERI from Yilan, 10 ERI from Pingtung, 11 ERI from Taichung, and 31 ERI from Minzu basins. Moreover, 10 ERI data were used to detect the Rusui Fault, 12 for the Qishan Fault, 13 for the Yuli Fault, and 25 for the Shanyi Fault. This data article contains 265 ERI profiles with a total survey length of 59,905 m. A single ERI profile contains hundreds to thousands of subsurface apparent resistivity data points. The data was collected between 2010 and 2022 from different regions of Taiwan. The main findings from the ERI data consisted here were reported by Lin et al. [1] for the Pingtung basin, Chang et al. [2] for the Minzu basin, and Jordi et al. [3] for the Taichung basin in order to estimate hydraulic parameters and characterize the aquifer systems. The ERI data presented here can be used for a variety of hydrogeological, geological, engineering, and environmental applications, and it can be further interpreted using machine learning and statistical methods. Therefore, the ERI data will helps in various subsurface applications, academic research, and educational purposes. ©

Value of the Data
• The ERI data from the various survey sites can be used to select sites for hydrogeological investigations, identify geological rock units, detect hazardous geological structures, and characterize engineering sites. • The ERI data can be used to monitor groundwater levels by comparing them with future survey findings. • The ERI data can be jointly inverted and interpreted with different geophysical data to obtain more reliable subsurface information. Other studies effectively integrated ERI data with seismic [ 4 ], electromagnetic [ 5 ] and ground penetrating radar [ 6 ]. • Using open-source inversion algorithms, raw ERI data can be reprocessed to generate 2D and 3D inverted models. Several studies [7][8][9][10][11] have successfully used open-source algorithms for ERI data inversion. Machine learning and statistical algorithms can be used to further interpret the inverted resistivity data. ERI data for several studies [12][13][14] have also been made available.

Objective
The primary focus of this work data was hydrogeological, land subsidence, and geological structural investigations such as aquifer characterization, groundwater level detection, groundwater recharge zone identification, hydraulic parameter estimation, clay layer identification, and fault detection [1][2][3]15 ]. The hydraulic parameter was estimated, and groundwater levels were monitored using time-lapse ERI measurement data from the Yilan, Minzu, Taichung, and Pingtung alluvial fans. To determine groundwater potential recharge zones, single survey ERI data from alluvial fans in Taipei, Yunlin, Yilan, Chiayi, and Tainan were collected. The ERI data were collected to detect geological structures such as Titung's Rusui and Yuli Faults, Tainan's Qishan Fault, and Taichung's Shanyi Fault.

Data Description
The electrical resistivity imaging data for hydrogeological and geological hazard investigations conducted in various Taiwan locations are presented. The data can be found at Mendeley's data repository [15] : https://data.mendeley.com/datasets/nkskkcpdg9/2 . The repository is divided into two major folders: ERI Data for Geohazard Detection and ERI Data for Hydeogeological Investigation. The ERI Data for Geohazard Detection is organized into four folders:  Tables 1 and 2 contain detailed data descriptions. Each survey site folder includes a detailed data description; for example, Table 2 shows ERI data descriptions for the Yunlin Chousui River Middle Alluvial Fan. Table 1 shows the total number of survey locations, the number of ERI profiles, and the survey's scope. The data collected by the 4-point light 10W LGM Lippmann resistivity meter is presented in 'URF' format, whereas the data collected by the AGI SuperSting R1 is presented in 'STG' format. If necessary, the ResIPy open-source software [ 8 , 10 ] can be used to convert the file to another format. Inverted resistivity models for each ERI profile are also presented in 'JPG' format. Google Earth KML and KMZ files containing the location of ERI profiles are also provided for each survey site.
Furthermore, the ERI data are described in terms of data collection time (dd/mm/year), profile start, center, and ending coordinates, survey site elevations, profile orientation, array type, electrode spacing, profile length, file types, and names, and the type of resistivity meter used for each survey area. For example, Table 2 shows ERI data descriptions for the Yunlin Chousui River Middle Alluvial Fan.   Fig. 1 depicts the overall ERI profile distributions for hydrogeological, land subsidence, and geological structural studies. It has 5 ERI from Changhua, 33 from Yunlin, 36 from Yilan, 23 from Taichung, 23 from Chiayi and Tainan, 23 from Taipei, 10 for Rusui Fault, 12 for Qishan Fault, 13 for Yuli Fault, and 25 for Shanyi Fault from a single survey. The time-lapse ERI profiles were collected from 10 ERI in Yilan (four time-lapse), 10 ERI in Pingtung (five time-lapse), 11 ERI in Taichung (five time-lapse), and 22 ERI in Minzu (five time-lapse), considering seasonal rainfall variation from Taiwan Central Weather Bureau Meteorological Observatory precipitation record. Yilan ERI data were collected for four months in 2020: February with 40 mm of monthly precipitation, May with 124.2 mm of precipitation, July with 56.0 mm of precipitation, and October with 621.0 mm of precipitation. Pingtung ERI data were collected over five months in 2019: February, May, July, September, and November, with precipitation of 1.0 mm, 5.2 mm, 19.6 mm, 7.1 mm, and 2.3 mm, respectively. Taichung ERI data were collected for five months in 2018: February (22.5 mm), May (73.0 mm), July (347.0 mm), September (20.0 mm), and October (7.5 mm). Furthermore, Minzu ERI data for five different months were collected: January, March, May, June, and September, with monthly precipitation of 10.1 mm, 8.3 mm, 5.0 mm, 16.1 mm, and 6.3 mm, respectively. Three survey sites, for example, are enlarged to show site-specific ERI profile distribution. Fig. 2 shows ERI data distribution for hydrogeological characterization of the Yilan Alluvial Basin. The Yunlin ERI data distribution for artificial groundwater recharge site selection and clay layer detection for land subsidence are presented in Fig. 3 . Besides, Fig. 4 displays the geological structural or fault detection in the Shanyi area, Taichung.

Data collection
The 2D resistivity imaging data were collected by galvanically injecting a low-frequency electrical current into the ground via two current electrodes and measuring the voltage difference between two potential electrodes. Variations in resistivity values caused by the flow of electric current through various subsurface mediums can be used to identify unknown materials. Electrical resistivity of the subsurface material is related to the nature of soil composition (particle size distribution, mineralogy), structure (porosity, pore size distribution, connectivity), fluid content, concentration of dissolved electrolytes, clay contents, and temperature [16] . Fig. 5 depicts the electrical resistivity/conductivity characteristics of common subsurface geological materials.
The advanced multi-electrode resistivity meters were used to measure hundreds to thousands of data points in a single ERI profile by automatically changing the current and potential electrodes. Wenner, Schlumberger, and dipole-dipole arrays were used to collect resistivity data in the normal (forward) and reverse survey directions. Fig. 6 depicts the acquisition of field ERI data with a 4-point light 10W LGM Lippmann ( Fig. 6 a) and an AGI SuperSting R1 ( Fig. 6 b) resistivity meter. Electrodes were placed along the profiles and connected to cables with connector boxes, which were then connected to a resistivity meter during resistivity measurements. The electrodes were then tested for contact resistance before data collection, and apparent resistivity was measured. Then, as shown in Fig. 7 a, hundreds to thousands of apparent resistivity data points can be measured for a single ERI profile.

Data processing and inversion
The measured data was filtered to remove noise. The EarthImager 2D software package [17] was used for data processing and inversion. Depending on the expected subsurface features, the commonly used smoothness-constrained least-square [18] and robust inversion [19] algorithms were used. The ERI data were mostly inverted using a smoothness-constrained leastsquare inversion algorithm. Iterative inversion was used until the misfit between measured and predicted resistivity data reached acceptable levels, where the root mean square (RMS) error is less than 5%, which may exceed for surveys in hard rock and noisy environments. For example, we show the ERI data distribution and inversion models ( Fig. 7 ) for resistivity data collected from the Yunlin Chousui River Alluvial Fan. The data was collected using a Wenner array with electrode spacing of 10 m. Fig. 7 a shows the distributions of apparent resistivity data, Fig. 7 b shows model calculated apparent resistivity data, and Fig. 7 c shows an inverted resistivity model. The overall ERI data acquisition, processing, and inversion approaches are depicted in Fig. 8 .

Ethics Statements
The authors declare that they have no known ethical issues in respect of the data reported in this article.

Declaration of Competing Interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data Availability
Electrical resistivity imaging data for hydrogeological and geological hazard investigations in Taiwan (Original data) (Mendeley Data).